Ogden, Charles. The Meaning of Meaning – T01

The meaning of meaning

Chapter I
Thoughts, words and things

Let us get nearer to the fire, so that we can see what we are saying.
— The Bubis of Fernando Po.

Sometimes, in fact, the wise have in this field
proved themselves the most foolish. Was it not the
great Bentley, Master of Trinity College, Cambridge,
Archdeacon of Bristol, and holder of two other livings
besides, who declared: “We are sure, from the names
of persons and places mentioned in Scripture before
the Deluge, not to insist upon other arguments, that
Hebrew was the primitive language of mankind”?
On the opposite page are collected other remarks on
the subject of language and its Meaning, and whether
wise or foolish, they at least raise questions to which,
sooner or later, an answer is desirable. In recent years,
indeed, the existence and importance of this problem
of Meaning have been generally admitted, but by some
sad chance those who have attempted a solution have
too often been forced to relinquish their ambition —
whether through old age, like Leibnitz, or penury, like
C. S. Peirce, or both. Even the methods by which
it is to be attacked have remained in doubt. Each
science has tended to delegate the unpleasant task to
1another. With the errors and omissions of metaphysicians
we shall be much concerned in the sequel,
and philologists must bear their share of the guilt.
Yet it is a philologist who, of recent years, has,
perhaps, realized most clearly the necessity of a broader
treatment.

“Throughout the whole history of the human
race,” wrote the late Dr Postgate, “there have been
no questions which have caused more heart-searchings,
tumults, and devastation than questions of the correspondence
of words to facts. The mere mention of
such words as ‘religion,’ ‘patriotism,’ and ‘property’
is sufficient to demonstrate this truth. Now, it is the
investigation of the nature of the correspondence
between word and fact, to use these terms in the widest
sense, which is the proper and the highest problem of
the science of meaning. That every living word is
rooted in facts of our mental consciousness and history
it would be impossible to gainsay; but it is a very
different matter to determine what these facts may be.
The primitive conception is undoubtedly that the name
is indicative, or descriptive, of the thing. From which
it would follow at once that from the presence of the
name you could argue to the existence of the thing.
This is the simple conception of the savage.”

In thus stressing the need for a clear analysis of the
relation between words and facts as the essential of a
theory of Meaning, Dr Postgate himself was fully aware
that at some point the philosophical and psychological
aspects of that theory cannot be avoided. When he
wrote (1896), the hope was not unreasonable that the
science of Semantics would do something to bridge
the gulf. But, although M. Bréal's researches drew
attention to a number of fascinating phenomena in the
history of language, and awakened a fresh interest in
the educational possibilities of etymology, the net result
was disappointing. That such disappointment was
inevitable may be seen, if we consider the attitude to
2language implied by such a passage as the following.
The use of words as though their meaning were fixed,
the constant resort to loose metaphor, the hypostatization
of leading terms, all indicate an unsuitable attitude in
which to approach the question.

“Substantives are signs attached to things: they contain exactly
that amount of truth which can be contained by a name, an
amount which is of necessity small in proportion to the reality of
the object. That which is most adequate to its object is the
abstract noun, since it represents a simple operation of the
mind. When I use the two words compressibility, immortality,
all that is to be found in the idea is to be found also in the
word. But if I take a real entity, an object existing in nature, it
will be impossible for language to introduce into the word all the
ideas which this entity or object awakens in the mind. Language
is therefore compelled to choose. Out of all the ideas it can
choose one only; it thus creates a name which is not long in
becoming a mere sign.

For this name to be accepted it must, no doubt, originally
possess some true and striking characteristic on one side or
another; it must satisfy the minds of those to whom it is first
submitted. But this condition is imperative only at the outset.
Once accepted, it rids itself rapidly of its etymological signification;
otherwise this signification might become an embarrassment.
Many objects are inaccurately named, whether through the ignorance
of the original authors, or by some intervening change which
disturbs the harmony between the sign and the thing signified.
Nevertheless, words answer the same purpose as though they
were of faultless accuracy. No one dreams of revising them.
They are accepted by a tacit consent of which we are not even
conscious” (Bréal's Semantics, pp. 171-2).

What exactly is to be made of substantives which
“contain” truth, “that amount of truth which can be
contained by a name”? How can “all that is found
in the idea be also found in the word”? The conception
of language as “compelled to choose an
idea,” and thereby creating “a name, which is not
long in becoming a sign,” is an odd one; while
‘accuracy’ and ‘harmony’ are sadly in need of elucidation
when applied to naming and to the relation between
sign and thing signified respectively. This is not
mere captious criticism. The locutions objected to
3conceal the very facts which the science of language
is concerned to elucidate. The real task before that
science cannot be successfully attempted without a far
more critical consciousness of the dangers of such loose
verbiage. It is impossible to handle a scientific matter
in such metaphorical terms, and the training of philologists
has not, as a rule, been such as to increase
their command of analytic and abstract language. The
logician would be far better equipped in this respect
were it not that his command of language tends to
conceal from him what he is talking about and renders
him prone to accept purely linguistic constructions,
which serve well enough for his special purposes, as
ultimates.

How great is the tyranny of language over those
who propose to inquire into its workings is well shown
in the speculations of the late F. de Saussure, a writer
regarded by perhaps a majority of French and Swiss
students as having for the first time placed linguistic
upon a scientific basis. This author begins by inquiring,
“What is the object at once integral and
concrete of linguistic?” He does not ask whether
it has one, he obeys blindly the primitive impulse to
infer from a word some object for which it stands, and
sets out determined to find it. But, he continues, speech
(le langage), though concrete enough, as a set of events
is not integral. Its sounds imply movements of speech,
and both, as instruments of thought, imply ideas. Ideas,
he adds, have a social as well as an individual side,
and at each instant language implies both an established
system and an evolution. “Thus, from whatever side
we approach the question, we nowhere find the integral
object of linguistic.” De Saussure does not pause at
this point to ask himself what he is looking for, or
whether there is any reason why there should be such
a thing. He proceeds instead in a fashion familiar in
the beginnings of all sciences, and concocts a suitable
object — ‘la langue,’ the language, as opposed to speech.
4“What is la langue? For us, it is not to be confounded
with speech (le langage); it is only a determinate part
of this, an essential part, it is true. It is at once a social
product of the faculty, of speech, and a collection of
necessary conventions adopted by the social body to
allow the exercise of this faculty by individuals.
It is a whole in itself and a principle of classification.
As soon as we give it the first place among the facts of
speech we introduce a natural order in a whole which
does not lend itself to any other classification.” La
langue is further “the sum of the verbal images stored
up in all the individuals, a treasure deposited by the
practice of speaking in the members of a given community;
a grammatical system, virtually existing in
each brain, or more exactly in the brains of a body of
individuals; for la langue is not complete in any one
of them, it exists in perfection only in the mass.” 11

Such an elaborate construction as la langue might,
no doubt, be arrived at by some Method of Intensive
Distraction analogous to that with which Dr Whitehead's
name is associated, but as a guiding principle for a
young science it is fantastic. Moreover, the same device
of inventing verbal entities outside the range of possible
investigation proved fatal to the theory of signs which
followed. 225

As a philologist with an inordinate respect for
linguistic convention, de Saussure could not bear to
tamper with what he imagined to be a fixed meaning,
a part of la langue. This scrupulous regard for fictitious
‘accepted’ uses of words is a frequent trait in philologists.
Its roots go down very deep into human nature,
as we shall see in the two chapters which follow. It
is especially regrettable that a technical equipment,
otherwise excellent, should have been so weak at this
point, for the initial recognition of a general science of
signs, ‘semiology,’ of which linguistic would be a
branch, and the most important branch, was a very
notable attempt in the right direction. Unfortunately
this theory of signs, by neglecting entirely the things
for which signs stand, was from the beginning cut off
from any contact with scientific methods of verification.
De Saussure, however, does not appear to have pursued
the matter far enough for this defect to become obvious.
The same neglect also renders the more recent treatise
of Professor Delacroix, Le Langage et la Pensée, ineffective
as a study of the influence of language upon thought.

“All speech,” says Dr Boas explicitly, “is intended
to serve for the communication of ideas.” Ideas, however,
are only remotely accessible to outside inquirers,
and we need a theory which connects words with things
through the ideas, if any, which they symbolize. We
require, that is to say, separate analyses of the relations
of words to ideas and of ideas to things. Further, much
language, especially primitive language, is not primarily
concerned with ideas at all, unless under ‘ideas’ are
included emotions and attitudes — a procedure which
would involve terminological inconveniences. The
omission of all separate treatment of the ways in which
speech, besides conveying ideas, also expresses attitudes,
desires and intentions, 13 is another point at which the
work of this active school is at present defective.7

In yet another respect all these specialists fail to
realize the deficiencies of current linguistic theory. Preoccupied
as they are — ethnologists with recording the
details of fast vanishing languages; philologists with an
elaborate technique of phonetic laws and principles of
derivation; philosophers with ‘philosophy’ — all have
overlooked the pressing need for a better understanding
of what actually occurs in discussion. The analysis of
the process of communication is partly psychological,
and psychology has now reached a stage at which this
part may be successfully undertaken. Until this had
happened the science of Symbolism necessarily remained
in abeyance, but there is no longer any excuse for vague
talk about Meaning, and ignorance of the ways in which
words deceive us.

Throughout the Western world it is agreed that
people must meet frequently, and that it is not only
agreeable to talk, but that it is a matter of common
courtesy to say something even when there is hardly
anything to say. “Every civilized man,” continues
the late Professor Mahaffy, to whose Principles of the
Art of Conversation we owe this observation, “feels, or
ought to feel, this duty; it is the universal accomplishment
which all must practise”; those who fail are
punished by the dislike or neglect of society.

There is no doubt an Art in saying something when
8there is nothing to be said, but it is equally certain that
there is an Art no less important of saying clearly what
one wishes to say when there is an abundance of material;
and conversation will seldom attain even the level
of an intellectual pastime if adequate methods of Interpretation
are not also available.

Symbolism is the study of the part played in human
affairs by language and symbols of all kinds, and
especially of their influence on Thought. It singles out
for special inquiry the ways in which symbols help us
and hinder us in reflecting on things.

Symbols direct and organize, record and communicate.
In stating what they direct and organize,
record and communicate we have to distinguish as
always between Thoughts and Things. 14 It is Thought
(or, as we shall usually say, reference) which is directed
and organized, and it is also Thought which is recorded
and communicated. But just as we say that the gardener
mows the lawn when we know that it is the lawn-mower
which actually does the cutting, so, though we know
that the direct relation of symbols is with thought, we
also say that symbols record events and communicate
facts.

By leaving out essential elements in the language
situation we easily raise problems and difficulties which
vanish when the whole transaction is considered in
greater detail. Words, as every one now knows,
‘mean’ nothing by themselves, although the belief
9that they did, as we shall see in the next chapter, was
once equally universal. It is only when a thinker
makes use of them that they stand for anything, or,
in one sense, have ‘meaning.’ They are instruments.
But besides this referential use which for all reflective,
intellectual use of language should be paramount,
words have other functions which may be grouped
together as emotive. These can best be examined
when the framework of the problem of strict statement
and intellectual communication has been set up. The
importance of the emotive aspects of language is not
thereby minimized, and anyone chiefly concerned with
popular or primitive speech might well be led to reverse
this order of approach. Many difficulties, indeed,
arising through the behaviour of words in discussion,
even amongst scientists, force us at an early stage
to take into account these ‘non-symbolic’ influences.
But for the analysis of the senses of ‘meaning’ with
which we are here chiefly concerned, it is desirable to
begin with the relations of thoughts, words and things
as they are found in cases of reflective speech uncomplicated
by emotional, diplomatic, or other disturbances;
and with regard to these, the indirectness of the
relations between words and things is the feature
which first deserves attention.

This may be simply illustrated by a diagram, in
which the three factors involved whenever any statement
is made, or understood, are placed at the corners
of the triangle, the relations which hold between them
being represented by the sides. The point just made
can be restated by saying that in this respect the base
of the triangle is quite different in composition from
either of the other sides.

Between a thought and a symbol causal relations
hold. When we speak, the symbolism we employ is
caused partly by the reference we are making and
partly by social and psychological factors — the purpose
for which we are making the reference, the proposed
10effect of our symbols on other persons, and our own
attitude. When we hear what is said, the symbols
both cause us to perform an act of reference and to
assume an attitude which will, according to circumstances,
be more or less similar to the act and the
attitude of the speaker.

image thought or reference | symbol | referent | correct, symbolyses (a causal relation) | adequate, refers to (other causal relations | stands for (an imputed relation) true *5

Between the Thought and the Referent there is also
a relation; more or less direct (as when we think about
or attend to a coloured surface we see), or indirect (as
when we ‘think of’ or ‘refer to’ Napoleon), in which
case there may be a very long chain of sign-situations
intervening between the act and its referent: word —
historian — contemporary record — eye-witness — referent
(Napoleon).

Between the symbol and the referent there is no
relevant relation other than the indirect one, which
consists in its being used by someone to stand for a
referent. Symbol and Referent, that is to say, are not
connected directly (and when, for grammatical reasons,
we imply such a relation, it will merely be an imputed, 16
11as opposed to a real, relation) but only indirectly
round the two sides of the triangle. 17

It may appear unnecessary to insist that there is
no direct connection between say ‘dog,’ the word, and
certain common objects in our streets, and that the
only connection which holds is that which consists in
our using the word when we refer to the animal. We
shall find, however, that the kind of simplification
typified by this once universal theory of direct meaning
relations between words and things is the source of
almost all the difficulties which thought encounters.
As will appear at a later stage, the power to confuse
and obstruct, which such simplifications possess, is
largely due to the conditions of communication.
Language if it is to be used must be a ready instrument.
The handiness and ease of a phrase is always more
important in deciding whether it will be extensively
used than its accuracy. Thus such shorthand as the
word ‘means’ is constantly used so as to imply a direct
simple relation between words and things, phrases and
situations. If such relations could be admitted then
there would of course be no problem as to the nature
12of Meaning, and the vast majority of those who have
been concerned with it would have been right in their
refusal to discuss it. But too many interesting developments
have been occurring in the sciences, through the
rejection of everyday symbolizations and the endeavour
to replace them by more accurate accounts, for any
naive theory that ‘meaning’ is just ‘meaning’ to be
popular at the moment. As a rule new facts in startling
disagreement with accepted explanations of other facts
are required before such critical analyses of what are
generally regarded as simple satisfactory notions are
undertaken. This has been the case with the recent
revolutions in physics. But in addition great reluctance
to postulate anything sui generis and of necessity undetectable 18
was needed before the simple natural notion
of simultaneity, for instance, as a two-termed relation
came to be questioned. Yet to such questionings the
theory of Relativity was due. The same two motives,
new discrepant facts, and distaste for the use of obscure
kinds of entities in eking out explanations, have led to
disturbances in psychology, though here the required
restatements have not yet been provided. No
Copernican revolution has yet occurred, although
several are due if psychology is to be brought into line
with its fellow sciences.

It is noteworthy, however, that recent stirrings in
psychology have been mainly if not altogether concerned
with feeling and volition. The popular success
of Psycho-analysis has tended to divert attention from
the older problem of thinking. Yet in so far as progress
here has consequences for all the other sciences
and for the whole technique of investigation in
psychology itself, this central problem of knowing or
of ‘meaning’ is perhaps better worth scrutiny and more
likely to promote fresh orientations than any other that
can be suggested. As the Behaviorists have also very
13properly pointed out, this question is closely connected
with the use of words.

But the approach to Meaning, far more than the
approach to such problems as those of physics, requires
a thorough-going investigation of language. Every
great advance in physics has been at the expense of
some generally accepted piece of metaphysical explanation
which had enshrined itself in a convenient,
universally practised, symbolic shorthand. But the
confusion and obstruction due to such shorthand
expressions and to the naive theories they protect and
keep alive, is greater in psychology, and especially in
the theory of knowledge, than elsewhere; because no
problem is so infected with so-called metaphysical
difficulties — due here, as always, to an approach to a
question through symbols without an initial investigation
of their functions.

We have now to consider more closely what the
causes and effects of symbols are. 19 Whatever may be
the services, other than conservative and retentive, of
symbolization, all experience shows that there are also
disservices. The grosser forms of verbal confusion
have long been recognized; but less attention has been
paid to those that are more subtle and more frequent.
In the following chapters many examples of these will
be given, chosen in great part from philosophical fields,
for it is here that such confusions become, with the
passage of time, most apparent. The root of the trouble
will be traced to the superstition that words are in some
way parts of things or always imply things corresponding
to them, historical instances of this still potent
14instinctive belief being given from many sources. The
fundamental and most prolific fallacy is, in other words,
that the base of the triangle given above is filled in.

The completeness of any reference varies; it is more
or less close and clear, it ‘grasps’ its object in greater
or less degree. Such symbolization as accompanies
it — images of all sorts, words, sentences whole and in
pieces — is in no very close observable connection with
the variation in the perfection of the reference. Since,
then, in any discussion we cannot immediately settle
from the nature of a person's remarks what his opinion
is, we need some technique to keep the parties to an
argument in contact and to clear up misunderstandings
— or, in other words, a Theory of Definition. Such a
technique can only be provided by a theory of knowing,
or of reference, which will avoid, as current theories do
not, the attribution to the knower of powers which it
may be pleasant for him to suppose himself to possess,
but which are not open to the only kind of investigation
hitherto profitably pursued, the kind generally known
as scientific investigation.

Normally, whenever we hear anything said we
spring spontaneously to an immediate conclusion,
namely, that the speaker is referring to what we should
be referring to were we speaking the words ourselves.
In some cases this interpretation may be correct; this
will prove to be what he has referred to. But in most
discussions which attempt greater subtleties than could
be handled in a gesture language this will not be so.
To suppose otherwise is to neglect our subsidiary
gesture languages, whose accuracy within their own
limited provinces is far higher than that yet reached
by any system of spoken or written symbols, with the
exception of the quite special and peculiar case of
mathematical, scientific and musical notations. Words,
whenever they cannot directly ally themselves with and
support themselves upon gestures, are at present a very
imperfect means of communication. Even for private
15thinking thought is often ready to advance, and only
held back by the treachery of its natural symbolism;
and for conversational purposes the latitude acquired
constantly shows itself to all those who make any
serious attempts to compare opinions.

We have not here in view the more familiar ways
in which words may be used to deceive. In a later
chapter, when the function of language as an instrument
for the promotion of purposes rather than as a
means of symbolizing references is fully discussed, we
shall see how the intention of the speaker may complicate
the situation. But the honnête homme may be
unprepared for the lengths to which verbal ingenuity
can be carried. At all times these possibilities have
been exploited to the full by interpreters of Holy Writ
who desire to enjoy the best of both worlds. Here,
for example, is a specimen of the exegetic of the late
Dr Lyman Abbott, pastor, publicist, and editor, which,
through the efforts of Mr Upton Sinclair, has now
become classic. Does Christianity condemn the
methods of twentieth-century finance? Doubtless there
are some awkward words in the Gospels, but a little
‘interpretation’ is all that is necessary.

“Jesus did not say ‘Lay not up for yourselves treasures upon
earth.’ He said ‘Lay not up for yourselves treasures upon earth
where moth and rust doth corrupt and where thieves break through
and steal.’ And no sensible American does. Moth and rust do
not get at Mr Rockefeller's oil wells, and thieves do not often
break through and steal a railway. What Jesus condemned was
hoarding wealth.”

Each investment, therefore, every worldly acquisition,
according to one of the leading divines of the
New World, may be judged on its merits. There
is no hard and fast rule. When moth and rust have
been eliminated by science the Christian investor will
presumably have no problem, but in the meantime it
would seem that Camphorated Oil fulfils most nearly
the synoptic requirements. Burglars are not partial
16to it; it is anathema to moth; and the risk of rust is
completely obviated.

Another variety of verbal ingenuity closely allied
to this, is the deliberate use of symbols to misdirect
the listener. Apologies for such a practice in the
case of the madman from whom we desire to conceal
the whereabouts of his razor are well known, but a
wider justification has also been attempted. In the
Christian era we hear of “falsifications of documents,
inventions of legends, and forgeries of every description
which made the Catholic Church a veritable seat of
lying.” 110 A play upon words in which one sense is
taken by the speaker and another sense intended by
him for the hearer was permitted. 211 Indeed, three sorts
of equivocations were distinguished by Alfonso de
Liguori, who was beatified in the nineteenth century,
which might be used with good reason; 312 a good reason
being “any honest object, such as keeping our goods,
spiritual or temporal.” 413 In the twentieth century the
intensification of militant nationalism has added further
‘good reason’; for the military code includes all
transactions with hostile nations or individuals as part
of the process of keeping spiritual and temporal goods.
In war-time words become a normal part of the
mechanism of deceit, and the ethics of the situation
have been aptly summed up by Lord Wolseley: “We
will keep hammering along with the conviction that
‘honesty is the best policy,’ and that truth always
wins in the long run. These pretty sentences do
well for a child's copy-book, but the man who acts
upon them in war had better sheathe his sword for
ever.” 51417

The Greeks, as we shall see, were in many ways
not far from the attitude of primitive man towards
words. And it is not surprising to read that after the
Peloponnesian war the verbal machinery of peace had
got completely out of gear, and, says Thucydides,
could not be brought back into use — “The meaning
of words had no longer the same relation to things,
but was changed by men as they thought proper.”
The Greeks were powerless to cope with such a situation.
We in our wisdom seem to have created institutions
which render us more powerless still. 115

But even where the actual referents are not in doubt,
it is perhaps hardly realized how widespread is the
18habit of using the power of words not only for bona fide
communications, but also as a method of misdirection;
and in the world as it is to-day the naïve interpreter
is likely on many occasions to be seriously misled if
the existence of this unpleasing trait — equally prevalent
amongst the classes and the masses without distinction
of race, creed, sex, or colour — is overlooked.

Throughout this work, however, we are treating of
bona fide communication only, except in so far as we
shall find it necessary in Chapter IX. to discuss that
derivate use of Meaning to which misdirection gives
rise. For the rest, the verbal treachery with which
we are concerned is only that involved by the use of
symbols as such. As we proceed to examine the
conditions of communication we shall see why any
symbolic apparatus which is in general use is liable to
incompleteness and defect.

But if our linguistic outfit is treacherous, it nevertheless
is indispensable, nor would another complete
outfit necessarily improve matters, even if it were ten
times as complete. It is not always new words that
are needed, but a means of controlling them as symbols,
a means of readily discovering to what in the world
on any occasion they are used to refer, and this is what
an adequate theory of definition should provide.

But a theory of Definition must follow, not precede,
a theory of Signs, and it is little realized how large a
place is taken both in abstract thought and in practical
affairs by sign-situations. But if an account of sign-situations
is to be scientific it must take its observations
from the most suitable instances, and must not derive
its general principles from an exceptional case. The
person actually interpreting a sign is not well placed
for observing what is happening. We should develop
our theory of signs from observations of other people,
and only admit evidence drawn from introspection when
we know how to appraise it. The adoption of the
other method, on the ground that all our knowledge of
19others is inferred from knowledge of our own states,
can only lead to the impasse of solipsism from which
modern speculation has yet to recoil. Those who allow
beyond question that there are people like themselves
also interpreting signs and open to study should not
find it difficult to admit that their observation of the
behaviour of others may provide at least a framework
within which their own introspection, that special and
deceptive case, may be fitted. That this is the practice
of all the sciences need hardly be pointed out. Any
sensible doctor when stricken by disease distrusts his
own introspective diagnosis and calls in a colleague.

There are, indeed, good reasons why what is
happening in ourselves should be partially hidden
from us, and we are generally better judges of what
other people are doing than of what we are doing
ourselves. Before we looked carefully into other
people's heads it was commonly believed that an
entity called the soul resided therein, just as children
commonly believe that there is a little man inside the
skull who looks out at the eyes, the windows of the
soul, and listens at the ears. The child has the
strongest introspective evidence for this belief, which,
but for scalpels and microscopes, it would be difficult
to disturb. The tacitly solipsistic presumption that
this naïve approach is in some way a necessity of
method disqualifies the majority of philosophical and
psychological discussions of Interpretation. If we
restrict the subject-matter of the inquiry to ‘ideas’
and words, i.e., to the left side of our triangle, and
omit all frank recognition of the world outside us, we
inevitably introduce confusion on such subjects as
knowledge in perception, verification and Meaning
itself. 11720

If we stand in the neighbourhood of a cross road
and observe a pedestrian confronted by a notice To
Grantchester displayed on a post, we commonly distinguish
three important factors in the situation. There
is, we are sure, (1) a Sign which (2) refers to a Place
and (3) is being interpreted by a person. All situations in
which Signs are considered are similar to this. A doctor
noting that his patient has a temperature and so forth
is said to diagnose his disease as influenza. If we talk
like this we do not make it clear that signs are here
also involved. Even when we speak of symptoms we
often do not think of these as closely related to other
groups of signs. But if we say that the doctor
interprets the temperature, etc., as a Sign of influenza,
we are at any rate on the way to an inquiry as to
whether there is anything in common between the
manner in which the pedestrian treated the object at
the cross road and that in which the doctor treated
his thermometer and the flushed countenance.

On close examination it will be found that very
many situations which we do not ordinarily regard as
Sign-situations are essentially of the same nature. The
chemist dips litmus paper in his test-tube, and interprets
the sign red or the sign blue as meaning acid or base.
A Hebrew prophet notes a small black cloud, and
remarks “We shall have rain.” Lessing scrutinizes
the Laocoön, and concludes that the features of Laocoon
père are in repose. A New Zealand school-girl
looks at certain letters on a page in her Historical
Manual for the use of Lower Grades and knows that
Queen Anne is dead.

The method which recognizes the common feature
of sign-interpretation 118 has its dangers, but opens the
21way to a fresh treatment of many widely different
topics.

As an instance of an occasion in which the theory
of signs is of special use, the subject dealt with in our
fourth chapter may be cited. If we realize that in all
perception, as distinguished from mere awareness, sign-situations
are involved, we shall have a new method
of approaching problems where a verbal deadlock seems
to have arisen. Whenever, we ‘perceive’ what we
name ‘a chair,’ we are interpreting a certain group
of data (modifications of the sense-organs), and treating
them as signs of a referent. Similarly, even before the
interpretation of a word, there is the almost automatic
interpretation of a group of successive noises or letters
as a word. And in addition to the external world we
can also explore with a new technique the sign-situations
involved by mental events, the ‘goings on’ or processes
of interpretation themselves. We need neither
confine ourselves to arbitrary generalizations from introspection
after the manner of classical psychology, nor
deny the existence of images and other ‘mental’ occurrences
to their signs with the extreme Behaviorists. 119
The Double language hypothesis, which is suggested
by the theory of signs and supported by linguistic
analysis, would absolve Dr Watson and his followers
22from the logical necessity of affecting general anæsthesia.
Images, etc., are often most useful signs of our present
and future behaviour — notably in the modern interpretation
of dreams. 120 An improved Behaviorism will have
much to say concerning the chaotic attempts at symbolic
interpretation and construction by which Psycho-analysts
discredit their valuable labours.

The problems which arise in connection with any
‘sign-situation’ are of the same general form. The
relations between the elements concerned are no doubt
different, but they are of the same sort. A thorough
classification of these problems in one field, such as the
field of symbols, may be expected, therefore, to throw
light upon analogous problems in fields at first sight
of a very different order.

When we consider the various kinds of Sign-situations
instanced above, we find that those signs which
men use to communicate one with another and as
instruments of thought, occupy a peculiar place. It
is convenient to group these under a distinctive name;
and for words, arrangements of words, images, gestures,
and such representations as drawings or mimetic sounds
we use the term symbols. The influence of Symbols
upon human life and thought in numberless unexpected
ways has never been fully recognized, and to this chapter
of history we now proceed.23

Chapter II
The power of words

Le mot, qu'on le sache, est un être vivant… le mot
est le verbe, et le verbe est Dieu. — Victor Hugo.

Athenians! I observe that in all respects you are
deeply reverential towards the Gods. — Paul of Tarsus.

He who shall duly consider these matters will find that
there is a certain bewitchery or fascination in words,
which makes them operate with a force beyond what
we can naturally give account of. — South.

From the earliest times the Symbols which men have
used to aid the process of thinking and to record their
achievements have been a continuous source of wonder
and illusion. The whole human race has been so
impressed by the properties of words as instruments
for the control of objects, that in every age it has
attributed to them occult powers. Between the attitude
of the early Egyptian and that of the modern poet,
there would appear at first sight to be but little difference.
“All words are spiritual,” says Walt Whitman,
“nothing is more spiritual than words. Whence are
they? Along how many thousands and tens of thousands
of years have they come?” Unless we fully realize the
profound influence of superstitions concerning words,
we shall not understand the fixity of certain widespread
linguistic habits which still vitiate even the most careful
thinking.

With the majority, and in matters of ordinary discussion,
the influence of this legacy is all-pervasive, in
language no less than in other spheres. “If we could
open the heads and read the thoughts of two men of
24the same generation and country, but at the opposite
ends of the intellectual scale, we should probably find
their minds as different as if the two belonged to different
species… Superstitions survive because, while they
shock the views of enlightened members of the community,
they are still in harmony with the thoughts
and feelings of others, who, though they are drilled by
their betters into an appearance of civilization, remain
barbarians or savages at heart.” 121

Most educated people are quite unconscious of the
extent to which these relics survive at their doors, still
less do they realize how their own behaviour is moulded
by the unseen hand of the past. “Only those whose
studies have led them to investigate the subject,” adds
Dr Frazer, “are aware of the depth to which the ground
beneath our feet is thus, as it were, honeycombed by
unseen forces.”

The surface of society, like that of the sea, may,
the anthropologist admits, be in perpetual motion, but
its depths, like the depths of the ocean, remain almost
unmoved. Only by plunging daily into those depths
can we come in contact with our fellow-men; only —
in the particular case of language — by forgoing the
advantages of this or that special scientific symbol
system, by drinking of the same unpurified stream, can
we share in the life of the community. If the clouds of
accumulated verbal tradition burst above us in the
open — in the effort to communicate, in the attempt at
interpretation — few have, as yet, evolved even the
rudiments of a defence.

The power of words is the most conservative force
in our life. Only yesterday did students of anthropology
begin to admit the existence of those ineluctable
verbal coils by which so much of our thought is
encompassed. “The common inherited scheme of
conception which is all around us, and comes to us as
naturally and unobjectionably as our native air, is none
25the less imposed upon us, and limits our intellectual
movements in countless ways — all the more surely and
irresistibly because, being inherent in the very language
we must use to express the simplest meaning, it is
adopted and assimilated before we can so much as begin
to think for ourselves at all.” 122 And from the structure
of our language we can hardly even think of escaping.
Tens of thousands of years have elapsed since we shed
our tails, but we are still communicating with a medium
developed to meet the needs of arboreal man. And
as the sounds and marks of language bear witness to
its primeval origins, so the associations of those sounds
and marks, and the habits of thought which have grown
up with their use and with the structures imposed on
them by our first parents, are found to bear witness to
an equally significant continuity.

We may smile at the linguistic illusions of primitive
man, but may we forget that the verbal machinery on
which we so readily rely, and with which our metaphysicians
still profess to probe the Nature of Existence,
was set up by him, and may be responsible for other
illusions hardly less gross and not more easily eradicable?
It may suffice at this point to recall the prevalence of
sacred or secret vocabularies, and of forbidden words
of every sort. Almost any European country can still
furnish examples of the tale in which a name (Tom-Tit-Tot,
Vargaluska, Rumpelstiltskin, Finnur, Zi) has to be
discovered before some prince can be wedded, or some
ogre frustrated. 223 And on the contextual account of
reference which is the outcome of modern developments
of associationism, with its immense stress on the part
played by language in memory and imagination, it is
clear that in the days before psychological analysis
was possible the evidence for a special world of words
26of power, for nomina as numina, must have appeared
overwhelming.

In ancient Egypt precautions were taken to prevent
the extinction of the eighth or Name-soul, and to cause
its continuance along with the names of the Gods. 124 In
the Pyramid texts we find mentioned a God called
Khern, i.e., Word: the Word having a personality like
that of a human being. The Creation of the world
was due to the interpretation in words by Thoth of the
will of the deity. The greater part of mankind must
once have believed the name to be that integral part
of a man identified with the soul, or to be so important
a portion of him that it might be substituted for the
whole, as employers speak of factory ‘hands.’ In
Revelation we read “There were killed in the earthquake
names of men seven thousand,” and again in the letter
to the Church of Sardis, “Thou hast a few names in
Sardis which did not defile their garments.” The
beast coming up out of the sea has upon his head
“names of blasphemy.” Blasphemy itself is just such
an instance; for the god is supposed to be personally
offended by the desecration of his name: and even in
the reign of Henry VIII. a boy was put to death by
burning because of some idle words he had chanced to
hear respecting the sacrament — which he ignorantly
repeated. 225

“Why askest thou after my name, seeing it is
secret” (or ‘ineffable’ with Prof. G. F. Moore), says the
angel of the Lord to Manoah in the book of Judges.
Nearly all primitive peoples show great dislike to their
names being mentioned; when a New Zealand chief
was called Wai, which means water, a new name had
to be given to water; and in Frazer's Golden Bough
numerous examples of word taboos are collected to
show the universality of the attitude. Not only chiefs
but gods, and moreover the priest in whom gods were
27supposed to dwell (a belief which induced the Cantonese
to apply the term ‘god-boxes’ to such favoured personages),
are amongst the victims of this logophobia.
We know how Herodotus (II. 132, 171) refuses to
mention the name of Osiris. The true and great name
of Allah is a secret name, 126 and similarly with the gods
of Brahmanism 227 and the real name of Confucius. 328
Orthodox Jews apparently avoid the name Jahweh
altogether. 429 We may compare ‘Thank Goodness’
‘Morbleu’ — and the majority of euphemisms. Among
the Hindus if one child has been lost, it is customary
to call the next by some opprobrious name. A male
child is called Kuriya, or Dunghill — the spirit of
course knows folk as their names and will overlook the
worthless. Similarly, God knows each man by his
name — “and the Lord said unto Moses ‘Thou hast
found grace in my sight and I know thee by thy name.’”
Every ancient Egyptian had two names — one for the
world, and another by which he was known to the
supernal powers. The Abyssinian Christian's second
name, given at baptism, is never to be divulged. The
guardian deity of Rome had an incommunicable name,
and in parts of ancient Greece the holy names of the
gods to ensure against profanation were engraved on
lead tablets and sunk in the sea.

Children are often similarly anxious to conceal their
names; and just as children always demand what the
name of a thing is (never if it has a name) and regard
that name as a valuable acquisition, so we know that
the stars all have names. “He telleth the number of
the stars and calleth them all by their names.” Here
we may note the delightful proverb which might appear
on the title-page of every work dealing with Symbolism:
“The Divine is rightly so called.”28

In some ways the twentieth century suffers more
grievously than any previous age from the ravages of
such verbal superstitions. Owing, however, to developments
in the methods of communication, and the creation
of many special symbolic systems, the form of the
disease has altered considerably; and, apart from the
peculiar survival of religious apologetic, now takes
more insidious forms than of yore. Influences making
for its wide diffusion are the baffling complexity of the
symbolic apparatus now at our disposal; the possession
by journalists and men of letters of an immense semi-technical
vocabulary and their lack of opportunity, or
unwillingness, to inquire into its proper use; the success
of analytic thinkers in fields bordering on mathematics,
where the divorce between symbol and reality is most
pronounced and the tendency to hypostatization most
alluring; the extension of a knowledge of the cruder
forms of symbolic convention (the three R's), combined
with a widening of the gulf between the public and
the scientific thought of the age; and finally the exploitation,
for political and commercial purposes, of
the printing press by the dissemination and reiteration
of clichés.

With the aid of this strange verbal rapier many
palpable hits were claimed. Thus the theory of
“adjectives or attributes or ideal things in some way
less substantial, less self subsistent, less self identical,
than true substantives, appears to be wholly erroneous”; 333
whole philosophical systems were excluded, for “the
admission (involved in the mention of a man and a
chimera) of many terms destroys monism”; 434 and a
modern Platonism reconstructed, whereby a world of
certain of the ‘things’ ‘mentioned’ by means of ‘terms’
the world of universals, was rehabilitated. Here the
reason builds a habitation, “or rather finds a habitation
eternally standing, where our ideals are fully satisfied
and our best hopes are not thwarted. It is only when
we thoroughly understand the entire independence of
ourselves, which belongs to this world that reason finds,
that we can adequately realize the profound importance
of its beauty.” 535 For here everything is “unchangeable,
rigid, exact, delightful to the mathematician, the
logician, the builder of metaphysical systems, and all
who love perfection more than life.” This world was
commended to the working man, in contrast to the
30world of existence which is “fleeting, vague, without
sharp boundaries, without any clear plan or arrangement
though it contains all thoughts and feelings.”
Both worlds are equally there, equally worth contemplation,
and “according to our temperaments, we
shall prefer the contemplation of the one or of the
other.” 136

It is regrettable that modern Platonists so seldom
follow Plato in his attempts at a scientific study of
Symbolism, but it is interesting to note that they
recognize the kinship of their theory with Greek
speculation, for both have their origin in the same
linguistic habits. The ingenuity of the modern logician
tends to conceal the verbal foundations of his
structure, but in Greek philosophy these foundations
are clearly revealed. The earlier writers are full of
the relics of primitive word-magic. To classify things
is to name them, and for magic the name of a thing
or group of things is its soul; to know their names
is to have power over their souls. Nothing, whether
human or superhuman, is beyond the power of words.
Language itself is a duplicate, a shadow-soul, of the
whole structure of reality. Hence the doctrine of the
Logos, variously conceived as this supreme reality, the
divine soul-substance, as the ‘Meaning’ or reason of
everything, and as the ‘Meaning’ or essence of a name. 237

The Greeks were clearly assisted in their acceptance
of an Otherworld of Being by the legacy of religious
material which earlier philosophers incorporated in
their respective systems. The nature of things, their
physis, was regarded, e.g., by Thales, as supersensible,
a stuff of that attenuated sort which has always been
attributed to souls and ghosts; differing from body
31only in being intangible and invisible. Consequently
the World of Being, in which bogus entities reside,
had at first that minimum of materiality without which
nothing could be conceived. But as logic developed
and the power of words attracted more attention, this
materiality was gradually lost, until in the Symposium,
211, and the Phaedo, 80, Plato has evolved a realm of
pure ideality, also described as physis, in which these
name-souls dwell, pure, divine, immortal, intelligible,
uniform, indissoluble and unchangeable.

This development has been shown to be due largely
to the influence of Pythagoreanism and the intervening
stages are of peculiar interest for the history of Symbols.
It was Heracleitus who first appealed to words as
embodying the nature of things, and his influence on
Plato is manifest in the Cratylus. Heracleitus saw in
language the most constant thing in a world of ceaseless
change, an expression of that common wisdom
which is in all men; and for him the structure of human
speech reflects the structure of the world. It is an embodiment
of that structure — “the Logos is contained
and in it, as one meaning may be contained in many
outwardly different symbols.” 138

The Pythagoreans on the other hand were chiefly
puzzled by number symbols. “Since everything
appeared to be modelled in its entire character on
numbers,” says Aristotle, 239 “and numbers to be the
ultimate things in the whole universe, they became
convinced that the elements of numbers are the elements
of everything.” In fact, in its final stages, Pythagoreanism
passed from a doctrine of the world as a
procession of numbers out of the One, to the construction
of everything out of Number-souls, each
claiming an immortal and separate existence. 34032

Parmenides, who followed, was occupied with the
functions of negative symbols. If ‘Cold’ only means
the same as ‘not hot,’ and ‘dark’ the same as ‘not
light,’ how can we talk about absences of things?
“Two bodies there are,” he says, “which mortals have
decided to name, one of which they ought not to name,
and that is where they have gone wrong.” They have
given names to things which simply are not, to the
not-things (μὴ ἐόν). But in addition to the problem of
Negative Facts, which involved Plato in the first
serious examination of the relations of thought and
language (Sophist, 261), Parmenides handed on to
Plato his own Orphic conundra about the One and
the Many, which also have their roots in language. So
that, quite apart from the difficulties raised by his Ideal
World where the Name-souls dwelt, and its relations
with the world of mud and blood (to which entities
on æsthetic grounds he hesitated to allow ‘ideas,’
much as theologians debated the existence of souls in
darkies), Plato had every reason to be occupied by
linguistic theory.

It is, therefore, all the more unfortunate that the
dialogue, The Cratylus, in which his views on language
are set forth, should have been so neglected in modern
times. Plato's theory of Ideas or Name-souls was
accepted from the Pythagoreans; but as a scientist he
was constantly approaching the problem of names and
their meaning as one of the most difficult inquiries which
could be encountered. His analysis, in an age when
comparative philology, grammar, and psychology were
all unknown, is a remarkable achievement, but he fails
to distinguish consistently between symbols and the
thought symbolized.33

The main tradition of Greek speculation remained
Faithful to the verbal approach. There are two ways,
wrote Dr Whewell, of comprehending nature, “the
one by examining the words only and the thoughts
which they call up; the other by attending to the facts
and things which bring these notions into being…
The Greeks followed the former, the verbal or notional
course, and railed.” And again, “The propensity to
seek for principles in the common usages of language
may be discovered at a very early period… In
Aristotle we have the consummation of this mode of
speculation.” 141 It has been generally accepted since
the time of Trendelenburg 242 that the Categories, and
similar distinctions which play a large part in Aristotle's
system, cannot be studied apart from the peculiarities
of the Greek language. “Aristotle,” says Gomperz,
“often suffers himself to be led by the forms of
language, not always from inability to free himself from
those bonds, but at least as often because the demands
of dialectic will not allow him to quit his arena…
Thus a distinction is drawn between knowledge in
general and the particular sciences, based solely on the
fact that the objects of the latter are included in their
names… His classification of the categories is
frequently governed by considerations of linguistic
expediency, a circumstance which, it must be allowed
(sic), ought to have restrained him from applying it to
ontological purposes.” 343

The practice of dialectical disputation in Aristotle's
time was based on the notion of a definite simple
meaning for every term, as we see from the Scholia of
Ammonius to the De Interpretatione. Thus the questioner
34asked, “Is Rhetoric estimable?”; and in one
form of the game, at any rate, the respondent was
expected to answer simply Yes or No. Certain words
were regarded as equivocal, chiefly as a result of
studying their ‘contraries,’ in the current vocabulary.
Aristotle enumerates various rules with regard to equivocation
and other devices conceived with the object of
driving an opponent into some form of verbal inconsistency,
in his Topics.

It is curious that in the De Interpretatione Aristotle
puts forward views which are hard to reconcile with
such a verbal approach. He there insists that words
are signs primarily of mental affections, and only
secondarily of the things of which these are likenesses. 346
35And he elaborates a theory of the proposition which,
though incomplete and a source of endless confusion,
yet indicates a far more critical attitude to language
than his logical apparatus as a whole would suggest.
For here Aristotle finds no difficulty in settling the
main question raised by Plato in the Cratylus. All
significant speech, he says, is significant by convention
only, and not by nature or as a natural instrument —
thereby neglecting Plato's acute observations as to the
part played by onomatopœia in verbal origins. In the
De Interpretatione various branches of significant speech
are deliberately excluded, and we are there invited to
consider only that variety known as enunciative, which,
as declaring truth or falsehood, is all that belongs to
Logic; other modes of speech, the precative, imperative,
interrogative, etc., being more naturally regarded as
part of Rhetoric or Poetic. 147

In their levies, Cicero informs us, they took care
“to enrol first such names as Victor, and Felix, and
Faustus, and Secundus; and were anxious to head the
roll of the census with a word of such happy augury
as Salvius Valerius. Cæsar gave a command in Spain
to an obscure Scipio simply for the sake of the omen
which his name involved. Scipio upbraids his mutinous
soldiers with having followed an Atrius Umber, a ‘dux
abominandi nominis,’ being, as De Quincey calls him,
a ‘pleonasm of darkness.’ The Emperor Severus
consoled himself for the immoralities of his Empress
Julia, because she bore the same name as the profligate
daughter of Augustus”; 148 just as Adrian VI., when he
became Pope, was persuaded by his Cardinals not to
retain his own name, on the ground that all Popes who
had done so had died in the first year of their reign. 249

When we reflect on the influences which might have
concentrated the attention of Græco-Roman thinkers
on linguistic problems, it is at first sight surprising
that many of those whose constructions were so largely
verbal were also in certain respects fully aware of the
misleading character of their medium. The appeal of
the Heracliteans to language as evidence for the doctrine
of Change was, as we know from the Cratylus, vigorously
opposed by the Parmenidean logicians, as well as by
believers in the Ideas. And an equal readiness to
admit that the presuppositions of Language have to be
combated was manifested by Plotinus. Language, in
the Neo-Platonic view, “can only be made to express
the nature of the soul by constraining it to purposes
for which most men never even think of employing
it”; moreover, “the soul cannot be described at all
except by phrases which would be nonsensical if applied
to body or its qualities, or to determinations of
particular bodies.” 35037

The Buddhists, whose attitude towards language
was exceptional, were quite ready to make use of
customary phrases for popular exposition, but it is not clear
whether any more subtle approach to fictional problems had
been developed. 252

But though all the post-Aristotelian schools, and
particularly the Stoics, whose view of language had
considerable influence on Roman jurists, 353 devoted some
attention to linguistic theory, nowhere in ancient times
do we find evidence of these admissions leading to a
study of symbols such as Plato and Aristotle seemed
at times to be approaching. As we shall see, this was
owing to the lack of any attempt to deal with signs as
such, and so to understand the functions of words in
relation to the more general sign-situations on which
all thought depends. Yet just before the critical spirit
was finally stamped out by Christianity, notable discussions
had taken place in the Græco-Roman world,
and the central problem was being examined with an
acuity which might have led to really scientific developments.
The religious leaders were aware of the danger,
38and there is even a passage in St Gregory of Nazianzus,
where trouble is complained of, since “the Sexti
and Pyrrhoneans and the spirit of contradiction were
perniciously intruded into our churches like some evil
and malignant plague.” 154 In fact the whole theory of
signs was examined both by Aenesidemus, the reviver
of Pyrrhonism in Alexandria, and by a Greek doctor
named Sextus between 100 and 250 a.d. The analysis
offered is more fundamental than anything which made
its appearance until the nineteenth century. 255

This brief survey of the Græco-Roman approach to
language must suffice to represent pre-scientific speculation
upon the subject. Moreover, it has had a greater
influence on modern European thought than the even
more luxuriant growth of oriental theories. The atmosphere
of verbalism in which most Indian philosophy
developed seems to have been even more dense than
that of the scholastics or of the Greek dialecticians.
In this respect the Mīmāmsā-Nyāya controversy, the
Yoga philosophy, the Vijñānavāda categories, the Prābhākara
Mīmāmsakas 356 are hardly less remarkable than
the doctrine of the Sacred Word aum and the verbal
ecstasies of the Sufi mystics, 457 a part of whose technique
was revived by Dr Coué.

The history of spells, verbal magic and verbal
medicine, whether as practised by the Trobriand
magician, 558 by the Egyptian priest of the Pyramid
texts, or by the modern metaphysician, is a subject in
39itself and is dealt with at length in Word Magic, which
is designed as an expansion of the present chapter.

The extent to which primitive attitudes towards
words are still exploited by the astute is fully revealed
only when the achievements of some cynical rhetorician
are accorded the limelight of the law courts, or when
some particularly glaring absurdity is substituted for
the more patient methods of suggestion favoured by
repetitive journalism. But these same attitudes are
universal in childhood, and are so strengthened by the
prevailing verbalism that even the most accurate
scientific training has often done little to render the
adult less subservient to his medium. Indeed, as we
have seen, the ablest logicians are precisely those who
are led to evolve the most fantastic systems by the aid
of their verbal technique. The modern logician may,
in time to come, be regarded as the true mystic, when
the rational basis of the world in which he believes is
scientifically examined.

Turning then to the more emotional aspects of
modern thought, we shall not be surprised to find a
veritable orgy of verbomania. The process whereby
the purely verbal systems so characteristic of pistic
speculation have attained such formidable dimensions,
has recently been examined by Rignano. 159 Attributes
found by experience to be contradictory are gradually
dematerialized, and in their place are put “verbal
envelopes, void of all intelligible content, so as to
eliminate the reciprocal contradiction and inhibition to
which these attributes would inevitably give rise if they
were allowed to furnish matter for the imagination in
however small a degree”; and parallel with this dematerialization,
a formidable dialectic edifice such as
that of scholasticism is constructed, with the object of
convincing human reason of the absence of logical
inconsistency in the greatest of absurdities. 26040

In this way the idea of Divinity, for example, has
been slowly reduced to a “conglomerate of attributes,
purely, or almost purely verbal.” So that finally, as
William James puts it, “the ensemble of the metaphysical
attributes imagined by the theologian” (God
being First Cause, possesses an existence a se; he is
necessary and absolute, absolutely unlimited, infinitely
perfect; he is One and only, Spiritual, metaphysically
simple, immutable, eternal, omnipotent, omniscient, omnipresent,
etc.) “is but a shuffling and matching of
pedantic dictionary adjectives. One feels that in
the theologians' hands they are only a set of titles
obtained by a mechanical manipulation of synonyms;
verbality has stepped into the place of vision,
professionalism into that of life.” 161

Similarly, in reasoning commonly spoken of as
metaphysical, language has chiefly the function of
furnishing “a stable verbal support, so that inexact,
nebulous, and fluctuating concepts may be recalled to
the mind whenever required, without any prejudice to
the elasticity of the concepts”; for which purpose the
phraseology adopted is “as vaporous and mysterious
as possible. Hence the so-called terms ‘written in
profundity,’ referred to by Ribot, and dear to all
metaphysicians, just because they are so admirably
suited both to contain everything that it is desired to
have them include, and to conceal the contradictions
and absurdities of the doctrines based on the concepts
in question… The function of the verbal symbol
is therefore to keep inconsistent attributes forcibly
united, though all of them could not possibly be
present to the mind at the same moment just because
they inhibit each other; it being important that the
metaphysician should have them at his disposal in
order to deduce from the concept, from their aggregate,
sometimes one set of conclusions and sometimes
41another, according to the presentation of reality
desired.”

Ultimately the word completely takes the place of
the thought — Denn eben wo Begriffe fehlen, da stellt
ein Wort zur rechten Zeit sich ein, as Mephistopheles
remarked. And Rignano aptly likens the process to
the shedding of the carapace by a crustacean. “Without
this verbal carapace the disappearance of all intellectual
content would involve the disappearance of all trace
of the past existence of such content. But the carapace
preserves something which, just because it proves the
past existence of a concept which formerly had a real
life, may quite well be taken for one still existing. So
that this something, although devoid of all intellectual
content, always constitutes a valuable point of attachment
and support for the corresponding emotion, which
is so intense that it does not perceive that the cherished
resemblances no longer clothe the beloved object.” 162

But the carapace, the verbal husk, is not merely a
valedictory point d'appui; it also has a certain bombic
capacity, an ‘affective resonance’ which enables the
manipulator of symbols such as the Absolute to assure
himself that his labours are not altogether vain.
“When language is once grown familiar,” says
Berkeley, “the hearing of the sounds or sight of the
characters is often immediately attended with those
passions which at first were wont to be produced by
the intervention of ideas that are now quite omitted.” 263
From the symbolic use of words we thus pass to the
emotive; and with regard to words so used, as in
poetry, Ribot has well remarked that “they no longer
act as signs but as sounds; they are musical notations
at the service of an emotional psychology.” 364 So that
though at this extreme limit “metaphysical reasoning
42may be intellectually quite incomprehensible; though,
that is to say, it may actually become ‘vocem proferre
et nihil concipere,’ it acquires by way of compensation,”
as Rignano says, “an emotive signification which is
peculiar to it, i.e., it is transformed into a kind of
musical language stimulative of sentiments and emotions.”
Its success is due entirely to the harmonious
series of emotional echoes with which the naive mind
responds — et reboat regio cita barbara bombum.

In practical affairs these influences are no less
potent and far more disastrous. We need only instance
the contention of the late Dr. Crookshank, supported by
an abundance of detailed evidence, that “under the
influence of certain schools of thought, and certain
habits of expression, we have become accustomed to
speak and write as if a disease were a natural object”;
that these disastrous verbal habits must be resisted, for
“no great advance is probable in the domain of
Medicine until the belief in the real existence of diseases
is abandoned”; and that the linguistic problem must
be faced at once, for “no measure of useful agreement
will be achieved unless we are first in accord concerning
the principles of method and thought.” 165 Coming from
one with thirty years' experience of the healing art, so
striking a confirmation of the views we have been
advancing cannot be lightly rejected; and on another
page Dr Crookshank himself gives further reasons for
considering that its rejection could only be based on
a failure to appreciate the facts. 266

Until recent times it is only here and there that
efforts have been made to penetrate the mystery by a
direct attack on the essential problem. In the fourteenth
century we have the Nominalist analysis of
William of Occam, in the seventeenth the work of Bacon
and Hobbes. The discussion rises to an apex with the
43Third Book of Locke's Essay and the interest of Leibnitz
in a Philosophical Language — a Characteristica Universalis.
Berkeley and Condillac kept the issue alive, and
with Horne Tooke and his followers we reach the
nineteenth-century movement, in which the work of
Bentham, Taine, and Mauthner was especially
significant. 167

With the disappointing achievements of Comparative
Philology, on which public interest was long
centred through the efforts of Steinthal, Max Müller, and
others, we need not here concern ourselves; the Philological
and Sociological approaches still, in fact, leave
the field-worker without guidance. To the chaos of
the Grammarians we address ourselves in Appendix A;
and in Appendix D, in addition to the summary of the
work of C. S. Peirce, will be found examples of what
has been achieved by others who have looked to Logic
for a solution, as well as by those who appear to have
relied mainly upon Terminology. With contemporary
writers who have made use of the two remaining
avenues (of the seven chief methods of approach) the
Metaphysicians and the Psychologists, we shall be
frequently occupied in our remaining chapters. For
the rest, an endeavour has been made to give credit
where credit is due — from Anselm's De Grammatico,
through Delgarno (1661), Wilkins (1668), Freke (1693),
to Silberer (1917) and Cassirer's Philosophie der symbolischen
Formen (1923) — in the survey of man's progress
towards verbal independence which has been appearing
intermittently in Psyche since 1927. 268

As a result of all these efforts a Science of
Symbolism has become possible, but it is necessary
constantly to bear in mind the special forms in which
the Power of Words may make itself felt in modern
times.44

That which we call a rose, we flatter ourselves, “by
any other name would smell as sweet.” But followers
of the late M. Coué should hesitate to regale themselves
with a rose named The Squashed Skunk. “When I
partake,” says Bergson, “of a dish that is supposed to
be exquisite, the name which it bears suggestive of
the approval given to it comes between my sensation
and consciousness; I may believe that the flavour
pleases me when a slight effort of attention would prove
the contrary.” 371

And words may come between us and our objects in
countless subtle ways, if we do not realize the nature
of their power. In logic, as we have seen, they lead
to the creation of bogus entities, the universals,
properties and so forth, of which we shall have more
to say in the sequel. By concentrating attention on
themselves, words encourage the futile study of forms
which has done so much to discredit Grammar; by the
excitement which they provoke through their emotive
force, discussion is for the most part rendered sterile;
by the various types of Verbomania and Graphomania,
the satisfaction of naming is realized, and the sense of
personal power factitiously enhanced.45

And again, for what do the words we use in
everyday life stand? “We do not often have occasion
to speak, as of an indivisible whole, of the group of
phenomena involved or connected in the transit of a
negro over a rail-fence with a melon under his arm
while the moon is just passing behind a cloud. But if
this collocation of phenomena were of frequent occurrence,
and if we did have occasion to speak of it often,
and if its happening were likely to affect the money
market, we should have some name as ‘wousin,’ to
denote it by. People would in time be disputing
whether the existence of a wousin involved necessarily
a rail-fence, and whether the term could be applied
when a white man was similarly related to a stone wall.” 172

That it is “all a matter of words,” or that “we can
never get anywhere — you put it one way and I put it
another, and how can we ever know that we are talking
46about the same thing?” are conclusions to which the
study of verbal difficulties not infrequently leads those
who are confronted by them for the first time. But a
thorough understanding of the ways in which these
difficulties arise — the two cases just quoted are good
specimens — gives no ground for linguistic nihilism.

The best means of escape from such scepticism as
well as from the hypnotic influences which we have
been considering, lies in a clear realization of the way
in which symbols come to exercise such power, and
of the various senses in which they are said to have
Meaning. As an essential preliminary we are confronted
by the need for an account of the simplest kind
of Sign-situation, which will enable us to understand
how we come to ‘know’ or ‘think’ at all.

The contextual theory of Signs to which, then, we
first proceed, will be found to throw light on the
primitive idea that Words and Things are related by
some magic bond; for it is actually through their
occurrence together with things, their linkage with
them in a ‘context’ that Symbols come to play that
important part in our life which has rendered them
not only a legitimate object of wonder but the source
of all our power over the external world.47

Chapter III
Sign-situations

Studium linguarum in universis, in ipsis primordiis
triste est et ingratutn; sed primis difficultatibus
labore improbo et ardore nobili perruptis, postea
cumulatissime beamur. — Valcknaer.

Meaning, that pivotal term of every theory of language,
cannot be treated without a satisfactory theory of
signs. With some of its senses (in which ‘my meaning’
= ‘what I am thinking of’) the question to be answered
is, in brief, “What happens when we judge, or believe,
or think, of something: of what kind of entities does
the something consist: and how is it related to the
mental event which is our judging, our believing, our
thinking?” The traditional approach to this question
has been through introspection and through the logical
analysis of Judgment, with the result that all the many
answers which have been given from this angle will be
found, in contrast to that which is outlined below, to
be variants of one opinion. They agree, that is, in
holding that, when we think of anything, we have to it
(or sometimes to something else) a relation of a quite
unique kind. In other words thinking is regarded as
an unparalleled happening. Thus the problems of
symbolization and reference come to be discussed in
isolation as though there were no allied fields of
inquiry.

This assumption of the uniqueness of the relation
between the mind and its objects is the central tenet
in views which otherwise have no point of agreement.
Thus it is plausibly held by some that when we are
believing (say) that we are alive, we are in a direct
48relation of a unique kind to an entity which is neither
in time nor in space, to be called the proposition ‘that
we are alive.’ Others pretend that there is nothing of
this sort, but that instead we are then related by a
multiple relation, again of an unique kind, with a
variety of entities — among which are (perhaps) we
ourselves and certainly something to be called a ‘concept’
(or ‘universal’ or ‘property’), namely aliveness
or being alive. On both views the uniqueness in kind
of the relation between a thought as a mental event
and the things, whatever they may be, which the thought
is ‘of,’ is too obvious to be questioned.

As a representative of the realist school which
claims to have assimilated the modern scientific
outlook, we may cite Keynes, who adopted the view
that philosophically we must start from various classes
of things with which we have direct acquaintance.
“The most important classes of things with which
we have direct acquaintance are our own sensations,
which we may be said to experience, the ideas and
meanings, about which we have thoughts and which
we may be said to understand, and facts or characteristics
or relations of sense data or meanings, which
we may be said to perceive… The objects of
knowledge and belief — as opposed to the objects of
direct acquaintance which I term sensations, meanings,
and perceptions — I shall term propositions.” As an
example of direct knowledge we are told that from
acquaintance with a sensation of yellow “I can pass
directly to a knowledge of the proposition ‘I have a
sensation of yellow.’” 173 Lest it should be supposed
that this odd, but very prevalent, doctrine is peculiar
to a school, we may refer to the justification of das
Urteil, “spaceless, timeless and impersonal,” the
specific object of logical inquiry, elaborated by Lipps; 274
49to the similar doctrine which vitiates so much of
Husserl's analysis of language; 175 and to the still
more extraordinary phantasies of van Ginneken, a
subtle linguistic psychologist who, influenced doubtless
by Meinong as well as by Theology, advances
the same view as a theory of ‘adhesion.’ No account
of thinking in terms of verbal images and representations
of things is, according to this author, sufficient.
“We find ourselves confronted by a new force: something
non-sensible, transcendental… by means of
which we understand and know in a new manner, and
a more perfect one than we could through our animal
nature. We… adhere to the present reality, to
that which is really and actually there… and also to
the possible, the essence.” 276 It is plain that on any such
view a scientific account of thinking is ruled out from
the very beginning.

“What happens when we think?” is a question
which should be of interest to every thinker. The
triteness of the answer “When we think, we think,”
offered by such views may help to explain the smallness
of the interest which is shown. In the following
pages an attempt is made to outline an account of
thinking in purely causal terms, without any introduction
of unique relations invented ad hoc. It is with this
end in view, the provision of a natural as opposed to
an artificial theory of thinking, that we begin with the
consideration of signs.

Throughout almost all our life we are treating things
as signs. All experience, using the word in the widest
possible sense, is either enjoyed or interpreted (i.e.,
treated as a sign) or both, and very little of it escapes
some degree of interpretation. An account of the
process of Interpretation is thus the key to the understanding
of the Sign-situation, and therefore the beginning
50of wisdom. It is astonishing that although
the need for such an account has long been a
commonplace in psychology, those concerned with
the criticism and organization of our knowledge have
with few exceptions entirely ignored the consequences
of its neglect.

Attempts to provide this account have been given
in many different vocabularies. The doctrines of the
associationists, 177 of apperception, 278 of suggestion, 379 have
led up to restatements in terms of process rather than
of content: ‘instinctive sequences’ 480 taking the place
of ‘mental chemistry,’ with advantage but without
essential change in the views maintained. The most
recent form in which the account appears is that
adopted by Semon, the novelty of whose vocabulary
seems to have attracted attention once more to considerations
which were no doubt too familiar to be
thought of any importance.

These otherwise valuable methods of approach tend
to separate the treatment of fundamental laws of mental
process from that of sign-interpretation, which is unfortunate
for psychology. They have led not only to
the discussion in isolation of problems essentially the
same, but also to a failure to realize the extent of the
ground already covered by earlier thinkers.

Since the formulation has always been given in
causal terms, it will be convenient to use that terminology.
Its use is indeed almost unavoidable in the
interests of intelligibility, and need not be misleading
if the correct expansion is remembered. Thus in this
preliminary account we are merely using causal language
as an expository convenience for the sake of its brevity
and its verbs. The fuller statement which follows
avoids all mention of causes, effects, and dependence,
51and deals merely with observable correlations or contextual
uniformities among events.

The effects upon the organism due to any sign,
which may be any stimulus from without, or any process
taking place within, depend upon the past history of
the organism, both generally and in a more precise
fashion. In a sense, no doubt, the whole past history
is relevant: but there will be some among the past
events in that history which more directly determine
the nature of the present agitation than others. Thus
when we strike a match, the movements we make and
the sound of the scrape are present stimuli. But the
excitation which results is different from what it would
be had we never struck matches before. Past strikings
have left, in our organization, engrams, 181 residual traces,
which help to determine what the mental process will be.
For instance, this mental process is among other things
an awareness that we are striking a match. Apart from
the effects of similar previous situations we should have
no such awareness. Suppose further that the awareness
is accompanied by an expectation of a flame.
This expectation again will be due to the effects of
situations in which the striking of a match has been
followed by a flame. The expectation is the excitation
of part of an engram complex, which is called up by
a stimulus (the scrape) similar to a part only of the
original stimulus-situation.

A further example will serve to make this clearer.
The most celebrated of all caterpillars, whose history
is in part recorded in the late Professor Lloyd Morgan's
Habit and Instinct, p. 41, was striped yellow and black and
was seized by one of the professor's chickens. Being
offensive in taste to the chicken he was rejected. Thenceforth
the chicken refrained from seizing similar caterpillars.
Why? Because the sight of such a caterpillar,
52a part that is of the whole sight-seize-taste
context of the original experience, now excites the
chicken in a way sufficiently 182 like that in which the
whole context did, for the seizing at least not to occur,
whether the tasting (in images) does or not.

This simple case is typical of all interpretation, the
peculiarity of interpretation being that when a context
has affected us in the past the recurrence of merely a
part of the context will cause us to react in the way
in which we reacted before. 283 A sign is always a
stimulus similar to some part of an original stimulus
and sufficient to call up the engram 384 formed by that
stimulus.

An engram is the residual trace of an adaptation 485
made by the organism to a stimulus. The mental
process 586 due to the calling up of an engram is a
similar adaptation: so far as it is cognitive, what it is
adapted to is its referent, and is what the sign which
excites it stands for or signifies.

The term ‘adapted,’ though convenient, requires
expansion if this account is to be made clear — and to
this expansion the remainder of the present chapter
is devoted. Returning to our instance, we will suppose
that the match ignites and that we have been
expecting a flame. In this case the flame is what we
53are adapted to. More fully, the mental process which
is the expectation is similar to processes which have
been caused by flames in the past, and further it is
‘directed to’ the future. If we can discover what this
‘directed to’ stands for we shall have filled in the chief
part of our account of interpretation.

Besides being ‘directed to’ the future our expectation
is also ‘directed to’ flame. But here ‘directed to’
stands for nothing more than ‘similar to what has been
caused by.’ A thought is directed to flame when it is similar
in certain respects to thoughts which have been caused
by flame. As has been pointed out above, we must not
allow the defects of causal language either to mislead us
here or alternatively to make us abandon the method of
approach so indicated. We shall find, if we improve
this language, both that this kind of substitute for
‘directed to’ loses its strangeness, and also that the
same kind of substitution will meet the case of ‘direction
to the future’ and will in fact explain the ‘direction’
or reference of thinking processes in general.

The unpurified notion of cause is especially misleading
in this connection since it has led even the
hardiest thinkers 187 to shrink from the identification of
54‘thinking of’ with ‘being caused by.’ The suggestion
that to say ‘I am thinking of A’ is the same thing as
to say ‘My thought is being caused by A,’ will shock
every right-minded person; and yet when for ‘caused’
we substitute an expanded account, this strange suggestion
will be found to be the solution.

A Cause indeed, in the sense of a something which
forces another something called an effect to occur, is
so obvious a phantom that it has been rejected even
by metaphysicians. The current scientific account, on
the other hand, which reduces causation to correlation,
is awkward for purposes of exposition, since in the
absence of a ‘conjugating’ vocabulary constant periphrasis
is unavoidable. If we recognize, however, as
the basis of this account the fact that experience has
the character of recurrence, that is, comes to us in more
or less uniform contexts, we have in this all that is
required for the theory of signs and all that the old
theory of causes was entitled to maintain. Some of
these contexts are temporally and spatially closer
than others: the contexts investigated by physics for
instance narrow themselves down until differential
equations are invoked; those which psychology has
hitherto succeeded in detecting are wide, the uniformly
linked events being often far apart in time. Interpretation,
however, is only possible thanks to these recurrent
contexts, a statement which is very generally admitted
55but which if examined will be found to be far more
fundamental than has been supposed. To say, indeed,
that anything is an interpretation is to say that it is a
member of a psychological context of a certain kind.
An interpretation is itself a recurrence.

A concrete illustration may be considered at this
point. There is a well-known dog in most books upon
animal behaviour which, on hearing the dinner-bell,
runs, even from parts of the house quite out of reach of
scents and savours, into the dining-room, so as to be
well placed, should any kind thoughts towards him
arise in the diners. Such a dog interprets the sound of
the gong as a sign. How does this happen? We shall
all agree about the answer; that it is through the dog's
past experience. In this experience there have been so
to speak recurrent clumps of events, and one such clump
has been made up roughly as follows: Gong, savoury
odour, longing contemplation of consumption of viands
by diners, donations, gratification. Such a clump
recurring from time to time we shall call an external
context. Now on a particular occasion the gong is
heard out of reach of savours. But thanks to past
experience of gong-sounds together with savours in
the interpretative dog, this present gong-sound gets into
a peculiar relation to past gongs and savours, longings,
etc., so that he acts in the sagacious manner described,
and is in evidence at the meal. Now this set of mental
events — his present hearing of the gong, his past
hearings of similar sounds, his past savourings together
with gongs, etc., and also his present mental process
owing to which he runs into the dining-room — such
a set we shall call psychological context. A context of
this sort may plainly recur as regards its more general
features. It is also clear that the members of it may
be indefinitely numerous and may be widely separated
in time, and that it is through this separateness in time
that such a psychological context is able to link together
external contexts, the recurrent clumps of experiences
56of the gong-savour kind above mentioned. In a similar
fashion all learning by experience will illustrate the
point that to be an act of interpretation is merely to be
a peculiar 188 member of a psychological context of a
certain kind; a psychological context being a recurrent
set of mental events peculiarly related to one another
so as to recur, as regards their main features, with partial
uniformity.

Little hesitation will be felt in granting that without
such recurrence or partial uniformity no prediction,
no inference, no recognition, no inductive generalization,
no knowledge or probable opinion as to what is not
immediately given, would be possible. What is more
difficult to realize is that this is so only because
these processes, recognitions, inferences or thinkings are
members of certain recurrent psychological contexts.
To say that I recognize something before me as a
strawberry and expect it to be luscious, is to say that
a present process in me belongs to a determinative
psychological context together with certain past processes
(past perceptions and consumptions of strawberries).
These psychological contexts recur whenever
we recognize or infer. Usually they link up with (or
form wider contexts with) external 289 contexts in a peculiar
fashion. 390 When they do not, we are said to have been
mistaken.

The simplest terminology in which this kind of
linkage can be stated is that of signs. Behind all
interpretation we have the fact that when part of an
external context recurs in experience this part is, through
its linkage with a member of some psychological context
(i.e., of a causally connected group of mental events often
widely separated in time) sometimes a sign of the rest
of the external context.

Two points require elucidation if this outline is to
57be filled in. The first concerns Contexts; 191 the second
the sense in which they are Uniform.

(1) A context is a set of entities (things or events) related
in a certain way; these entities have each a character such
that other sets of entities occur having the same characters
and related by the same relation; and these occur ‘nearly
uniformly.’ In our instance of the match-scrape event
and the flame event the uniting relation evidently includes
proximity in time and space — a scrape in America
and a flame in China would not constitute such a
context — but it is important to realize that no restriction
need be initially imposed as to the kind of relation which
may occur as the uniting relation in a context, since
which relations actually occur will be discovered only
by experience. Contexts, moreover, may have any
number of members; dual contexts containing only two
members seem to be rare, though for purposes of
exposition it is convenient to suppose them to occur.
The constitutive characters involved present a certain
difficulty. In our instance of the match-scrape event
and the flame event they may be written ‘being a
scrape’ and ‘being a flame,’ but these are plainly
shorthand names for very elaborate sets of properties.
It is not all scrapes from which we expect flames,
and we would be surprised if our match flamed like
magnesium ribbon.58

In the first case the context is said to be determinative
in respect of the character flame; in the second in
respect of the character scrape; in the third in respect
of both characters.

A dual context is here taken for the sake of simplicity,
a fact which tends to make the account appear artificial.
Multiple contexts of three or more terms involve no
further problems. They must be determinative in
respect of one constitutive character, and may be so in
respect of any number.

In this account we have carefully avoided all mention
59of images — those revivals or copies of sensory experience
which figure so prominently in most accounts of
thinking. There are good reasons why attempts to
build a theory of interpretation upon images must be
hazardous. One of these is the grave doubt whether
in some minds they ever occur or ever have occurred.
Another is that in very many interpretations where
words play no recognizable part, introspection, unless
excessively subtle and therefore of doubtful value as
evidence, fails to show that imagery is present. A third
and stronger reason is that images seem to a great
extent to be mental luxuries. Before the appearance
of an image, say, of an afanc, something can be observed
to occur which is often misleadingly described as ‘an
intention of imagining’ an afanc. But that this is not
merely an intention becomes plain upon reflection.
When we speak of an intention in this way we are
speaking of affective-volitional characters, those, roughly
speaking, on account of which a state of mind changes
from a relatively inchoate to a relatively organized and
articulate condition. An intention by itself is as impossible
as an excitement. There has to be something
which is excited, and there has to be something for
the intention to belong to. Now what is this in such
cases as we are examining?

Whatever it is it has that peculiar character of being
directed towards one thing rather than another, which
we here call reference. This reference may be uncertain
and vague, but seems to be the same in kind as that
which occurs in more articulate and clear-cut cases of
thinking, where symbols in the form of images or words
have been provided. In the initial stages of such
references it is hard to suppose that images are playing
any essential part. Any image which does arise is at
once accepted or rejected as it accords or disaccords
with the reference, and this accordance is not a question
of matching between images or of similarity in any
intrinsic characters. If images of any sort are involved
60in these states of beginning to think of things, it is
certain that they are not always involved qua images,
i.e., as copying or representing the things to which
the reference points, but in a looser capacity as mere
signs and not in their capacity as mimetic or simulative
signs.

Indeed, it may be questioned whether mimetic
imagery is not really a late, sporadic product in mental
development. We are so accustomed to beginning
psychology with images that we tend to think that
minds must have begun with them too. But there is
no good reason to suppose that the mind could not
work equally well without them. They have certain
oddly limited uses as economizing effort in certain
restricted fields. The artist, the chess-player, the mathematician
find them convenient. But these are hardly
primitive mental occupations. Hunger rarely excites
taste images, the salivary flow occurs without them.
Route-finding in pathless wilds or Metropolitan suburbs
is best done by sense of direction and perception alone.
On the whole, a mimetic sign is not the kind of thing
that a primitive mind would be able to make much use
of. Other signs would serve equally well for most
purposes, and the few advantages of images would be
more than counterbalanced by ‘the risk of danger’ to
which their users expose themselves. An inaccurate
or irrelevant image is worse than no image at all.
Such arguments as there are in favour of images as
very primitive and fundamental products, the argument
from dreams, for example, or the alleged prevalence of
images among children and primitive peoples, are
obviously difficult to estimate. Imagery may be
prevalent without necessarily serving any important
function; in day-dreaming, for instance, the gratifications
which it affords are no proof that the references concerned
could not occur without it. Similarly those who
naturally produce exhaustive images of their breakfast-table
can often know all about it without a glimmer of
61an image, unless too much indulgence in images has
impaired their natural ability.

For these reasons, any theory of interpretation which
can refrain from making images a corner-stone has clear
advantages over those which cannot. It is mainly on
this point that the view here developed differs from Mr
Russell's account 193 of meaning, which should, however,
be consulted by those who desire a more simple discussion
of the part played by Mnemonic causation in
knowledge than our brief outline provides.

Suppose now that we have struck our match and
have expected a flame. We need some means of
deciding whether our expectation has been true or false.
Actually we look to see whether there was a flame or
not, but the question we have to answer is, how do we
pick out, amongst all the other possible events which
we might have selected, this particular flame as the
event on which the truth or falsity of our expectation
depended. 294 We pick it out by means of certain external
contexts to which it belongs: namely, it is that event,
if any, which completes the context whose other member
in this case is the scrape, and thus comes to be linked
to the expectation through the psychological context
made up of that expectation and past experiences of
scrapes and flames.

If now there be an event which completes the external
context in question, the reference is true and the event
is its referent. If there be no such event, the reference
is false, and the expectation is disappointed.

The above account covers beliefs of the form ‘a
flame will follow this scrape’ prompted by a present
62sensation. Instead of a present sensation a belief may
itself be a sign for a further belief which will then be
an interpretation of this belief. The only cases of this
which appear to occur are introspective beliefs of the
form ‘I believe that I am believing, etc.’ which may,
it is important to recognize, be false as often as, or
more often than, other beliefs. As a rule a belief not
prompted by a sensation requires a number of beliefs
simultaneous or successive for its signs. The beliefs,
‘There will be a flame’ and ‘I am in a powder factory,’
will, for most believers, be signs together interpreted
by the belief ‘The end is at hand.’ Such is one of
the psychological contexts determinative in respect of
the character of this last belief. 195 Whether the belief
in question is true or not will depend upon whether
there is or is not some entity forming together with
the referents of the two sign beliefs, in virtue of its
characters and their characters and a multiple relation,
a context determinative in respect of their characters.
In other words — upon whether the place does blow up.

In this way the account given can be extended to
all cases of particular expectations. Further, since the
uniting relations of contexts are not restricted to successions
it will also apply to all cases of inference or
interpretation from particular to particular. The next
step, therefore, is to inquire what kind of account can
be given of general references.

The abstract language which it is necessary to
employ raises certain difficulties. In a later chapter
arguments will be brought in favour of regarding such
apparent symbols as ‘character,’ ‘relation,’ ‘property,’
63‘concept,’ etc., as standing for nothing beyond (indirectly)
the individuals to which the alleged character
would be applicable. The most important of these
arguments is the natural incredibility of there being
such universal denizens of a world of being. As we
shall see, these apparent symbols are indispensable as
machinery, and thus for some purposes such credulity
is harmless. But for other purposes these baseless (or
purely symbolically based) beliefs are dangerous impediments.
Thus a chief source of opposition to an
extension of the account here outlined to general
references, is phantom difficulties deriving from faith
in this other world.

Such references may be formulated in a variety
of ways: — ‘All S is P’ and ‘(x): φ (x) ψ (x)’ are
favourites. What we have to discover is what happens
when we have a belief which can be symbolized in
these ways. Let us take as an instance the belief
‘All match-scrapes are followed by flames.’ There is
good reason to suppose that such beliefs are a later
psychological development than beliefs of the form
which we have been considering. It is plausible to
suppose that some animals and infants have particular
expectations but not any general beliefs. General beliefs,
it is said, arise by reflection upon particular beliefs.
Thus we may expect to find that general beliefs arise
in some way out of particular beliefs. But the generality
and particularity to be attributed to simple or
primordial references are certainly not those which
logical formulation endeavours to introduce. Nor
should it be supposed that genetically a stage or era
of particular reference precedes general thinking. It
is rather the case that in all thought processes two
tendencies are present, one towards greater definiteness
or precision, the other towards wider scope and range.
It is the conditions under which this second tendency
takes effect that we are here considering.

The second of these conditions is plainly more
important than the first. We often seem to pass to
general beliefs from single experiences and not to
require a plurality, but (exceptionally powerful thinkers
apart) we do not base general beliefs upon directly
contradictory evidence. We may therefore retain the
second condition, but must revise the first. In some
cases, no doubt, repeated verified expectations do
condition the general expectation, but they condition
its degree rather than its reference. On the other hand
some experience of repetition would seem to be required.
A primordial mind's first thought could hardly be a
general thought in the sense here considered. It seems
justifiable to assume that some series of similar verified
interpretations should be included in the context of a
general belief, though how closely this need be connected
with the particular interpretation which is being
generalized must at present be left uncertain.

Another condition which can only be put rather
vaguely concerns the inclusiveness of a general
reference. The togetherness involved in such a reference
does not seem to require any properties in a
‘mind’ beyond those already assumed and stated, but
the inclusiveness might be thought to raise an additional
problem. The kind of experience required,
however, is not difficult to discover. On many occasions
so far as the verifying stimuli are concerned it is
65indifferent whether we think of all of a given set of
objects or of each of them in turn. The child who finds
all his fingers sticky might equally well have found
each of them sticky. On other occasions his smallest
fingers will not need to be washed. Thus the difference
between inclusive and non-inclusive sets of objects
as referents, the difference between ‘some’ and ‘all’
references, will early develop appropriate signs.
Individuals can be found who throughout their lives
‘think’ of these differences by means of such images,
i.e., use such images as adjunct-signs in their interpretations.
In other cases no such imagery nor even
the use of the words ‘all’ or ‘some,’ or any equivalents,
is discoverable. Yet even in these cases some lingering
trace of the engraphic action due to situations of
this sort may reasonably be supposed as conditioning
interpretations which ‘employ these notions.’ In
attempting therefore to set out the kind of psychological
context of which a general reference consists, terms
representing them would require inclusion.

Such in very tentative outline is the account which
the causal theory of reference would give of general
beliefs. The detailed investigation of such contexts is
a task to which sooner or later psychology must address
itself, but the methods required are of a kind for which
the science has only recently begun to seek. Much
may be expected when the theory of the conditioned
reflex, due to Pavlov, has been further developed. 196

It remains to discuss in what sense, if any, a false
belief, particular or general, has a referent. From the
definitions given it will be plain that the sense in which
a false belief may be said to have a referent must be
quite other than that in which a true belief has a
referent. Thus the arguments now to be given for
a more extended use of the term in no way affect what
has been said; and it will also be purely as a matter
66of convenience that we shall use the term in connection
with false beliefs.

In the first place it is clear that true and false
references alike agree in a respect in which processes
such as sensing, breathing, contracting muscles,
secreting, desiring, etc., do not agree with them. It
is convenient to have a term, such as reference, to stand
for this respect in which they agree. The term ‘belief’
which might at first appear most suitable is less convenient,
both because of its association with doctrines
such as those above discussed which postulate an
unique relation ‘thinking of,’ and because it is becoming
more and more often used with special reference
to the affective-volitional characters of the process. A
second and stronger reason derives from what may be
called the analysis of references. If we compare, say,
the references symbolized by ‘There will be a flash
soon,’ and ‘There will be a noise soon,’ it is at least
plausible to suppose that they are compounds containing
some similar and some dissimilar parts. The
parts symbolized by ‘flash’ and ‘noise’ we may
suppose to be dissimilar, and the remaining parts to
be similar in the two cases. The question then arises:
“What are these parts from which it would seem
references can be compounded?”

The answer which we shall give will be that they
are themselves references, that every compound
reference is composed wholly of simple references
united in such a way as will give the required structure
to the compound reference they compose. But in
attempting to carry out this analysis a special difficulty
has to be guarded against. We must not suppose that
the structure of the symbol by which we symbolize the
reference to be analysed does in any regular fashion
reflect its structure. Thus in speaking of the parts
symbolized by ‘flash’ and ‘noise’ above we are
running a risk. Illegitimate analyses of symbols are
the source of nearly all the difficulties in these subjects.67

Another point which must first be made clear concerns
the sense in whch references may be compounded.
To speak of a reference is to speak of the contexts
psychological and external by which a sign is linked to
its referent. Thus a discussion of the compounding of
references is a discussion of the relations of contexts
to one another.

What are usually called the ‘logical forms’ of
propositions, and what we may call the forms of
references, are, for the view here maintained, forms or
structures of the determinative contexts of interpretations.
They are at present approached by logicians
mainly through the study of symbolic procedure. A
more direct approach appears however to be possible,
though, as yet, difficult. Thus the remaining portions
of the complete contextual theory of reference, namely
the accounts of references of the forms ‘p or q,’ ‘p and
q,’ ‘not p,’ and of the difference between ‘S’ and
‘some S,’ regarded as concerned with the interweaving
of contexts, are, if still conjectural, plainly not beyond
conjecture.

With this proviso, we may resume the consideration
of the referents of false and of the analysis of compound
beliefs.

We have seen that true and false beliefs are members
of the same kinds of psychological contexts, and that
they differ only in respect of external contexts. 197 Let
68us consider this difference again, taking for the sake
of simplicity the case of particular beliefs. Suppose
that of two possible beliefs, ‘There will be something
green here in a moment,’ ‘There will be something red
here in a moment,’ the first is true and the second false.
But the second, if it can be regarded, as having
contained or included the belief, ‘There will be something
here in a moment,’ will have included a belief
which is true and similar to a belief included in the
first belief. Reverting now to our definition of a
context let us see in what sense this belief is included
and how it can be true.

In such a case the external context may consist of
two entities, say s (a sign) and g (something green),
having the characters S, G, and related by space and
time relations which may be taken together. But it is
clear that both s and g will have other characters
besides S and G. For instance, s has succeeded other
entities and may be interpreted in respect of this
character as well as in respect of S, so 198 interpreted it
gives rise to the belief, ‘There will be something here
in a moment’; interpreted also in the further respect
of S it gives rise to the complex belief, ‘There will be
something green here in a moment,’ or to the complex
belief, ‘There will be something red here in a moment,’
true and false interpretation of sin this further respect
as the case may be. In either case, however, the
contained belief, ‘There will be something here in a
moment,’ will be true if there is something (say g)
which forms with s, in virtue of s's character of being a
successor (or other temporal characters) and g's temporal
characters, a context determinative of this
character of s. Thanks to the generality of these
characters such contexts never fail to recur, a fact which
accounts for the ease with which true predictions of
this unspecific kind can be made.69

It appears then that a belief may contain other less
specific beliefs, and that a compound definite belief is
composed of simpler, less specific beliefs, united by
such relations as will yield the required structure. 199

One objection to such a view derives from language.
It is usual to restrict the term belief to such processes
as are naturally symbolized by propositions and further
to those among such processes as have certain affective-volitional
characters in addition to their characters as
cognitions. The simple references which would be
required if the analysis suggested were adopted would
rarely lend themselves to propositional formulation and
would be lacking as a rule in accompanying belief,
feelings and promptings to action. Thus the terms
‘idea’ and ‘conception’ would often be more suitable
for such processes. To extend a metaphor which is
becoming familiar, these might be regarded as
‘electronic’ references. But the ideas or conception
with which we are here concerned would have to be
clearly distinguished from the ‘concepts’ of those
metaphysicians who believe in a world of universals.
We shall deal at greater length with the question in
Chapter V.

Let us consider the idea or conception of green.
It arises in the reader in this case through the occurrence
of the word ‘green.’ On many occasions this
word has been accompanied by presentations of green
things. Thus the occurrence of the word causes in him
a certain process which we may call the idea of green.
But this process is not the idea of any one green thing;
such an idea would be more complex and would require
a sign (or symbol in this case) with further characters
for him to interpret — only so will his idea be specific.
70The psychological context to which it belongs is not of
a form to link any one green thing with the sign rather
than any other. If now we write instead, ‘a green
thing,’ the same process occurs — unless the reader is a
logician or philosopher with special theories (i.e., peculiar
linguistic contexts). In both cases the idea can
be said to be ‘of’ any sensation similar to certain
sensations which have accompanied in the past the
occurrence of the sensation taken as a sign. Compare
now the indefinite belief symbolized by ‘There are
green things.’ Here any one of the same set of
sensations that the idea was said to be ‘of’ will verify
the belief. For if there be one or more entities similar
to certain entities which are members of its psychological
context, it will be true; otherwise it will be
false. We may therefore extend the term ‘referent’ to
cover these entities, if there be any such, without the
usage leading to confusion.

It will be noticed that strictly simple indefinite
beliefs (illustrated by, ‘There are green things’ as
opposed to ‘There are green things now’) only require
for their truth a condition which is present among
their psychological contexts. This happy state of
things has its parallel in the fact that strictly simple
ideas raise no problem as to whether they are ideas
of anything or not. But complex ideas, such as
glass mountains, phœnixes, round squares, and
virtuous triangles may be made to bristle with such
problems. The distinction between an idea and a belief
is, however, one of degree, although through symbolic
conventions it can sometimes appear insuperable.

We can now define the usage of the term ‘referent’
for false beliefs. All beliefs whether true or false are
theoretically analysable into compounds whose constituents
are simple references, either definite or indefinite,
united by the relations which give its ‘logical
form’ to the reference.

Definite simple references are not very common.
71Sometimes when we say ‘this!’ ‘there!’ ‘now!’ we
seem to have them. But usually, even when our reference
is such that it can have but one referent, it can
be analysed. Even references for which we use simple
symbols (names), e.g., Dostoevski, are perhaps always
compound, distinct contexts being involved severally
determinative of distinct characters of the referent. 1100
What is more important is to understand the peculiar
dispersion which occurs in false reference. Illustrations
perhaps make this clearer than do arguments.

Thus, if we say, ‘This is a book’ and are in error,
our reference will be composed of a simple indefinite
reference to any book, another to anything now, another
to anything which may be here, and so on. These
constituents will all be true, but the whole reference to
this book which they together make up (by cancelling
out, as it were, all but the one referent which can be
a book and here and now) will be false, if we are in
error and what is there is actually a box or something
which fails to complete the three contexts, book, here,
and now. To take a slightly more intricate case, a
golfer may exclaim, “Nicely over!” and it may be
obvious to the onlooker that his reference is to a divot
and its flight, to his stroke, to a bunker, and to a ball.
Yet the ball remains stationary, and these constituent
or component references, each adequate in itself, are
combined in his complex reference otherwise than are
their separate referents in actual fact. There is clearly
no case for a non-occurrent flight of a golf-ball as an
object of his belief; though he may have been referring
to the feel of his stroke, or to an image of a travelling
ball. In these last cases we should have to suppose
him to be shortening his own interpretative chain
instead of breaking loose and venturing a step too far
72by what may be called saltatory interpretation. His
language (cf. also Canon IV., page 103 infra) does not
bind us to either alternative. Thus we see in outline
how compound false beliefs may be analysed.

The referent of a compound false belief will be the
set of the scattered referents of the true simple beliefs
which it contains. We shall, in what follows, speak of
beliefs, and interpretations, whether true or false, and
of ideas, as references, implying that in the senses
above defined they have referents.

We thus see how the contextual theory of reference
can be extended to cover all beliefs, ideas, conceptions
and ‘thinkings of.’ The details of its application to
special cases remain to be worked out. Logicians will
no doubt be able to propound many puzzles, 1101 the
solving of which will provide healthy exercise for
psychologists. The general hypothesis that thinking
or reference is reducible to causal relations ought however
to commend itself more and more to those who
take up (at least sometimes) a scientific attitude to the
world. Subject to the proviso that some satisfactory
account of probability can be given, ‘meaning’ in the
sense of reference becomes according to this theory a
matter open to experimental methods.

A satisfactory account of probability, however,
though very desirable, does not seem likely to be
forthcoming by current methods. Evidently a change
of attack is required. The late Lord Keynes' Treatise
starting as it does with an unanalysable logical relation,
called probability, which holds between equally mysterious
and unapproachable entities, called propositions,
is too mediæval in its outlook to be fruitful; and it remains
to be seen whether scientists will be able to profit by
Reichenbach's more empirical Wahrscheinlichkeitslehre.

It seems possible on the contextual theory of reference
73to suggest an expansion of this kind of obscure
shorthand and so come nearer the formulation of the
yet undiscovered central question of probability. What
are talked about by logicians as propositions are,
according to this theory, relational characters of acts of
referring — those relational characters for which the term
‘references’ is used. Thus to believe, or entertain, or
think of, a proposition, is on this view simply to refer,
and the proposition as a separate entity is to be regarded
as nothing but a linguistic fiction foisted upon us by
the utraquistic subterfuge. 1102 Two ‘thinkings of’ the
same ‘proposition’ are two thinkings with the same
reference, the same relational property, namely ‘being
contextually linked in the same way with the same
referent.’ It will be noted that on this account of
propositions the logical relations of propositions to one
another must be dealt with far less summarily and
formally than has hitherto been the case.

With propositions so understood there occurs a sense
in which a single proposition by itself without relation
to other propositions, can intelligibly be said to be
probable. Probability here has still a relational aspect,
and it is only because propositions (i.e., references) are
relational that they can be said to be probable. This
very fundamental sense is that in which the uniformity
of the context upon which the truth of a reference depends
is probable.

We have seen that by taking very general constitutive
characters and uniting relation, we obtain contexts
of the highest probability. Similarly by taking too
specific characters and relation the probability of the
context dwindles until we should no longer call it a
context. In this way, whether a context is probable
can be seen to be a question about the degree of
generality of its constitutive characters and uniting
relation; about the number of its members, the other
contexts to which they belong and so on… a question
74not about one feature of the context but about many.
We can always for instance raise the probability of
a context by adding suitable members. But this last
though a natural remark suffers from the linguistic
redundance to which the difficulties of the problem are
chiefly due. ‘Probability’ in the fundamental sense
in which a context is probable is a shorthand symbol
for all those of its features upon which the degree of
its uniformity depends.

In considering conscious and critical processes of
interpretation we must not fail to realize that all such
activity, e.g., of the kind discussed in the theory of
induction, rests upon ‘instinctive’ interpretations. If
we recognize how essential ‘instinctive’ interpretation
is throughout, we shall be able to pursue our investigations
undisturbed by the doubts, of causal purists or
the delay of the mathematicians in bringing their
differential equations into action. For the working
of a differential equation itself, that most rational
process of interpretation, will break down unless many
‘instinctive’ interpretations, which are not at present
capable of any mathematical treatment, are successfully
performed.

It is sometimes very easy by experimental methods
to discover what a thought process is referring to. If
for instance we ask a subject to ‘think of’ magenta
we shall, by showing various colours to him, as often
as not find that he is thinking of some other colour. It
is this kind of consideraton which makes the phrase
‘adapted to’ so convenient an equivalent for ‘referring
to,’ and if we bear in mind that ‘being adapted to’
something is only a shorthand symbol for being linked
with it in the manner described, through external and
psychological contexts, we may be able to use the
term without its purposive and biological associations
leading to misunderstanding.

We have still to give an account of misinterpretation,
and to explain how unfounded beliefs can arise. To
75begin with the first, a person is often said to have
introduced irrelevant, or to have omitted relevant, considerations
or notions when he has misinterpreted some
sign. The notion of relevance is of great importance
in the theory of meaning. A consideration (notion,
idea) or an experience, we shall say, is relevant to an
interpretation when it forms part of the psychological
context which links other contexts together in the
peculiar fashion in which interpretation so links them. 1103
An irrelevant consideration is a non-linking member of
a psychological context. The fact that ‘baseless’
convictions occur might be thought to be an objection
to the view of thinking here maintained. The explanation
is however to be found in the fact that mental
processes are not determined purely psychologically
but, for example, by blood pressure also. If our interpretation
depended only upon purely psychological
contexts it might be that we should always be justified
in our beliefs, true or false. We misinterpret typically
when we are asleep or tired. Misinterpretaton therefore
is due to interference with psychological contexts,
to ‘mistakes.’ Whether an interpretation is true or
false on the other hand does not depend only upon
psychological contexts — unless we are discussing psychology.
We may have had every reason to expect
a flame when we struck our match, but this, alas! will
not have made the flame certain to occur. That depends
upon a physical not a psychological context.76

Chapter IV
Signs in perception

La Nature est un temple ou de vivants piliers
Laissent parfois sortir de confuses paroles;
L'homme y passe a travers des forêts de symboles
Qui l'observent avec des regards familiers. —
Baudelaire.

Though with the growth of knowledge we have become
much less certain than our ancestors about what chairs
and tables are, physicists and philosophers have not
yet succeeded in putting the question entirely beyond
discussion. Every one agrees that chairs and tables
are perfectly good things — they are there and can be
touched — but all competent to form an opinion are
equally agreed that whatever we see is certainly not
them. What shall we do about it?

The assumption of the reverend gentleman is that,
having asked a definite question, he was entitled to a
definite answer. Very little study of what he actually
saw or tapped might have saved him the trouble of
discovering at a later stage that “one lad had thought
we wanted the word for tapping; another understood
we were seeking the word for the material of which the
table was made; another had an idea that we required
the word for hardness; another thought we wished for
a name for that which covered the table; and the last,
not being able, perhaps, to think of anything else
gave us the word meza, table — the very word we were
seeking.”

A similar discovery awaits the experts, and it may
not be inapposite to indicate the main features of this
imminent advance in knowledge. It is at first sight
surprising that modern investigators should have been
so long in taking up the analysis of sign-situations as
begun by Aenesidemus and Occam. But their uneasiness
in matters which they supposed to fall within
the domain of ‘the metaphysicians,’ seems to have
been sufficient to inhibit their curiosity as to the principles
of interpretation involved at every stage of their
work. Moreover, so long as controversy with specialists
in other fields was avoided, a great deal could be
achieved without the realization that perception can
only be treated scientifically when its character as a
sign-situation is analysed.

The isolated utterance of Helmholtz is therefore all
the more significant, for not only was Helmholtz one
of the profoundest scientific thinkers of modern times,
78but, as we know from his correspondence, he took
throughout his life a lively interest in philosophic controversies.
In 1856 we even find him referring to
the problem of the way in which we pass from simple
sensations to judgments of perception as one to which
no modern philosopher had devoted serious attention.
He was much influenced by Kant, who, in spite of
his disconcerting technique, seems constantly on the
verge of approaching the central issues of interpretation,
and who has been claimed as the most convinced
Nominalist of modern times: 1105 but there is nothing
particularly Kantian about the theory of signs which
can be found in various parts of Helmholtz' writings. 2106
Our knowledge, he contended, takes the form of signs,
and those signs we interpret as signifying the unknown
relation of things in the external world. The sensations
which lie at the basis of all perceptions are subjective
signs of external objects. 3107 The qualities of sensations
are not the qualities of objects. Signs are not pictures
of reality.

“A sign need have no kind of similarity whatever
with what it signifies. The relation consists simply
in the fact that the same object acting under similar
circumstances arouses the same sign, so that different
signs correspond always to different sensations.” 4108

In discussing the way in which we interpret sensations
in terms of an external world, Helmholtz has
occasion to point out that the multiplicity of the optical
signs which we use is such that we need not be surprised
at the variety and complexity of the news which
they give us. The elementary signs of language are
only 26 letters. If out of these 26 letters we can
get the whole of literature and science, the 250,000
optic nerve fibres can be relied on for an even richer
and more finely graded knowledge.79

What do we see when we look at a table? First
and foremost, a lighted region containing some air,
lit by rays coming partly from the direction of the table,
partly from other sources; then the further boundaries
of this region, surfaces of objects, including part of the
surface of the table. If now we point at what we see
and name it This, we are in danger, if our attention is
concentrated on the table, of saying: This is a Table.
So that we must be careful. And where is colour
according to this scheme? Somewhere in the eye, as
anyone who cares to strike his eye will discover.

What we have described is not the Table, though
part of what we have described is part of the table.
Anything which we say under these circumstances which
involves the Table must also involve Interpretation. We
interpret a sign, some part of what is given, 1109 as signifying
something other than itself, in this case the table.

But this is not the whole of the story, and here it
seems possible to say something quite new. It would
be strange to suggest that we see anything which is
not in front of the eye, or which does not, like a musca
volitans, throw images on the retina. Thus purists will
have to maintain that we never see colours. Yet it is
colours and such directly apprehended entities that are
the initial signs on which all interpretation, all inference,
all knowledge is based. And what is it that by interpretation
we come to know? It is what is present —
a whole which, as we learn in course of time, is composed
80of the lighted region, the air, etc., to which
we allude above, but in which we only distinguish
these namable components after a long process of
interpretation conducted on experimental methods —
“The infant learns first, etc., etc.”

What then is this direct apprehending to which so
important a role is assigned? The correct answer is
usually rejected without hesitation, so contrary is it
to some of our favourite verbal habits. To be directly
apprehended is to cause certain happenings in the
nerves, as to which at present neurologists go no
further than to assert that they occur. Thus what is
directly apprehended is a modification of a sense organ,
and its apprehension is a further modification of the
nervous system, about which we may expect information
at some future date. 1110

But this is mere materialism? Suitably misunderstood,
it is. In itself, however, it is no more than a highly
probable step in the most plausible systematic account
of ‘knowing’ which can be given. On all other
accounts yet suggested, at least one indefinable idea
has at some point to be introduced, at least one
ultimately and irredeemably mysterious extra entity
has to be postulated — some relation of ‘immediate
knowing’ and further inexplicables in its train. Meanwhile
it is generally granted that much is known. There
are the sciences; and it is here urged that we already
have the material for an account of knowing itself —
provided, that is, certain symbolic entanglements are
first penetrated or swept aside.81

The chief of these rest upon misunderstanding as
to the nature of statement. To make a statement is to
symbolize a reference. What a reference is we have
seen in the preceding chapter. However much we
may try, we cannot go beyond reference in the way
of knowledge. True reference is reference to a set of
referents as they hang together. False reference is
reference to them as being in some other arrangement
than that in which they actually hang together. The
advance in knowledge is the increase in our power of
referring to referents as they actually hang together.
This is all we can do. By no manner of make-believe
can we discover the what of referents. We can only
discover the how. This is, of course, old and familiar
doctrine but it needs to be reaffirmed whenever the
metaphysician intervenes, whether he comes as a
materialist, spiritualist, dualist, realist or with any other
answer to an impossible question. Unfortunately in
our present ignorance of the mechanism of language,
he has a good opportunity of setting up apparently
impenetrable barriers. The only way by which these
may be avoided is to set out from the known facts as
to how we acquire knowledge. Then with an account
of interpretation, such as that which is here sketched,
the way is open to the systematization of all that is
known and further of all that will ever come to be
known. 1111

To resume our outline sketch of a systematic account
of perception. Directly apprehended retinal modifications
such as colours, are therefore initial signs of
‘objects’ and ‘events’ (or however we agree to symbolize
82referents); characters of things which we discover by
interpretation, such as shapes of cones or tables, are
signs of second or third order respectively. On the
other hand shapes of initial signs, e.g. retinal modifications,
are first order signs.

Place a new nickel florin on the palm of the hand
with the arm extended horizontally, and note that a
truthful person would describe its shape as elliptical.
Now look at it vertically from above and agree that it
is round. Is the florin circular or elliptical? What an
insoluble problem!

If we say that it is the surface of the florin which is
given us in both cases, then it seems to be both circular
and elliptical. Which is absurd — since we ‘know,’
and every physicist stoutly maintains, 1112 that it has not
measurably changed, and is actually circular. We
have, therefore, the option on the one hand of
opining with the Metaphysicians that the Universe
is very paradoxical, with the polite Essayists that it is
very odd, or with the Bishops that it is very wonderful;
or, on the other, of saying that it is not the surface
which is given in either sense.

Anyone who watched our procedure with the florin,
if appealed to for assistance at this point, would say
that what was present in each case was a whole containing
as parts, cones 2113 whose apices are in the eye,
and whose bases are the limits of our vision, or, where
objects such as florins are about, their surfaces. Here
there are two cones with the circular surface of the
florin for base. In the first case the cone is elliptical
83in cross-section, and the surface of the florin is an
oblique section; in the second case the cone is circular,
and the surface of the florin a cross-section, also circular.
What here is taken as the apparent shape of the florin
is most plausibly said to be the cross-section of the
cone. This is the sign which we interpret as the
surface, and in no case is that surface a ‘datum datissimum’ — directly
given. This simple application of
the Theory of Signs frees us from the paradox, the
oddness, and the wonder, restores our faith in the
physicist, and enables us to get on with our business,
viz., a proper account of perception of the Nature
of Things.

The method by which this ancient scandal is removed
may be applied with equal success to all the
other ‘fundamental problems.’ Whenever the ingenious
mind discovers a self-contradiction (“This
same florin that I see is both round and elliptical,” or
“This same stick which I see in the water is both
straight and bent”) bad symbolization is indicated, and
we must expand the peccant symbol 1114 until we discover
the ambiguous sign-situation which caused the trouble.
We then note this ambiguity, and improve our symbolism
so as to avoid the nonsense to which we shall
otherwise be led. Thus in the case of the florin we
say: “The base of this cone that is my sign is oblique
and circular, and is the surface of the florin that I see;
but a normal section of this cone is elliptical. I can
equally be said to see the florin or to see any section
of the cone, but no one of these is directly given. Even
the whole cone of which they are parts is picked out
from the wider cone which includes besides the florin
cone the cones of all that I am seeing, the total datum
which is my field of view.”

This selection of partial cones out of the total cone
84which is the visual field is, in normal circumstances,
effected without mistake. It might, in fact, never have
been suspected that even here interpretation is at work,
were it not for the case of ‘double images.’ For each
eye there is a separate total cone, but we learn normally
to identify certain partial cones within these as having
the same base. If the retinal correspondence through
which we do this is upset (as when we push the eyeball
a little, or look past a near at a distant object) we fail
to make the right identification, and say we see two
florins (double-images). Here once again we let our
language trick us. What is present is, as always in
binocular vision, two cones with a common base.
Thanks to the retinal shift, the normal, automatic
method of identification breaks down, and we ‘see’
one florin as though it were in two places; we interpret
two cones with a common base as though they were
cones with separate bases. Reflection and refraction —
the whole of the theory of vision is full of such
‘puzzles,’ to be solved by the above Theory of Signs. 1115

Through this Theory of Signs then we can not only
remove the standard pre-scientific paradoxes, but provide
a new basis for Physics. It is commonly assumed
that contrasted with what we see are the things we
imagine, which are in some sense unreal. This distinction
between Vision and Imagination is misleading,
and of those things which we rightly claim to see the
parts we do not see are as real as those we do. The
85other side of the moon, which we never see, is as real
as the side which vision perceives. The atoms, whose
paths are photographed, the electrons which we do not
‘see’, are, if this interpretative effort of the physicist
be sustained, as real as the signs given to perception
from which he starts. When we look at our chairs
and tables we ‘see’ a datum datissimum, then cones,
then surfaces, chair, legs-seat-back, wood, bamboo,
fibres, cells, molecules, atoms, electrons… the many
senses of ‘see’ proceeding in an ordered hierarchy as
the sign-situations change. And as the point of view,
interest, scientific technique or purpose of investigation
alters, so will the levels represented by these references
change in their turn.86

Chapter V
The canons of symbolism

A happy nomenclature has sometimes been more
powerful than rigorous logic in allowing a new train
of thought to be quickly and generally accepted. —
Prof. A. Schuster.

For the rest I should not be displeased, sir, did you
enter a little farther into the details of the turns of
mind which appear marvellous in the use of the
particles. — Leibnitz.

At the basis of all communication are certain postulates
or pre-requisites — regulative presumptions without
which no system of symbols, no science, not even
logic, could develop. Their neglect by logicians is
not surprising, since it has hitherto been nobody's
business to discuss them. Logic, which may be
regarded as the science of the systematization of
symbols, has been preoccupied either with judgments
which are psychological, or with ‘propositions,’ which
were treated as objects of thought, distinct from
symbols and not psychological. Modern mathematicians,
who have done so much for the formal development
of symbolic method, either tacitly assume these
Canons, or when confronted by difficulties due to their
neglect, introduce additional ad hoc complexities 1116 into
their systems. Actually they are as essential to all
discourse as chemistry to physiology, dynamics to
ballistics, or psychology to æsthetics. In any logic
which is not purely formal, in the sense of being
87devoted to some elaboration of the possibilities of
symbol-manipulation, 1117 the study of these Canons is a
first essential, and their strict observance would render
otiose whole tracts of the traditional treatment.

It will be convenient to state some of these Canons
in terms of Symbols and Referents. The triangle of
Reference given on p. 11 should be consulted. The
First Canon of Symbolism, the Canon of Singularity,
is as follows: —

I. — One symbol stands for one and only one referent.

This one referent may be, and in most cases is,
complex. ‘All Mongolian Imbeciles,’ for instance, is
a symbol which has one referent. Similarly (x or y)
has one referent. The symbols of mathematics, however,
are peculiar in that they are symbols either of
other symbols or of operations with symbols. This
peculiarity is what is often expressed by saying that
pure mathematics is abstract, or formal, or that it does
not mention anything at all. Symbols may contain
necessary parts, e.g., the negative, and words like ‘the’
and ‘which,’ which themselves have no specific referents.
The study of such non-symbolic structural elements of
symbols is the business of grammar.

These indications of structure appear in ordinary
language in a bewildering variety of forms. The
inflexions, the conjunctions, distributives, auxiliary verbs,
some of the prepositions, the main use of the copula, etc.,
all have this function. In mathematics, owing to the
simplicity of its outlook, these structural elements are
reduced to the minimum; otherwise such symbols for
counting operations as two and three, or such symbols
of symbols as algebraic expressions could never be
handled systematically. Recent views on mathematics
show a refreshing reaction from the logical mysticism or
88arithmosophy of Frege, Couturat and others, prevalent
at the beginning of the century. It is clearly felt that
an account which does not invoke supersensible entities
must be given of what mathematicians do.

Some, like Wittgenstein, have been able to persuade
themselves that “The propositions of mathematics are
equations, and therefore pseudo-propositions,” and that
“the method by which mathematics arrives at its equations
is the method of substitution. For equations
express the substitutability of two expressions, and we
proceed from a number of equations to new equations,
replacing expressions by others in accordance with the
equations.” 1118 Such a view can be presented without the
background and curtain of mysticism which this author
introduces. Those parts of mathematics, the Theory
of Sets of Points, for instance, which do not seem to be
merely concerned with equations then remain to be
accounted for.

Others maintain with Rignano 2119 that mathematics
throughout is merely the performance of imagined
physical experiments, recorded and represented in
symbols. This amplification of the view of James Mill 3120
and Taine, though it fits some parts of mathematics
well enough, is less plausible for others. As Rignano
develops it, too little importance is assigned to symbols;
hightly systematized sets of symbols such as those of
mathematics are something more than a mere means of
representing our mental performances. They become,
as it were capable of performing on their own account.
They become thinking machines which, suitably manipulated,
89yield results which cannot be foreseen by any
process of imagining physical experiments.

A third school would present mathematics not as a
thinking machine, but as a set of directions for the use
of such a machine, the machine in question being the
mind. For this school mathematics would contain no
statement but only commands or directions. The
problem then becomes what exactly mathematicians
are told to do.

It is probable that the answer to this vexed question
as to the nature of mathematics will be found to consist
of a combination of these varied doctrines. There is
no good reason for supposing that mathematics is
fundamentally homogeneous, although its possession of
a single symbol system makes it appear so. The known
readiness with which not only single symbols but whole
systems of symbols may acquire supernumerary uses
should make us ready to allow this possibility. It is
plain that some parts of mathematics are concerned in a
special way with the discussion of other parts. “It may
be that when logic is wholly emancipated from metaphysics,
logicians will devise a grammar of logistic
language. Perhaps they will then call it the grammar
of logic, and logistic language will be called logic. All
that is valuable in the so-called logic will remain as
component elements of a grammar — a grammar of the
science of reasoning with language.” 1121

Returning from this excursus, it is important to
remember that a reference, as described above at page
62, is a set of external and psychological contexts
linking a mental process to a referent. Thus it is
extremely unlikely that any two references will ever be
strictly similar. In asking, therefore, whether two
symbols are used by the same reference — especially
when the users thereof are two persons with their
different histories — we are raising a question of degree.
90It is better to ask whether two references have sufficient
similarity to allow profitable discussion. When such
discussion is possible the references are said to be ‘the
same.’ No means are at present available for directly
comparing references. We have to judge by indirect
evidence derived mainly from observing the further
behaviour of the parties concerned. We notice whether
doubt and certainty arise at the same points, whether
both admit alternatives at the same points, and so
on. But for many important questions in the theory
of Grammar, especially when discussing the degree to
which the emotive functions of language interfere with
the referential, there is urgent need for some more
easily applicable test. The only hope is in further
analysis of the contexts operative in reference, with a
view to selecting from the many contextual factors
those which are determinative; and meanwhile a clear
realization of the complexities involved may prevent
unnecessary dogmatism.

When a symbol seems to stand for two or more
referents we must regard it as two or more symbols,
which are to be differentiated. This Canon guards
against the most obvious kind of ambiguity, that of
top (mountain), and top (spinning), for instance. We
differentiate these symbols by the aid of a Second Canon
which concerns what is usually called Definition, and is
also of the utmost importance.

When we encounter a symbol which we do not comprehend
we take steps, if interested, to have another
symbol, which we can interpret, provided, whose
referent is the same. Then we can say “I know what
symbol A means; it means the same as symbol B.”
(When scholars say ‘chien’ means ‘dog,’ they should
say that ‘chien’ and ‘dog’ both mean the same.)
Similarly if a symbol is long or awkward to use, or
likely to be misunderstood, we take a new convenient
symbol and use it instead. In both cases the same
process, Definition, is occurring. The details of the
91technique of definition, as required constantly in discussion,
call for special study and will be dealt with in
Chapter VI. below. A foundation-stone is laid in the
Second Canon of Symbolism, the Canon of Definition: —

II. — Symbols which can be substituted one for
another symbolize the same reference.

By means of this Canon we substitute for the
ambiguous symbol ‘top’ the synonym ‘mountain top’
or ‘spinning top,’ and the ambiguity is removed. But
this is not the only use which we make of the Canon.
Its importance is belied by its modest simplicity. It
is the guarantor of mathematics. The systematization
of our symbols (for which we may substitute the phrase
“the organization of our thought”) is achieved by its
application. It is plain for instance that the two
symbols ‘The King of England’ and ‘the owner of
Buckingham Palace’ have the same referent. They
do not however symbolize the same reference, quite
different psychological contexts being involved in the
two cases. Accordingly they are not substitutes one
for another in the sense required in this Canon.
Symbols which are substitutes and so can be used to
‘define’ 1122 one another not only have the same referent
but symbolize the same reference. Such symbols are
usually said to have the same ‘connotation,’ a misleading
and dangerous term, under cover of which the
quite distinct questions of application of reference and
correctness of symbolization (cf. p. 102 below) are
unwittingly confused. Connotation will be further
discussed in Chapter IX.

But there are more dangerous booby-traps in
language than the plain equivoque, and “certain it is,”
as Bacon has it, “that words like a Tartar's bow do
92shoot back upon the understanding and mightily
entangle and pervert the judgment.” Those complex
symbols, known as propositions, which ‘place’ referents
(cf. Canon VI infra) can be either Contracted
or Expanded. “Hamlet was mad” is a contracted
symbol, needing to be expanded before it can be
discussed. “Hamlet was mad on the stage” or “in
my interpretation of the play” may be expanded
symbols for what is referred to. The question is of
the greatest importance because of its bearing on the
distinction between true and false. It leads to the
Third Canon of Symbolism, the Canon of Expansion: —

III. — The referent of a contracted symbol is the referent
of that symbol expanded.

The consequences of infringing this Canon are
sometimes called Philosophy, as little by little we
shall proceed to show.

It is an obvious result of this Canon that the first
thing to do when a disputed symbol is encountered is
to expand it, if possible, to its full form — to such a form,
that is, as will indicate the sign-situations behind
the reference it symbolizes. Instances of this expansion
occur continually in all scientific discussion. In the
last chapter we had occasion to expand ‘table’ and
‘see’ and later on we shall endeavour to expand
‘meaning’ in all possible directions. Unfortunately in
the absence of any systematic theory of interpretation,
no definite ordering of the levels at which we refer has
hitherto been made. The idea even of a level of
reference remains vague. Yet when we refer to ‘that
animal,’ and then later, after further study of its footprints
perhaps, to ‘that lynx,’ 1123 our reference will be to
the same referent but at different levels of interpretation
93in a definite sense involving the number of applications
of interpretative processes and the complexity of these
processes. In such relatively simple cases matters are
easy to set straight; in more complicated cases — if we
speak of government, credit, patriotism, faith, beauty,
emotion, etc. — it is not so. All our usual discussion of
subjects of general interest suffers from the uncertainty,
difficult even to state, as to the level of interpretation, of
reference, at which we are symbolizing. All those
engaged in education know what ‘levels of reference’
stand for. The fuller analysis of the question is of
great urgency. Something towards it was attempted
in Chapter IV. It is a pity, however, that those very
persons who by their analytic ability would be most
likely to succeed, should be so reluctant to take up
problems until they have been elaborately formulated.

Meanwhile such is the chaos of symbolic apparatus
in general that, instead of expansions, mere symbolic
overgrowths are most usually what are provided by way
of elucidation of doubtful symbols, thus leading to
greater confusion than would the contractions which
they replace. Instances are given in the following
paragraph. Both contractions and pseudo-expansions
have the same result — the peopling of the universe
with spurious entities, the mistaking of symbolic
machinery for referents. The only permanent cure is
the discovery of the appropriate expansion by inquiry
into the sign-situation leading to the reference which
is doubtfully symbolized. 1124

It can in fact be recognized without difficulty that
until this is done it is idle to raise such further questions
as its truth or its relations to other symbols; for
a contracted symbol does not make plain the ‘place’
of its referent, and so cannot be investigated. The
distinction between true and false symbols is a matter
94which cannot be discussed profitably in general terms,
i.e., by means of contractions or linguistic shorthand.
It must be left in each case to the specialist, who being
familiar with the actual sign-situations involved can
decide within his particular field of reference which
symbols are true and which not. It is owing to such a
discussion in contracted symbols that what is known
as the Problem of Truth has arisen. Instead of treating
each case of adequacy on its own merits, epistemologists
will have it that because they can use one word
as a convenient shorthand sign to refer to all true
symbols, there must be something for them to investigate
apart from true and false propositions. No
problem arises over any true proposition when recognized
as such, and to raise a bogus problem here is
quite as unnecessary as to assume a universal ‘redness’
because red things are every one of them red. Classes
are now recognized as symbolic fictions, and logisticians
will only be logical when they admit that universals
are an analogous convenience. The World of Pure
Being will then be definitively denuded of its quondam
denizens, for which the theory of Universals was an
attempted explanation. It should be noted that our
symbolic machinery (similarity, etc.), becomes both
more valuable and more comprehensible when these
desiccated archetypes have faded away.

By way of explanation of these symbolic conveniences
a few considerations may be added. Modifications
of our sense organs, and ‘things’ as we come
to know them through the interpretation of these signs,
are always complex or parts of a complex. Even the
tiny speck which, in virtue of a certain disturbance
in the colour apparatus of an eye, we call a barely visible
star is surrounded by a dark field. All that there is
in such a sign for us to talk about is this complex, and
we can talk about it in various ways. We can say
“the speck is in the field” or “surrounded by the
field” or “part of the field” or “related to the field
95by the relation of being enclosed by”; or we can say
“this which has the property of being a speck is
related to that which has the property of being a field
by the relation of inclusion.” These are alternative
locutions, equally true. ‘Speck in field’ is a name,
and so is ‘speck.’ On other occasions, however, we
wish to symbolize references under circumstances in
which the same names are correctly reapplied. We
have to economize in our symbolic material; we have to
use it over and over again, and in a systematic fashion,
under pain of failure to communicate. Now if instead
of the name ‘this speck’ we use the more luxuriant
symbolic growth, ‘this which has the property of being
a speck,’ we shall be tempted to suppose that the
‘thises’ on different occasions stand for different referents
but that ‘the property of being a speck’ stands for one
and the same.

In this way universal ‘qualities’ arise, phantoms
due to the refractive power of the linguistic medium;
these must not be treated as part of the furniture of
the universe, but are useful as symbolic accessories
enabling us to economize our speech material. Universal
‘relations’ arise in a precisely similar fashion,
and offer a similar temptation. They may be regarded
in the same way as symbolic conveniences. The claims
of ‘similarity’ and ‘dissimilarity’ which on account of
purely symbolic arguments (cf. Russell, Some Problems
of Philosophy, p. 150) are often supposed to be peculiar
are in no way different.

In all cases, even in this case of similarity, the
invention of non-existent entities in order to account
for the systematic use of symbols is an illegitimate
procedure. Were there other evidence for them not
deriving merely from symbolic necessities 1125 it would be
96a different matter. As it is they stand on the same
footing as the ‘faculty’ of knowing in psychology.
The occurrence of similars does not compel us to
recognize ‘similarity,’ a universal, any more than the
occurrence of knowledge forces us to recognize a faculty
of knowing. It merely compels us to recognize that
similars do occur. That things are similar is natural
knowledge. To make it, by exploiting the economy
of symbolisms, into a basis of metaphysical knowledge
— into a proof of another world of pure being where
entities ‘subsist’ but do not exist — is unwarrantable.
No argument about the world is valid if based merely
upon the way a symbol system behaves. 1126 Such arguments
97can give knowledge only about the symbol system
in question. This knowledge is often of great value.
All methods of distinguishing symbols proper, i.e.,
names, from symbolic accessories are important.

We have spoken above of reflection and refraction
by the linguistic medium. These metaphors if carefully
considered will not mislead. But language, though
often spoken of as a medium of communication, is best
regarded as an instrument; and all instruments are
extensions, or refinements, of our sense-organs. The
telescope, the telephone, the microscope, the microphone,
and the galvanometer are, like the monocle or the eye
itself, capable of distorting, that is, of introducing new
relevant members into the contexts of our signs. And
as receptive instruments extend our organs, so do
manipulative instruments extend the scope of the motor
activities. When we cannot actually point to the bears
we have dispatched we tell our friends about them or
draw them; or if a slightly better instrument than
language is at our command we produce a photograph.
The same analogy holds for the emotive uses of
language: words can be used as bludgeons or bodkins.
But in photography it is not uncommon for effects due
to the processes of manipulation to be mistaken by
amateurs for features of the objects depicted. Some of
these effects have been exploited by experts so as greatly
to exercise the late Sir Arthur Conan Doyle and his
friends. 1127 In a similar fashion language is full of elements
with no representative or symbolic function, due solely to
its manipulation; these are similarly misinterpreted or
exploited by metaphysicians and their friends so as
greatly to exercise one another — and such of the laity
as are prepared to listen to them.

The fictitious entities thus introduced by language
98form a special variety of what are called fictions. But,
as Vaihinger's own use shows, this term is very vague
and so-called fictions are often indistinguishable from
hypotheses, which are simply unverified references.
Certain abstractions, like the ‘economic man,’ are of
this nature, though, being purely methodological, they
are not believed in; on the other hand, many idealizations
and imaginative creations, such as Don Juan and
the Übermensch, may some day find their referents.
Hamlet and Goethe's Urtier appear not to be hypotheses,
since they are dated and placed where history has no
room for them; they are fictitious in the sense that
Shakespeare or Goethe's thought had no single referent.
We, of course, may refer to these thoughts; more
usually we attempt only to reproduce them. But all
fictions of this kind must be clearly distinguished from
those due to manipulations of language itself. Vaihinger
has not sufficiently emphasized this distinction;
owing perhaps to an incomplete analysis of the relations
of language and thought — shown by his use of the
terms ‘Begriff’ and ‘begreifen’ in the discussion of
abstractions and knowledge. 1128 Linguistic fictions occur
in two ways, either through a misunderstanding of the
function of symbolic accessories such as ‘liberty’ or
‘redness,’ so that in making a reference to free actions
or red things the user supposes himself to be referring
to something not in time and space; or through hypostatization
of such connective structural machinery as
‘or,’ ‘if,’ ‘not,’ etc., to which only logicians are prone.
The use of the term ‘concept’ is particularly misleading
in linguistic analysis. There is a group of
words, such as ‘conception,’ ‘perception,’ ‘excitation,’
which have been a perpetual source of controversy
since the distinction between happenings inside and
happenings outside the skin was first explicitly recognized.
Processes of perceiving caused in an interpreter
by the action on him of external objects have been
99commonly called ‘perceptions,’ and so, too, by a very
intelligible confusion, discussed in our next chapter as
the ‘utraquistic fallacy,’ have those objects themselves.
Other processes, more abstract or less obviously caused
references, have similarly been called ‘conceptions.’
But whereas the double sense of the term ‘perception’
involves merely a confusion between two possible
referents or sets of referents, the one inside the head
and the other outside, the term ‘concept’ when thus
duplicated has been a special inducement to the creation
of bogus entities. It has often been assumed that the
referents of these more abstract processes, since they
appeared to be simple, were quite different from those
of the mental processes which occurred when the referents
were ‘given’ in perception. A transcendental
world of ‘concepts’ has therefore been envisaged by
philosophers; while even psychologists who elected
to call themselves ‘conceptualists’ in recognition of
the fact that concepts are mental — as opposed to the
transcendental (scholastic ‘realist’) or the non-psychological
(nominalist) account — have frequently been led
by their terminology to take an inaccurate view of
symbol situations.

In discussions of method or of mental processes,
‘concepts’ or abstract references may, of course, be
themselves talked about; and in this special case words
will properly be said to stand for ideas. But it is not
true to say that in ordinary communication we are thus
referring to our own mental machinery rather than to
the referents which we talk ‘about’ by means of that
machinery. Words, as we have seen, always symbolize
(cf. p. 11) thoughts, and the conceptualist is apt to imply
that the very special case of the construct or concept imagined
for the purpose of an attempted scientific reference
or classification, and then itself examined, can be
generalized. He then states that the word is not a
mere word as the nominalist holds, but stands for a
conceptual symbol. In opposition to the believer in
100a single discoverable entity for which words symbolizing
general references stand, he is right; but by those who
do not admit that they are talking ‘about’ nothing
when they appear to have referred to unjustifiable
entities, his vocabulary is likely to be misunderstood. 1129

Such linguistic accessories may be used without
danger, provided they are recognized for what they are.
They are conveniences in description, not necessities
in the structure of things. This is shown by the fact
that various alternatives are open to us in describing
any referent. We can either use a grammar of ‘substantives’
and ‘attributes’ 2130 (nouns and adjectives),
or one of ‘Events’ and ‘Objects,’ 3131 or of ‘Place’ and
‘Referent,’ 4132 according as we favour an Aristotelian
outlook, or that of Modern Physics, or a pictorial exposition
of the views here advocated. To discuss such
questions in any other spirit than that in which we
decide between the merits of different Weed killers is
to waste all our own time and possibly that of other
people.

In a similar way, from the question, What is Truth?
an apparently insoluble problem has arisen. In
Chapter III. however the problem was seen to be
soluble as part of the theory of Interpretation. It will
101be convenient here to define a true Symbol as distinguished
from a true Reference. The definition is
as follows: — A true symbol = one which correctly records
an adequate 1133 reference. It is usually a set of words
in the form of a proposition or sentence. It correctly
records an adequate reference when it will cause a
similar reference to occur in a suitable interpreter.
It is false when it records an inadequate reference.

It is often of great importance to distinguish between
false and incorrect propositions. An incorrect symbol
is one which in a given universe of discourse 2134 causes
in a suitable interpreter a reference different from that
symbolized in the speaker. Thus if we say, “Charles I.
died in his bed, making witty remarks,” our symbol
is more likely to be incorrect than our reference false, for
it is no rash suggestion that the referent is Charles II.'s
death in his bed. But in many cases such an audacious
exegetic is unwarranted, and it will then be a more
difficult matter to decide which is occurring. In the
opposite case when, e.g., we say, “The sun is trying
to come out,” or “The mountain rises,” we may clearly
be making no different references than if we were to
give a scientific description of the situation, but we may
mean these assertions to be taken ‘literally.’ By taking
an assertion literally is meant interpreting our symbols
as primary symbols, i.e., as names used with a reference
fixed by a given universe of discourse. When for any
reason, such as poverty of language, no symbol is
at hand we can choose a symbol whose referent is
102analogous to our referent and transfer this symbol.
Then if the speaker fails to see that such symbols are
metaphorical or approximative only, i.e., takes them
literally, falsity arises, namely the correct symbolization
of a false reference by which the interpreter could
be misled. If on the other hand the speaker makes
a true reference, but uses symbols such that a suitable
interpreter, rightly interpreting, makes a false reference,
then the symbol is incorrect.

Incorrectness may plainly have degrees, for if, when
my pipe is out, I say, “My pipe is alight,” then this
symbol, “My pipe is alight,” is sufficiently correct to
characterize its referent but not to place it. In other
words, it is good enough for the investigator to be able
to look for its referent among events, and to exclude it
on the ground that the place it claims is filled by the
referent of “My pipe is out.” It may also be good
enough, according to the actual context, for him to go
and look for it among other likely orders of referents,
gustatory, olfactory and thermal sensations, images and
so forth. If he can find it he may be able to expand
the incorrect symbol, possibly changing every word in
the process. Similarly, once convinced that my pipe is
out, I may be able myself to expand my symbol to “My
pipe feels as though it were alight.”

A group of questions arise out of this instance,
which require a Fourth Canon, the Canon of Actuality,
to clarify the situation: —

IV. — A symbol refers to what it is actually used to
refer to; not necessarily to what it ought in
good usage, or is intended by an interpreter, or
is intended by the user to refer to

The assertion considered above may or may not have
referred to a referent like that for which it would be
correctly used. I may admit or deny that my referent
was some feeling and not burning tobacco. Accordingly,
by Canon 1., we have here a group of symbols
103appearing to be one symbol, and we must select that
which is actually being used. When we cannot so
select, nothing more can be done beyond framing a
collection of unambiguous symbols for future use in
analogous cases. 1135 But suppose that we were led to
state, after the manner of formal logicians, that a
referent such as ‘non-existent combustion of tobacco’’
is involved, we should appear to be confronted by a
problem as to how we can refer to what is not there
to be referred to. This problem, which is of no interest
in itself, is mentioned here because it is typical
of the difficulties which arise through treating an incomplete
system of defective symbols as though it were
a complete system of perfect symbols. Within a minor
system of symbols which has been wrought into a high
degree of complexity, such contradictions, if they ensue
from a legitimate manipulation of symbols, are a helpful
indication of some imperfection still remaining. Mathematics
is a case in point. Faced with such a contradiction,
the mathematician proceeds to improve his symbolism,
and we should follow his example rather than suppose
that we have proved some curious eccentricity in the
universe.

Two other questions arise which deserve an answer.
The first is “How do we know that ‘pipe alight now’
claims the same place as ‘pipe out now,’ while ‘pipe
foul now’ does not?” The answer is, in the words of
the old tale, “By experience.” We possess in familiar
fields vast accumulations of such knowledge. We
know, for instance, that ‘x is green’ and ‘x is red’
and ‘x is blue’ all claim the same place for their
referents; as do ‘x is dark’ and ‘x is light.’ We
also know that ‘x is green’ and ‘x is dark’ and ‘x is
vivid’ do not make conflicting claims. In fields with
which we are unfamiliar the main difficulty is precisely
104in gaining such knowledge. We need this knowledge
in order to perfect our symbols, just as we need
perfected symbols in order to advance our knowledge.

The other question is, “Why not say that since no
referent for ‘My pipe is out’ was to be found where
we were led to look for it, there was no referent?”
But there was a reference — though not to the referent
suggested at first sight. The problem of finding the
actual referent is here, as always, that of tracing out the
causal connections or contexts involved, in the manner
indicated in Chapter III.

One special difficulty with regard to complex symbols
calls for a Canon whose functions may not be evident
at first sight, though it is necessary for the avoidance
of nonsense in our discourse. It concerns the building
up of complex symbols from those which are simple
or less complex. It is plain that if we incorporate in
one symbol signs which claim the same place, whether
e.g., colour (red — yellow) or shape (round — square), our
proposed symbol is void. This Fifth Canon is called
the Canon of Compatibility: —

V. — No complex symbol may contain constituent
symbols which claim the same ‘place.’

It is therefore important at once to make clear what
is done when a symbol ‘places’ a referent. Since the
days of Aristotle, three formulæ, traditionally known
as the Laws of Thought, have received much attention,
civil and uncivil, from logicians. They have been
variously interpreted as laws which the mind obeys but
which things need not, as laws which things obey but
which the mind need not, as laws which all things (the
mind included) obey, or as laws which nothing need
obey but which logic finds strangely useful. For
Symbolism they become a triad of minor Canons which
help to keep the Cathedral of Symbolism in due order.
First comes the Law of identity — quaintly formulated
as ‘A is A’; a symbol is what it is; i.e., Every symbol
105has a referent. The second is the Law of Contradiction —
‘A is not not-A’; no symbol refers to what it does
not refer to; i.e., No referent has more than one place
in the whole order of referents. The third is the Law
of Excluded Middle — ‘A is either B or not B’; a symbol
must have a given referent or some other; i.e.,
Every referent has a fixed place in the whole order of
referents. For this triad, by Canon II. we may substitute
the following formula, which is then the Sixth
Canon of Symbolism: The Canon of Individuality —

VI. — All possible referents together form an order,
such that every referent has one place only in that order.

One difficulty with regard to ‘place’ may be usefully
commented on. It is rather a symbolic accessory (cf.
p. 94 above) than an actual symbol. In any false
assertion, we have implied, two things must be clearly
distinguished (1) the referent to which we are actually
referring (2) an alleged referent to which we believe
ourselves to be referring. Only the first of these has a
‘place’ in the whole order of referents.

We can, using alternative language, say either that
in a false assertion we are believing a referent to be in
a ‘place’ in which it is not, or that we are believing
ourselves to be referring to a different referent from
that to which we are actually referring. We can for
instance either say that in two contradictory assertions
we are referring to the same referent but assigning to it
different ‘places,’ or we can say that we are referring to
two different referents and assigning them to the same
‘place.’ These alternative locutions involve subtle
shifts in the references using both ‘referent’ and ‘place,’
and accentuate the important consideration that the
distinction between the reference of these terms is
merely artificial. There is no difference between a
referent and its place. There can be no referent out
of a place, and no place lacking a referent. When a
referent is known its place also is known, and a place
106can only be identified by the referent which fills it.
‘Place,’ that is, is merely a symbol introduced as a
convenience for describing those imperfections in
reference which constitute falsity.

We have shown that for all references, between the
referent and the act there are always intervening sign-situations.
In the simplest case, that of the true direct
judgment of perception, there may be only one such sign-situation
(discussed in Chapter III.). In a false proposition
there will be a similar sign chain with the
difference that some misinterpretation occurs. It is not
however always necessary in order to translate a false
proposition into a true one to discover where the misinterpretation
occurred; a new sign chain abutting on
the same referent may be substituted. In expansion,
however, such discovery is necessary, and the difficulty
explains our preference for Translation over Expansion.
In education and controversy the discovery of the
misinterpretation is usually the more essential step.

These Canons control the System of Symbols known
as Prose. If by themselves they do not prove sufficient
to keep our speech from betraying us, any others which
may be required will be of the same nature. A set of
symbols will only be well organized, or form a good
prose style, when it respects these Canons. Only such
a set will allow us to perform with safety those transformations
and substitutions of symbols by which
scientific language endeavours to reflect and record its
distinctions and conclusions — those operations which,
as we have seen, appeared to primitive man to partake
of the nature of magic. Moreover, only such a set
will enable the philosopher to discuss more important
matters than his own or his colleagues' peculiarities of
expression.108

Chapter VI
The theory of definition

The first cause of absurd conclusions I ascribe to
the want of method; in that they begin not their
ratiocination from definitions. — Hobbes.

“Do, as a concession to my poor wits, Lord
Darlington, just explain to me what you really
mean.” — “I think I had better not, Duchess.
Nowadays to be intelligible is to be found out.” —
Lady Windermere's Fan.

There is at present no theory of Definition capable
of practical application under normal circumstances.
The traditional theory, in so far as it has not been lost
in the barren subtleties of Genus and Differentia, and in
the confusion due to the term ‘Connotation,’ has made
little progress — chiefly on account of the barbarous
superstitions 1136 about language which have gathered on
109the confines of logic from the earliest times. Four
difficulties have stood in the way and must first be
removed.

Firstly, do we define things or words? To decide
this point we have only to notice that if we speak about
defining words we refer to something very different from
what is referred to, meant, by ‘defining things.’ When
we define words we take another set of words which may
be used with the same referent as the first, i.e., we substitute
a symbol which will be better understood in a
given situation. With things, on the other hand, no
such substitution is involved. A so-called definition of
a horse as opposed to the definition of the word ‘horse,’
is a statement about it enumerating properties by means
of which it may be compared with and distinguished
from other things. There is thus no rivalry between
‘verbal’ and ‘real’ definitions. 1137

The words by means of which these properties are
enumerated do, of course, give us a substitute symbol —
either a complete analysis, or as abbreviated by classificatory
methods (the usual ‘genus and differentia’ type)
— with the same referent (the horses) as the original
symbol; but rather by way of corollary than as the
main purpose of the analysis. Moreover, this process
is only possible with complex objects which have been
long studied by some science. With simple objects, or
those which for lack of investigation are not known to
be analysable, as well as with everything to which
classificatory methods have not yet been applied, such
a method is clearly not available, and here other symbols
must be found as the substitutes which symbol-definition
seeks to provide. Such, in outline, is the solution of
the long-standing dispute between the advocates of real
and symbolic definitions.110

The second difficulty is closely related to the above.
Though definition be symbol-substitution, definitions
have usually, for grammatical reasons, to be stated in
a form which makes them appear to be about things.
This is because we are in the habit of abbreviating such
symbols as “the word ‘fire’ refers to the same referent
as the words ‘what burns’” to “fire is what burns”;
and of saying “Chien means ‘dog,’” when we ought to
say “the word chien and the word ‘dog’ both mean the
same animal.” 1138

Thirdly, all definitions are essentially ad hoc. They
are relevant to some purpose or situation, and consequently
are applicable only over a restricted field or
‘universe of discourse.’ For some definitions, those of
physics, for instance, this universe is very wide. Thus
for the physicist ‘energy’ is a wider term than for the
schoolmaster, since the pupil whose report is marked
“without energy” is known to the physicist as possessing
it in a variety of forms. Whenever a term is thus taken
outside the universe of discourse for which it has been
defined, it becomes a metaphor, and may be in need
of fresh definition. Though there is more in metaphor
than this, we have here an essential feature of symbolic
metaphorical language. The distinction between this
and emotive metaphorical language is discussed later at
pages 239-40.

Fourthly there is the problem of ‘intensive’ as
opposed to ‘extensive’ definition which comes to a head
with the use of the terms ‘denote’and ‘connote.’ In
Chapter IX. the artificiality of these distinctions will be
urged. Here it is only necessary to point out that two
symbols may be said to have the same connotation when
111they symbolize the same reference. An intensive or
connotative definition will be one which involves no
change in those characters of a referent in virtue of
which it forms a context with its original sign. In an
extensive definition there may be such change. In other
words when we define intensively we keep to the same
sign-situation for definiendum and definiens, when we
define extensively this may be changed.

We are now in a position to grapple with the
difference between definitions and ordinary assertions.
“Gorillas are animals” and “Gorillas are affable” are
unlike one another in the respect that the first appears
to be certainly true as soon as we understand it, while
the second may be doubted. From “This is a gorilla”
it follows directly that “This is an animal,” but not
that it is an affable one. If we look for a distinction in
essential connection between animality and gorillarity
on the one hand, and gorillarity and affability on the
other we shall make but indifferent use of our leisure.
But if the difference be sought in its proper place, that
is, between or in 1139 the references, it will be found that
the definition actually used in the first case includes
animal, so that in speaking of a gorilla we have spoken
of an animal, and are therefore able to refer again without
diffidence to what we have already referred to; while
affability was not so included. The relevant definition,
in fact, is the one actually used. 2140112

To attempt now a fresh attack upon the essential
problem of how we define, or attain the substitute
symbols required in any discussion. We know 1141 that
‘A symbol refers to what it has actually been used to
refer to.’ We shall cease then to assume that people are
referring to what they ‘ought’ to have referred to, and
consider only what they actually do refer to. The point
to be met in every discussion is the point actually
advanced, which must be first understood. We have,
that is, in all cases to find the referent. How can this
best be done?

The answer is simple and obvious. Find first, it
runs, a set of referents which is certainly common to
all concerned, about which agreement can be secured,
and locate the required referent through its connection
with these.

It is fortunate that the types of fundamental connections
with which discussions are concerned are few
in number, though we are apt to believe, such is the
multifarious complexity of our talk, that things are
connected in any number of ways. Whether this
poverty is due to the trammelling influence of language,
a larger number of connections being quite, not merely
partially, unmanageable by naive talkers, whether it
is due to the structure of the brain, or whether it is due
to an actual simplicity in the universe, need not here
be considered. For practical purposes the fundamental
connections which can be used in definition are limited
to those which the normal mind can think of when
directly named. Let us consider, for instance, the
growth of the abstraction which we name a spatial
relation. In all our references to spatial objects certain
common elements or strands are active. Originally to
think of space as opposed to spatial objects we had to
think in rapid succession of a variety of spatial objects
in order that the common elements in the references
should stand out. In time we became able to use these
113common, i.e., general references independently without
requiring them to be built up anew on each occasion.
We are now able to use them merely upon the vicarious
stimulus of the symbol ‘spatial relation.’ A normal
mind, however, except in the few cases in which such
abstractions have universal value, still requires the aid
of instances, analogies and metaphors. The fewness
of these abstractions saves the linguistic situation. If
we employed, say, a hundred radically different types
of connections (still a small number) the task of limiting
the misunderstandings due to the variety in our
references would have proved impossible.

The fundamental connections being thus so few,
the task of a theory of definition narrows itself down to
the framing of a list. All possible referents are connected
in one or other, or several, of these fundamental
ways with referents which we can all succeed in identifying.
We must not assume in referring to any given
fixed point of agreement from which we find we are
able to start that we do more than agree in identifying
this. We must be careful to introduce our starting-points
in such a way that they do not raise fresh problems
on their own account. That is to say, we must select
them with reference to the particular universe of discourse
in which our definienda fall. Thus, if we wish
to indicate what we are referring to when we use the
word ‘Beauty’ we should proceed by picking out certain
starting-points, such as nature, pleasure, emotion, or
truth, and then saying that what we refer to by
‘Beauty’ is anything lying in a certain relation (imitating
nature, causing pleasure or emotion, revealing truth) to
these points. How this may be done is shown in detail
in the following chapter.

The importance of starting-points having thus been
indicated, namely, to act as signs by which the required
referents may be reached, we may now enumerate some
of the main routes which are useful in finding our way
about the field of reference. The sign-situations here
involved, we must not forget, arise only through and
upon many other simpler interpretations of the kind
discussed in the preceding chapters. It is easy
symbolically to make the situation which arises when
we define appear simple, but if we realize the delicacy
of the processes and adaptations required we shall not
place overmuch trust in face-value comparisons of
symbols (the usual method), but will attempt instead
to consider what actually is happening.

When in a discussion we are asked, ‘Can you
define your terms?’ or complain ‘I do not understand
what you mean by the words you use,’ we endeavour
to discover some route by which understanding, i.e.,
identification of referents, may be secured.

A person thoroughly acquainted with his subject
115and with the technique of Definition should be able,
like the man up aloft in a maze, to direct travellers
from all quarters to any desired point; and it may be
added that to go up the ladder and overlook the maze
is by far the best method of mastering a subject.

Although in no case, as we have already seen, are
relations to be regarded as part of the stuff of nature,
and although when we appear to speak of them we are
merely using them as tools, which does not involve
actual referents corresponding to them, yet when they
are so used there are various distinctions which it is
desirable to make as a matter of convenience. At the
beginning of our inquiry we described the relation
which could be said to hold between symbol and
referent as an imputed relation. To have described it
simply as an indirect relation would have omitted the
important difference between indirect relations recognized
as such, and those wrongly treated as direct.
Thus the relation between grandfather and grandson
is much more indirect than that between father and son,
and can be analysed into two paternal relations — ‘ being
the father of the father (or mother) of.’ Few people
would suppose that a direct relation was here involved,
since all family relations are highly indirect. But love,
hate, friendship, sympathy, etc., are very commonly
spoken of and regarded as direct, though on examination
their indirectness is at once discovered. The whole
of social psychology is, however, infested with imputed
relations of this type, for an explanation of which such
hypotheses as that of group-consciousness are often
invoked.

The distinction between simple and complex relations
on the other hand is somewhat different. Indirectness
is only one kind of complexity, and direct
relations need not be simple. For instance, the relation
of ‘being a benevolent uncle to’ is complex; it is a
blend of the two relations ‘well disposed towards’
and ‘avuncularity.’ The similarity between one pea
116and another is complex, being a blend of similarities
in respect of greenness, hardness, edibility, etc. These
considerations, elementary though they may appear,
are of use whenever we have to treat of relations.

The routes, then, which we seek in our endeavour
to reach a desired referent are the obvious relations
in which that referent stands to some known referent.
The number of possible relations is indefinitely large,
but those which are of practical use fortunately fall,
as we have already explained, into a small number of
groups. So that as a preliminary classification 1142 we get
such a list as this: —

1. Symbolization

This is the simplest, most fundamental way of
defining. If we are asked what ‘orange’ refers to,
we may take some object which is orange and say
“‘Orange’ is a symbol which stands for This.” Here
the relation which we use in defining is the relation
discussed in Chapter I. as constituting the base of our
triangle. It is, as we mentioned, an imputed relation
reducible to a relation between symbol and act of
reference and a relation between act of reference and
referent. Our starting-point is the word ‘orange,’
our route of identification is this relation. The required
referent is This. What we are doing in fact here is
directly naming.

But, it will be said, This merely tells us that
‘orange’ is applicable in one case; what we wish to
know is how it is applicable in general; we wish to
have the definition extended so as to cover all the
referents for which ‘orange’ is a suitable symbol.
This generalization may be performed for all types
of definitions in the same manner by the use of
similarity relations. We may say “‘Orange’ applies
to this and to all things similar in respect of colour to
this.” In practice the discrimination of one similarity
relation from others generally requires the use of
117parallel instances, analogies in fact, of the simplest
order.

2. Similarity

Thus similarity itself may be used as a defining
relation. Our required referent is like a chosen referent.
If we are asked what the symbol ‘orange’ refers to,
we may define this symbol by taking something which
is orange and saying “To anything which is like this
thing in respect of colour the symbol ‘orange’ is
applicable.” Here we have substituted for ‘orange’
‘like this in respect of colour,’ and the referents of
both symbols are the same. Our starting-point is
This and the relation is Likeness, and anyone who
knows what ‘This’ stands for (i.e., is not blind) and
knows what ‘Likeness’ stands for will get there.

3. Spatial relations

In, On, Above, Between, Beside, To the right of,
Near, Bigger than, Part of, are obvious examples.
“‘Orange’ is a symbol for the colour of the region
between red and yellow in a spectrum (and of any
colour like this).” It will be noted that the naming
relation is involved in this as in every definition, and
that the definition is always extendable by a similarity
relation. It is curious that some of these symbols for
spatial relations are unsymmetrical. Thus we have
‘on’= ‘above and in contact with,’ but no abbreviation
for ‘under and in contact with,’ except such ambiguous
words as ‘supporting.’ We may further note that
most of the common uses of ‘on’ are so strangely
metaphorical that it has even been doubted whether
there is not some simple unanalysable relation which has
not yet been noticed. The right approach to problems
of metaphorical extension will be considered later in
this chapter.

4. Temporal relations

‘Yesterday’ is the day before to-day; ‘Sunday’
118is the first day of the week; ‘The end of the war’
is x months after event y; ‘Lighting-up time’ is x
minutes after sunset.

5. Causation: physical

‘Thunder’ is what is caused (not by two clouds
bumping but) by certain electrical disturbances. ‘Sawdust’
is what is produced, etc.

6. Causation; psychological

‘The Unconscious’ is what causes dreams, fugues,
psychoses, humour and the rest. ‘Pleasure’ is ‘the
conscious accompaniment of successful psychic activity.’

7. Causation: Psycho-physical

In addition to the examples given in the following
chapter in connection with Beauty, we may define
‘A perception of orange’ as ‘the effect in consciousness
of certain vibrations falling on the retina.’

Causal relations are probably the routes of identification
most commonly employed in general discussion,
as well as in science. Thus a view of great historical
consequence defines the Deity as the Cause of the
Universe, while the importance of Embryology in
zoological classification is due to the causal defining
relations which are thereby provided.

8. Being the Object of a Mental State

The right-hand side of our triangle, Referring, is
one of these; so are Desiring, Willing, Feeling, etc.
Thus ‘Piteous things’ may be defined as those towards
which we feel pity, and ‘Good things’ are those which
we approve of approving.

9. Common complex relations

Some definitions are most conveniently formulated
in complex form. While capable of being analysed
out into sets of simple relations falling under one or
other of the above headings, they are more readily
applicable as popularly symbolized.119

Examples are ‘utility’ (analysable into Nos. 7 and
8), ‘Imitation’ (2 and 7), ‘Implication’ (1 and 8).

10. Legal relations

These are so frequently employed and implied,
though often disguised, that it seems worth while to
give them a separate heading; moreover, they are
subject to an arbitrary test — satisfying the judge.

Examples: ‘Belonging to’ (when = ‘owned by’),
‘Subject of,’ ‘Liable to,’ ‘Evidence of.’ All legal
definitions are highly complex, though none the less
serviceable.

The above relations are those which considerable
experience has shown to be commonly employed in
definitions. Any other relations which might be required
for special purposes equally deserve to be
included in a complete list — Shape, Function, Purpose, or
Opposition, for example. It is therefore neither claimed
that the first eight groups exhaust the relevant elementary
relations, nor that those complex relations which we
have cited can be reduced without remainder to relations
of these types. The whole classification is on a pragmatic
basis, and merely on the level of the most usual
universes of discourse.

It has also proved unnecessary to discuss whether
and in what sense all relations may be logically
reducible to one or more ultimate kinds, 1143 for any such
reduction would make no difference to the value of the
definitions we have been considering in their appropriate
field. Even definitions of considerable complexity,
involving a variety of theories, can be reduced
without difficulty to discussable morsels, and their
validity as substitutes the better examined. This further
illustrates the fact that definitions often go by stages,
as when our inquirer for Cambridge Circus is not
120familiar with the British Museum and requires first
to be directed thither via the Tube from the Marble
Arch.

The question of multiple relations raises no difficulty
in this connection. A multiple relation holds
between a number of terms greater than two. Thus,
Perceiving, as Dr Whitehead has recently insisted,
is a multiple relation holding between a percipient,
an object, and the conditions; and Giving is a multiple
relation holding between a philanthropist, a donation,
and a beneficiary. In defining any of these terms, or
in taking any of them as a starting-point for a route
of definition, we proceed in exactly the same fashion
as with dual relations — except that bearings must be
taken from more than one landmark, when the universe
of discourse demands special accuracy. Otherwise
the Definiendum is not reached. Thus, in defining
some object as what so-and-so saw, it may on some
occasions be necessary to state the conditions — as in
a séance we need to know the strictness of the test;
or in identifying a passing train as an Express we have
to consider the speed of our own train. But much
discussion can be profitably undertaken without such
complex situations arising.

The practical aspect of the above list of routes of
definition deserves to be insisted upon. The reason
for using definitions at all is practical. We use them
to make discussion more profitable, to bring different
thinkers into open agreement or disagreement with
one another. There is, it is true, a more recondite
use of definition derived from this simple primitive
use. Definitions are of great importance in the
construction of deductive, scientific systems, those
automatic thinking-machines for which logic and
mathematics are, as it were, the rules or instructions.
In such a deductive system as mechanics, for example,
it is through the definitions employed that the parts
of the symbolic system are linked together, so that a
121given manipulation of the symbolism will yield comparable
results even when their precise nature is not
foreseen by the manipulator. Thus, for such systems
there comes to be something which is regarded as the
definition of a particular symbol. Given the system,
there will be one and only one definition of a symbol
which is the right or proper definition, in the sense that
the working of the system depends upon the employment
of this definition.

Specialists who are much concerned with such
systems naturally tend to regard all definitions in
the same manner. Yet for many of the most interesting
topics of discussion a quite different attitude
and habit of mind as regards definitions is not
only desirable, but, in fact, necessary, if fruitful
discussion is to be possible. In æsthetics, politics,
psychology, sociology, and so forth, the stage of
systematic symbolization with its fixed and unalterable
definitions has not been reached. Such studies
as these are not far enough advanced for anyone yet
to decide which system is most advantageous and
least likely to exclude important aspects. The most
highly systematized sciences are those which deal
with the simplest aspects of nature. The more difficult
and to many people, naturally, the more attractive
subjects are still in a stage in which it is an open
question which symbolization is most desirable. At
this stage what has chiefly to be avoided is the veiled
and hidden strife between rival systems in their early
forms, which more than anything else prevents mutual
understanding even between those who may be in
agreement. Many terms used in discussions where
‘faith,’ ‘beautiful,’ ‘freedom,’ ‘good,’ ‘belief,’ ‘energy,’
‘justice,’ ‘the State’ constantly occur are used with
no distinct reference, the speaker being guided merely
by his linguistic habits and a simple faith in the
widespread possession of these habits. Hence the
common sight of anger aroused by the hearer's apparent
122obtuseness and wrong-headedness “where the matter
is surely self-evident.”

But even in those rarer discussions in which the
speakers are capable of greater explicitness, the curious
instinctive tendency to believe that a word has its own
true or proper use, which we have seen has its roots
in magic, too often prevents this ability to produce
definitions from taking effect. No doubt other factors
are involved. Lack of practice, literary fetishes concerning
elegance of diction, reluctance to appear
pedantic, defensive mimicry and other protective uses
of language all contribute. But far more important
than these is the instinctive attitude to words as natural
containers of power, which has, as we have shown, from
the dawn of language been assumed by mankind, and
is still supported and encouraged by all the earlier
stages of education.

The correction for this persistent tendency is a
greater familiarity with the more common routes of
definition, and a lively sense, which might easily be
awakened as a part of education, that our use of any
given word to stand for our referent on any occasion
is not due to any particular fitness of the word for that
particular referent, but is determined by all sorts of odd
accidents of our own history. We ought to regard
communication as a difficult matter, and close correspondence
of reference for different thinkers as a
comparatively rare event. It is never safe to assume
that it has been secured unless both the starting-points
and the routes of definition, whereby the referents of
at least a majority of the symbols employed have been
reached, are known.

In this chapter we are, for the sake of simplicity,
confining our attention to reference alone. In actual
discussion terms are used at least as much for the sake
of their suasory and emotive effects as for their strictly
symbolic value. Any substitute for ‘beautiful,’ for
example, inevitably falls so flatly and heavily that
123many people prefer to use the term with all its dangers
rather than the psychological jargon which they may
agree is more satisfactory from a scientific as opposed
to an emotive point of view.

It is often, indeed, impossible to decide, whether a
particular use of symbols is primarily symbolic or
emotive. This is especially the case with certain kinds
of metaphor. When the Psalmist cries of his enemies,
“They have sharpened their tongues like a serpent;
adders' poison is under their lips,” it is hard to
determine whether an elusive similarity between the
reptile and the persons he is describing is enabling him
metaphorically to state something about them, or
whether the sole function of his utterance is not to
express his abhorrence of them and to promote similar
attitudes towards them in his hearers. Most terms of
abuse and endearment raise this problem, which, as a
rule, it is, fortunately, not important to settle. The
distinction which is important is that between utterances
in which the symbolic function is subordinate to the
emotive act and those of which the reverse is true. In
the first case, however precise and however elaborate
the references communicated may be, they can be seen
to be present in an essentially instrumental capacity,
as means to emotive effects. In the second case, however
strong the emotive effects, these can be seen to
be by-products not essentially involved in the speech
transaction. The peculiarity of scientific statement,
that recent new development of linguistic activity, is
its restriction to the symbolic function.

If this restriction is to be maintained, and if scientific
methods of statement are to be extended to fields such
as those traditionally tended by philosophers, certain
very subtle dangers must be provided for. Amongst
these is the occurrence, in hitherto quite unsuspected
numbers, of words which have been erroneously
regarded without question as symbolic in function.
The word ‘good’ may be taken as an example. It
124seems probable that this word is essentially a collection
of homonyms, such that the set of things, roughly,
those in connection with which we heard it pronounced
in early years (a good bed, a good kick, a good baby,
a good God) have no common characteristic. But
another use of the word is often asserted to occur, of
which some at least of those which we have cited are
supposed to be degenerations, where ‘good’ is alleged
to stand for a unique, unanalysable concept. This
concept, it is said, is the subject-matter of Ethics. 1144
This peculiar ethical use of ‘good’ is, we suggest, a
purely emotive use. When so used the word stands
for nothing whatever, and has no symbolic function.
Thus, when we so use it in the sentence, ‘This is good,’
we merely refer to this, and the addition of ‘is good’
makes no difference whatever to our reference. When
on the other hand, we say ‘This is red,’ the addition
of ‘is red’ to ‘this’ does symbolize an extension of our
reference, namely, to some other red thing. But ‘is
good’ has no comparable symbolic function; it serves
only as an emotive sign expressing our attitude to this,
and perhaps evoking similar attitudes in other persons,
or inciting them to actions of one kind or another.

The recognition that many of the most popular
subjects of discussion are infested with symbolically
blank but emotively active words of this kind is a
necessary preliminary to the extension of scientific
method to these questions. Another is some technique
by which to ascertain which words are of this nature
and on what occasions. Whether experimental and
physiological methods can at present yield any result
may be doubted, but the ultimate settlement of the
matter can hardly be expected until tests in some
125way independent of the opinion of the speaker are
obtained.

In all discussions we shall find that what is said is
only in part determined by the things to which the
speaker is referring. Often without a clear consciousness
of the fact, people have preoccupations which
determine their use of words. Unless we are aware of
their purposes and interests at the moment, we shall
not know what they are talking about and whether their
referents are the same as ours or not.

Purpose affects vocabulary in two ways. Sometimes
without affecting reference it dictates the choice of
symbols specially suited to an occasion. Thus, the
language of a teacher in describing his spectroscope to
a child may differ from that in which he describes it to
his colleague or to his fiancée without there being any
difference in his reference. Or an elegant writer will
ring the changes on a series of synonyms 1145 without
changing his reference. On the other hand, a physicist
uses different language from that employed by his guide
in order to discuss the Spectre of the Brocken; their
different purposes affect their language in this case
through altering their references.

It is plain that cases of the first kind are much
simpler than those of the second; only the latter are
likely to lead to vain controversies. Thus, if one disputant
talks of public opinion he may be referring to
what others would call the views of certain newspaper
owners, in which case an argument as to whether the
Press influences public opinion would tend to be inconclusive
in the absence of some third party familiar with
the technique of definition. Such arguments are of
constant occurrence even in the most intelligent circles,
although when examined in the clear light of criticism
they usually appear too foolish to be possible.

But how should a discussion whose aim is the
126removal of uncertainty as to whether the parties to it
are referring to the same things or not be conducted?

The first necessity is to remember that since the
past histories of individuals differ except in certain very
simple respects, it is probable that their reactions to
and employment of any general word will vary. There
will be some to whom a word is merely a stimulus to
the utterance of other words without the occurrence of
any reference — the psittacists, that is to say, who
respond to words, much as they might respond to the
first notes of a tune which they proceed almost automatically
to complete. At the other extreme there will
be some for whom every word used symbolizes a definite
and completely articulated reference. With the first we
are not here concerned, but as regards the others, unless
we have good evidence to the contrary we should assume
that, clear though their ideas may be, they will probably
not be ideas of the same things. It is plain that we can
only identify referents through the references made to
them. Different references then, may be to the same
referent, sufficiently similar ones must be; and it is
only by ensuring similarity of reference that we can
secure identity in our referents. For this it is desirable
to symbolize references by means of the simple routes
of definition discussed above. We must choose as
starting-points either things to which we can point, or
which occur freely in ordinary experience. The routes
by which we link these starting-points to our desired
referents must be thoroughly familiar, which in practice
confines us to four main routes and combinations of
these. They are those which we must know and
unerringly recognize if we are to survive — Similarity,
Causation, Space and Time. In practice, however, it
is often sufficient to start from less primitive initial
points and follow more complicated and dangerous
routes. Thus ‘razor’ = ‘instrument used for shaving’
unambiguously, without it being necessary to reduce
‘used for’ any further by analysis.127

At what point our definitions are thorough enough
must be left for the occasion to decide. In viva voce
discussion, unless unduly prolonged and pertinacious,
little can be hoped for except stimulus and hints which
will be of use in more serious endeavours. But where
there is reason to suppose that a slippery term is being
employed, it is a wise policy to collect as wide a range
of uses as possible without at this stage seeking for a
common element. A good dictionary attempts this for
certain words, but usually from an historical standpoint
and on no theoretical principle. The next step is to
order these uses with a view to discovering which main
routes of identification have been adopted for the
referents concerned. It is not necessary that the
separate definitions so formulated should be mutually
exclusive; very often they will cover the same referents
but with different references. In such cases we may
be confronted by the problem of levels of reference above
alluded to. ‘Animal’ in current speech, and ‘mammal’
in zoology stand for almost the same referents; but the
references differ very greatly in the definiteness and
complexity of the sign-chains involved. These differences
should, if possible, be indicated in the formulation
of the definitions. What is required is that each definition
should unmistakably mark out a certain range
of referents. If two definitions mark out the same range
no harm is done, the essential consideration being that
each range should be clearly separated from the others
so as to be capable of treatment on its own merits.

The natural tendency of those accustomed to traditional
procedure is to expect that since what appears
to be one word is being defined, the alternative substitute
symbols will stand for referents with some common
character of a more or less recondite nature. This may
sometimes occur, but the inquiry as to whether there
is such a common character should be postponed to a
much later stage. The slightest study of the way in
which words in ordinary speech gain occasional derivative
128and supernumerary uses through metaphorical
shifts of all degrees of subtlety, and through what can
be called linguistic accidents, is enough to show that
for a common element of any interest or importance to
run through all the respectable uses of a word is most
unlikely. Each single metaphorical shift does, of
course, depend upon some common element which is
shared by the original reference and by the reference
which borrows the symbol. Some part of the two
contexts of the references must be the same. But the
possible overlaps between contexts are innumerable,
and there is no reason to expect that any word at all
rich in context will always be borrowed on the strength
of the same similarity or overlap. Thus, Beautiful^A
and Beautiful^B may symbolize references with something
in common; so may Beautiful^B and Beautiful^c, but it
by no means follows that these common elements will be
the same or that the three symbols will stand for referents
which share anything whatever of interest. Yet few
writers who concern themselves with such wandering
words resist the temptation to begin their inquiry with
a search for essential or irreducible meanings.

The whole of this development took place in Latin,
but when in English the word was borrowed in the
form persoun, which Chaucer uses, a transference and
fading out of the metaphor in B produces Bi, the shift
to ‘personage’; and parson is a phonetic spelling of this
older form. In this manner about a dozen uses of a
word may often be found; and where the historical or
phonetic separation is not clearly defined confusion
is inevitable unless the objects referred to are so readily
distinguishable as to encourage the punster.

If we wish to mediate between rival views it is far
better to assume that the disputants are terminologically
independent than to assume that they must in all respects
use their words alike. With the first procedure, if there
actually is a common element involved, we shall be in
a good position to discover it. With the second we
shall inevitably tend to misrepresent all the views concerned
and to overlook most of their really valuable and
peculiar features. The synthesis of diverse opinions,
if it is attempted at all, should be postponed until each
view has been examined as completely as possible in
isolation. Premature efforts, to which all our natural
attitudes to symbols conspire to tempt us, are an
unfailing source of confusion.

For those whose approach to symbols is unreflective
it is often difficult to believe that such convenient words
as ‘beauty,’ ‘meaning,’ or ‘truth’ are actually not
single words at all, but sets of superficially indistinguishable
yet utterly discrepant symbols. The reasons
why this is so are, however, not hard to point out.
Language, which has developed chiefly to satisfy the
exigencies of everyday practical intercourse, presents
a remarkable unevenness in the density of distribution
of its units when we regard it from the standpoint of
our theoretical needs. Thus it constantly happens
that one word has to serve functions for which a hundred
would not be too many. Why language is often so
recalcitrant to growth at these points is a puzzling
130problem. Shortage of terms in the established sciences
is met without difficulty by the introduction of new
terms. But with sciences in their initial stages, before
they have developed into affairs for specialists, and
while they are still public concerns, the resistance to
new terms is very great. Probably the explanation
of this is to be found in the lack of emotive power which
is a peculiarity of all technicalities.

The result of this scarcity of terms is that any
reference whatever made to these symbolically starved
topics is forced to make use of the few words which are
available, no matter how distinct its referents may be
from those of other references which also use the same
words. Thus any reference to human activities which
are neither theoretical nor practical tends to be symbolized
by the word ‘æsthetic’; and derivatively anything
which we are not merely concerned either to
know or to change tends to be described as beautiful.
And this, no matter how many fundamentally different
attitudes to things we may come to distinguish. We
have here a cause for the extravagant ambiguity of all
the more important words used in general discussion;
one which supplements and reinforces the processes of
metaphorical shift just considered.

At the beginning, then, of any serious examination
of these subjects we should provide ourselves with as
complete a list as possible of different uses of the
principal words. The reason for making this list as
complete as possible, subject, of course, to common
sense and ordinary discretion, is important. It is
extraordinarily difficult in such fields to retain consistently
what may be called a ‘sense of position.’
The process of investigation consists very largely of
what, to the investigator, appear to be flashes of
insight, sudden glimpses of connections between
things and sudden awareness of distinctions and differences.
These, in order to be retained, have to be
symbolized, if, indeed, they do not, as is most often
131the case, originally occur in an already symbolized
state.

Without such a map of the separable fields covered
by the investigation any constatation géniale is liable to be
confused with another, to their common detriment, or
to yield an apparent contradiction of purely verbal
origin. If, however, we are able at once to locate the
idea in its proper province, the accident that we have
to use the same words as totally distinct symbols is
deprived of its power to disturb our orientation. The
mere ad hoc distinction between two or perhaps three
senses of a word made in response to particular exigencies
of controversy is insufficient. We can never
foretell on what part of the total field light will next be
vouchsafed, and unless we know in outline what the
possibilities are we are likely to remain ignorant of
what it is into which we have had insight.

Not all words are worth so much trouble. It might
be supposed that it is rather certain subjects which do
not merit attention, but closer scrutiny suggests that
these subjects, of which Theology appears to be a good
example, are themselves merely word systems. But
even the most barren fields have their psychological
interest, and those who approach a discussion armed
with a symbolic technique and able to apply such
principles as the Canons dealt with in the last chapter
may hope every day and in every way to find themselves
better and better.

Something, however, can be achieved even by those
who shrink from the severities of the Six Canons. In
his Art of Controversy, of which he remarked “I am
not aware that anything has been done in this direction
although I have made inquiries far and wide,”
Schopenhauer says, “It would be a very good thing
if every trick could receive some short and obviously
appropriate name, so that when a man used this or
that particular trick, he could be at once reproached for
it.” This suggestion is supported by Professor Dewey's
132characterization of the verbal sign as a fence; a label;
and a vehicle: that is to say it selects and detaches
meanings from out of the void, and makes what was
dim and vague stand out as a clear-cut entity —
secondly, it conserves the meaning thus fixed for
future use, and, thirdly, enables it to be applied and
transported to a new context and a new situation. Or
in less metaphysical language, a symbol assists us in
separating one reference from another, in repeating a
reference we have already made, and in making partially
analogous references in other contexts. In all these ways
a notation of the devices of the controversialist would
be very desirable.

Three such tricks may thus be readily stigmatized.
The first, the Phonetic subterfuge, would be considered
too simple to be dangerous if history bore no testimony
to its effects. It consists in treating words which sound
alike as though their expansions must be analogous.
The most famous case is Mill's use of ‘desirable’ as
though it must expand in the same way as ‘visible’ or
‘knowable.’ The subterfuge is to be charged against
language rather than against Mill, and is plainly verbal.
‘Desirable,’ in the sense equivalent to ‘ought to be
desired,’ may be reducible to ‘can be desired by a mind
of a certain organization,’ 1147 but is not on all fours as a
symbol with ‘visible’ in the sense of ‘able to be seen
by somebody.’

The second subterfuge, the Hypostatic, is more
difficult to discourage because it is a misuse of an
indispensable linguistic convenience. We must, if
we are ever to finish making any general remark,
contract and condense our language, but we need not
hypostatize our contractions. The point has been
referred to in connection with Universals, but how
popular and how influential is this practice may be
133shown by such a list of terms as the following: — Virtue,
Liberty, Democracy, Peace, Germany, Religion, Glory.
All invaluable words, indispensable even, but able
to confuse the clearest issues, unless controlled by
Canon III.

The third, the Utraquistic subterfuge, has probably
made more bad argument plausible than any other
controversial device which can be practised upon
trustful humanity. It has long been recognized that
the term ‘perception’ may have either a physical or a
mental referent. Does it refer to what is perceived, or
to the perceiving of this? Similarly, ‘knowledge’ may
refer to what is known or to the knowing of it. The
Utraquistic subterfuge consists in the use of such terms
for both at once of the diverse referents in question.
We have it typically when the term ‘beauty’ is employed,
reference being made confusedly both to
qualities of the beautiful object and to emotional effects
of these qualities on the beholder.

Sometimes two or more of these subterfuges may be
located in the same word. Thus ‘Beauty’ on most
occasions is a double offender, both hypostatic and
utraquistic.

In addition to this labelling of controversial tricks,
a further set of Rules of Thumb may be laid down for
practical guidance in conformity with the six Canons.
In a recent Symposium of the Aristotelian Society on
Mental Activity, carried on for the most part in inverted
commas, it was not surprising to find Professor
Carveth Read remarking once more that “the commonest
cause of misunderstanding has long been recognized
to lie in the ambiguity of terms, and yet we make
very little progress in agreeing upon definitions. Even
if we sometimes seem to be agreed upon the use of an
important word, presently a new interest awakens or an
old interest acquires new life; and then, if its adherents
think it would be strengthened by using that word in
another sense they make no scruple about altering it.” 134

Over two years later at the tenth annual meeting
of the American Philosophical Association we find
Professor Lovejoy breaking in on a similar series of
misunderstandings with the remark, “More adherence
to definitions is required if we are to come to an
understanding. Appoint a committee to define the
fundamental terms which are to be used in the discussion.”

When we consider the amount of time we spend
to-day in such discussion and the number of words we
utter in the course of a single day — it is calculated
that when vocal we emit between 150 and 250 words
per minute — it is of some importance to recognize
certain classes of these words which are liable to mislead
in controversy.

“In Psychology what seems ‘is’” it has been
happily said. Is what ‘seems’ Real? “Everything,”
replies Bosanquet, “is Real so long as we do not take
it for what it is not.” “I somewhat uncautiously speak
of mind as a Thing,” confessed Professor Alexander —
and still more regretfully “I have used the unfortunate
word Phenomenon. I have made up my mind that I
shall never use the word Phenomenon again without
carefully defining its meaning. How Mr Stout can
say I describe the mind as if it were not a Phenomenon
passes my comprehension. I meant by the word
Almost Nothing at all.” This is reminiscent of Croce's
dictum with regard to the Sublime: “the Sublime is
everything that is or will be so called by those who
have employed or shall employ the name.” The chief
function of such terms in general discussion is to act
as Irritants, evoking emotions irrelevant to the determination
of the referent. This is an abuse of the
poetical function of language to which we shall return.

There is much scope for what may be called the
Eugenics of Language, no less than for the Ethics of
Terminology.

Foreshadowing the conscious process of Linguistic
135elimination Mr Alfred Sidgwick has drawn attention
under the title “Spoilt Words” to terms ambiguous
beyond remedy. But having thus stated the problem,
he leaves it. Language, as we know, was made before
people learned to think: in the phraseology of
Mill, by the ‘vulgar’; and it is still being so made in
the form in which we use it in conversation, however
much we may regret the fact. It is very questionable
how far we do but add to the existing confusion by
endeavouring to restrict the meaning of these Unfortunates.
When we remember that it is not round words
only that emotional and other associations gather,
but that Victor Hugo, for instance (as Ribot has
pointed out), saw in each letter, even, a symbolic representation
of some essential aspect of human knowledge, 1148
it is somewhat optimistic to put trust in the
efficacy of restriction of meaning in discussion. “I believe,”
said Max Müller, “that it would really be of the
greatest benefit to mental science if all such terms as
impressions, sensations — soul, spirit, and the rest,
could, for a time, be banished, and not be readmitted
till they had undergone a thorough purification.”
And in his remarkable analysis of the Economics of
Fatigue and Unrest (1924) Dr Sargant Florence has
successfully employed this method by eliminating
altogether the terms ‘fatigue’ and ‘unrest’ in the earlier
stages (Chapters V.-XI.) of his argument.

“Never change native names, for there are Names
in every nation God-given, of unexplained power in
the mysteries.” So says a Chaldean Oracle with true
insight. But in prose discussions which aim at the
avoidance of mysteries, both Irritants and Degenerates
must be ruthlessly rejected — Irritants because of their
power to evoke disturbing emotions, and Degenerates
because of the multiplicity of their associated referents.136

It is not necessary here to compile the Index Expurgatorius
from ‘Appearance’ to ‘Reality,’ or as
near Z as possible.

There is another class of words which may profitably
be placed beyond the range of legitimate dispute.
Matthew Arnold speaks of “terms thrown out, so to
speak, at a not fully grasped object of the speaker's
consciousness.” So long as the true function of these
Mendicants, as they might be designated, is recognized,
they will cause little trouble. They must never
receive harsh treatment; decasualization is the remedy.

To be distinguished from Mendicants, which may be
assumed to possess the homing instinct, are Nomads,
whose mode of life was first described by Locke.

“Men having been accustomed from their cradle to learn
words which are easily got and retained, before they knew or had
framed the complete ideas which they express, they usually continue
to do so all their lives; and without taking the pains
necessary to settle in their minds determined ideas, they use their
words for such unsteady and confused notions as they have;
contenting themselves with the same words as other people use,
as if the very sound necessarily carried with it the same meaning.
This (although men make a shift with it in ordinary occurrences
of life, yet when they come to reason concerning their Tenets) it
manifestly fills their discourse with abundance of empty noise
and jargon — especially in moral matters where the bare sound
of the words are often only thought on, or at least very uncertain
and obscure notions annexed to them.

Men take the words they find in use amongst their neighbours,
and that they may not seem ignorant what they stand for use
them confidently without much troubling their heads about a
certain fixed meaning, whereby besides the ease of it they obtain
this advantage that as in such discourse they are seldom in the
right so they are seldom to be convinced they are in the wrong,
it being all one to draw these men out of their mistakes, who have
no settled notions, as to dispossess a Vagrant of his habitation,
who has no settled abode. This I guess to be so; and every one
may observe in himself or others whether it be so or not.”

We can still agree to-day that there is little doubt
as to whether it be so or not; and if we were able more
readily to recognize these Nomads, we should spend
137less time in the frenzied rifling of Cenotaphs which is
at present so much in favour.

When we enter the Enchanted Wood of Words, our
Rules of Thumb may enable us to deal not only with
such evil genii as the Phonetic, the Hypostatic and the
Utraquistic subterfuges, but also with other disturbing
apparitions of which Irritants, Mendicants and Nomads
are examples; such Rules, however, derive their virtue
from the more refined Canons, whose powers we have
already indicated.

It may, however, be asked, What is the use of
knowing the nature of definition, for does not the
difficulty consist in hitting upon the precise definition
which would be useful? There are two answers to this.
In the first place, the ability to frame definitions comes
for most people only with practice, like surgery, diagnosis
or cookery, but, as in these arts, a knowledge
of principles is of great assistance. Secondly, such a
knowledge of general principles renders any skill
acquired in the course of special study of one field
available at once when we come to deal with other
but similar fields. In all the main topics of discussion
— Æsthetics, Ethics, Religion, Politics, Economics,
Psychology, Sociology, History — the same types of
defining relations occur, and thus a theoretical mastery
of any one of them gives confidence in the attack upon
the others.138

Chapter VII
The meaning of beauty

This I have here mentioned by the bye to show of what
Consequence it is for Men to define their Words when
there is Occasion. And it must be a great want of Ingenuity
(to say no more of it) to refuse to do it: Since a
Definition is the only way, whereby the precise Meaning
of moral Words can be known. — Locke.

“Disputes are multiplied, as if everything was uncertain,
and these disputes are managed with the greatest
warmth, as if everything was certain. Amidst all this
bustle 'tis not reason which gains the prize, but eloquence;
and no man need ever despair of gaining proselytes
to the most extravagant hypothesis, who has art
enough to represent it in any favourable colours. The
victory is not gained by the men at arms, who manage the
pike and sword; but by the trumpeters, drummers, and
musicians of the army.” — Hume.

In order to test the value of the account of Definition
given in the previous chapter, we may best select a
subject which has hitherto proved notoriously refractory
to definitive methods. Many intelligent people indeed
have given up æsthetic speculation and take no interest
in discussions about the nature or object of Art, because
they feel that there is little likelihood of arriving at
any definite conclusion. Authorities appear to differ
so widely in their judgments as to which things are
beautiful, and when they do agree there is no means
of knowing what they are agreeing about.

What reason is there to suppose that one æsthetic
doctrine can be framed to include all the valuable kinds
of what is called Literature.

Yet, surprising though it may seem, the only author
who appears to have expressly admitted this difficulty
and recognized its importance is Rupert Brooke. “One
of the perils attending on those who ask ‘What is Art?’
is,” he says, “that they tend, as all men do, to find what
they are looking for: a common quality in Art…
People who start in this way are apt to be a most
intolerable nuisance both to critics and to artists…
Of the wrong ways of approaching the subject of ‘Art,’
or even of any one art, this is the worst because it is
the most harmful.” He proceeds to point out how
“Croce rather naively begins by noting that ‘æsthetic’
has been used both for questions of Art and for perception.
So he sets out to discover what meaning it
can really have to apply to both. He takes it for the
140one necessary condition a true answer about ‘Æsthetics’
must satisfy, that it shall explain how Art and Perception
are both included. Having found such an explanation,
he is satisfied.” The same lively awareness
of linguistic pitfalls which enabled Rupert Brooke
wisely to neglect Croce also allowed him to detect the
chink in Professor G. E. Moore's panoply, and so to
resist the inexorable logic of the Cambridge Realists,
then at the height of their power. “Psychologically,”
he says, “they seem to me non-starters. In the first
place I do not admit the claims of anyone who says
‘There is such a thing as Beauty, because when a
man says, “This is beautiful,” he does not mean “This
is lovely.”’… I am not concerned with what men
may mean. They frequently mean, and have meant,
the most astounding things. It is, possibly, true that
when men say, ‘This is beautiful’ they do not mean
‘This is lovely.’ They may mean that the æsthetic
emotion exists. My only comments are that it does
not follow that the æsthetic emotion does exist; and
that, as a matter of fact, they are wrong.” 1149

His own sympathies, at least as they appear in the
volume from which we quote, were with views of type
XI. in the list given below, though he does not seem
to have considered the matter very deeply, and had no
opportunity of following up the promise of his admirable
approach.

Whenever we have any experience which might be
called ‘æsthetic,’ that is whenever we are enjoying,
contemplating, admiring or appreciating an object,
there are plainly different parts of the situation on
which emphasis can be laid. As we select one or other
of these so we shall develop one or other of the main
æsthetic doctrines. In this choice we shall, in fact, be
141deciding which of the main Types of Definition we are
employing. Thus we may begin with the object itself;
or with other things such as Nature, Genius, Perfection,
The Ideal, or Truth, to which it is related; or with its
effects upon us. We may begin where we please, the
important thing being that we should know and make
clear which of these approaches it is that we are taking,
for the objects with which we come to deal, the referents
to which we refer, if we enter one field will not as a rule
be the same as those in another. Few persons will be
equally interested in all, but some acquaintance with
them will at least make the interests of other people
more intelligible, and discussion more profitable.
Differences of opinion and differences of interest in these
matters are closely interconnected, but any attempt at a
general synthesis, premature perhaps at present, must
begin by disentangling them.

We have then to make plain the method of Definition
which we are employing. The range of useful
methods is shown in the following table of definitions,
most of which represent traditional doctrines, while
others, not before emphasized, render the treatment
approximately complete. It should be remarked that
the uses of ‘beautiful’ here tabulated are not by any
means fully stated. Any definition is sufficiently explicit
if it enables an intelligent reader to identify the reference
concerned. A full formulation in each of these cases
would occupy much space and would show that the
field of the beautiful is for some of them more extensive
than that of works of art, while certain restrictions,
such as those which would exclude the Police from
No. VIII., for example, will readily occur to the
reader.

It will be noticed that each of these definitions
illustrates one or more of the fundamental defining
relations discussed in the last chapter. Thus, the
definitions in Group C, Definitions X.-XVI., are all
in terms of the effects of things upon consciousness
and so are cases of type VII. Of the two definitions in
Group A, the first is a case of simple naming, type I.
We postulate a quality Beauty, name it, and trust the
identification of this mythological referent to the
magical efficacy of our name. The discussion of the
143Beautiful in terms of an intrinsic quality Beauty is in
fact an excellent example of the survival of primitive
word-superstitions, and of the risks run by any discussion
which is symbolically uncritical. The second
Definition (II.), by Form, is either Spatial or Temporal
according to the Art to which it is applied. If any
others than these relations seem to be involved on any
occasion, we shall find on examination that the definition
has had its starting-point surreptitiously changed
and has become actually psychological, a change which
can easily occur in this field, without any immediately
apparent change in the symbolism. As a glaring
instance the use of the word ‘great’ in literary and
artistic criticism shows this process, the transition,
without symbolic indication, from the ‘objective’ to the
‘subjective’ as they used to be called.

The Definitions in Group B are all more or less
complex.

Both Imitation (III.), and Exploitation (IV.), the
definition by reference to the capacities of the medium,
are evidently compounded of Causation, Similarity,
Cognizing and Willing Relations; Exploitation being
in fact as fine an instance as can be found of a complex
definition easy to understand in its condensed shorthand
form and difficult or impossible to analyse. Few
people, however, will suffer any temptation to postulate
a special property of being an exploitation, though
such devices are the penalty we usually have to pay for
convenient short cuts in our symbolization.

The other definitions of Group B offer similar
problems in analysis. The degree to which routes of
type VIII., mental attitudes of believing (VI. and VII.)
or approving (VIII.), appear is an interesting feature,
which again helps to account for the tendency of such
views to become psychological (Group C). Thus
definition XVI. tends to absorb and replace VI.; and
XV. in a refined and explicit form often supersedes V.
These variations in reference, even for definitions of
144symbols specially provided to control such inconstancy,
serve to remind us of the paramount importance of
Canon IV. for all discussion. The use of a symbolic
theory of definition lies not in any guarantee which it
can offer against ambiguity, but in the insight which it
can give as to what, since we are using symbols, will
be happening; and in the means provided of detecting
and correcting those involuntary wanderings of the
reference which are certain in all discourse to occur.

In the case of the above definitions our ‘starting-points,’
synæsthesis, specific emotion, desirable social
effects, etc., are plainly themselves arrived at by
intricate processes of definition. For the particular
purposes for which definitions of ‘beautiful’ are likely
to be drawn up these starting-points can be assumed to
be agreed upon, and the methods by which such
agreement can be secured are the same for ‘emotion’
or ‘pleasure,’ as for ‘beautiful’ itself.

Equally we can proceed from these definitions or
from any one of them, to terms cognate (Ugliness,
Prettiness, Sublimity) or otherwise related (Art,
Æsthetic Decoration), and to define these in their turn
we may take as starting-points either some one of the
now demarcated fields of the beautiful and say: —
Æsthetics is the study of the Beautiful, or: — Art is
the professed attempt to produce Beauty, or we may
return to our starting-point for the definition of Beauty
and box the compass about it.

The fields indicated by the above definitions may in
some cases be co-extensive, e.g., V. and XV.; or they
may partially overlap, e.g., X. and XIII.; or they may
be mutually exclusive, a condition not realized here
or indeed in any probable discussion. The question
whether two such fields do co-extend, do overlap or do
exclude, is one to be decided by detailed investigation
of the referents included in the fields. The ranges of
overlap between fields, in fact, give rise to the special
empirical problems of the sciences. Thus, for instance,
145we find that beautiful things defined as Imitations of
Nature (III.) only coincide with beautiful things defined
as producers of Illusion (VII.) under certain strict
conditions among which is to be found the condition
that neither shall be included in the range defined by
IV. The investigation of such correlations and the
conditions to which they are subject is the business of
Æsthetic as a science.

The advantage of a grammatically extensional form
for the definitions is that, so put, the symbols we use
are least likely to obscure the issues raised, by making
questions which are about matters of fact into puzzling
conundra concerning the interlinking of locutions.

The fields reached by these various approaches can
all be cultivated and most of them are associated with
well-known names in the Philosophy of Art.

Let us, then, suppose that we have selected one
of these fields and cultivated it to the best of our
ability; for what reasons was it selected rather than
some other? For if we approach the subject in the
spirit of a visitor to the Zoo, who, knowing that all
the creatures in a certain enclosure are ‘reptiles,’ seeks
for the common property which distinguishes them as
a group from the fish in the Aquarium, mistakes may
be made. We enter, for example, Burlington House,
and, assuming that all the objects there collected are
beautiful, attempt similarly to establish some common
property. A little consideration of how they came there
might have raised serious doubts; but if, after the
manner of many æstheticians, we persist, we may even
make our discovery of some relevant common property
appear plausible.

We have seen (pp. 124-5) how widely such a respected
word as ‘good’ may wander; and there are good
reasons for supposing that ‘beauty’ will not be more
faithful to one particular kernel of reference. In discussion
we must in fact always bear in mind that there
is an indefinitely large number of ways in which any
146symbol may acquire derivative uses; any similarity,
any analogy may provide a sufficient reason for an
extension of ‘meaning,’ or semantic shift. It no more
follows that the two or more symbols which it then
becomes (cf. p. 91) will stand for referents with some
relevant common property, than it would follow from
the common name of a man's step-mother and his
daughter-in-law that they share his gout or his passion
for the turf.

If, therefore, terms such as Beauty are used in discussion
for the sake of their emotive value, as is usually
the case, confusion will inevitably result unless it is
constantly realized that words so used are indefinable,
i.e., admit of no substitution, there being no other
equally effective stimulus-word. Such indefinable uses
are no doubt what have often led to the assumption of
a simple quality of Beauty (Definition I.) to account for
verbal difficulties; as was also suggested above in the
case of Good (p. 125). If, on the other hand, the term
Beauty be retained as a short-hand substitute, for some
one among the many definitions which we have elicited,
this practice can only be justified as a means of indicating
by a Word of Power that the experience selected
is regarded as of outstanding importance; or as a
useful low-level shorthand.

In addition to providing a test case for any general
technique of definition a consideration of the problem
of Beauty is perhaps the best introduction to the question
of the diverse functions of language. As is well
known, those whose concern with the arts is most direct
often tend to deprecate a scientific approach as being
likely to impair appreciation. This opinion if carefully
examined will be found to be a typical symptom of
a confusion as to the uses of language so constantly
present in all discussions that its general recognition
would be one of the most important results which a
science of symbolism could yield.

If we compare a body of criticism relating to any
147of the arts with an equally accredited body of remarks
dealing with, let us say, physics or physiology, we
shall be struck by the frequency, even in the best
critics, of sentences which it is impossible to understand
in the same way as we endeavour to understand those
of physiologists. “Beautiful words are the very and
peculiar light of the mind,” said Longinus. According
to Coleridge “the artist must imitate that which is
within the thing, that which is active through form
and figure, and discourses to us by symbols — the Naturgeist,
or spirit of nature.” “Poetry,” Dr Bradley
writes, “is a spirit. It comes we know not whence.
It will not speak at our bidding, nor answer in our
language. It is not our servant; it is our master.” 1151
And Dr Mackail is even more rhapsodic: “Essentially
a continuous substance or energy, poetry is historically
a connected movement, a series of successive integral
manifestations. Each poet, from Homer to our own
day, has been to some extent and at some point, the
voice of the movement and energy of poetry; in him
poetry has for the moment become visible, audible,
incarnate, and his extant poems are the record left of
that partial and transitory incarnation… The
progress of poetry… is immortal.” 2152

No one who was not resolved to waste his time
would for long try to interpret these remarks in the
same way as he would, let us say, an account of the
circulation of the blood. And yet it would be a mistake
to regard them as not worth attention. It is clear that
they require a different mode of approach. Whether
their authors were aware of the fact or not, the use of
words of which these are examples is totally distinct
from the scientific use. The point would be made still
more plain, if sentences from poetry were used for the
experiment. What is certain is that there is a common
and important use of words which is different from the
148scientific or, as we shall call it, the strict symbolic use
of words.

In ordinary everyday speech each phrase has not
one but a number of functions. We shall in our final
chapter classify these under five headings; but here a
twofold division is more convenient, the division
between the symbolic use of words and the emotive use.
The symbolic use of words is statement; the recording,
the support, the organization and the communication of
references. The emotive use of words is a more simple
matter, it is the use of words to express or excite feelings
and attitudes. It is probably more primitive. If
we say “The height of the Eiffel Tower is 900 feet” we
are making a statement, we are using symbols in order
to record or communicate a reference, and our symbol
is true or false in a strict sense and is theoretically
verifiable. But if we say “Hurrah!” or “Poetry is
a spirit” or “Man is a worm,” we may not be making
statements, not even false statements; we are most
probably using words merely to evoke certain attitudes.

Each of these contrasted functions has, it will be
seen, two sides, that of the speaker and that of the
listener. Under the symbolic function are included
both the symbolization of reference and its communication
to the listener, i.e., the causing in the listener of
a similar reference. Under the emotive function are
included both the expression of emotions, attitudes,
moods, intentions, etc., in the speaker, and their communication,
i.e., their evocation in the listener. As
there is no convenient verb to cover both expression
and evocation, we shall in what follows often use the
term ‘evoke’ to cover both sides of the emotive function,
there being no risk of misunderstanding. In many
cases, moreover, emotive language is used by the
speaker not because he already has an emotion which
he desires to express, but solely because he is seeking
a word which will evoke an emotion which he desires
to have; nor, of course, is it necessary for the speaker
149himself to experience the emotion which he attempts
to evoke.

It is true that some element of reference probably
enters, for all civilized adults 1153 at least, into almost all
use of words, and it is always possible to import a
reference, if it be only a reference to things in general.
The two functions under consideration usually occur
together but none the less they are in principle distinct.
So far as words are used emotively no question as to
their truth in the strict sense can directly arise. Indirectly,
no doubt, truth in this strict sense is often
involved. Very much poetry consist of statements,
symbolic arrangements capable of truth or falsity,
which are used not for the sake of their truth or falsity
but for the sake of the attitudes which their acceptance
will evoke. For this purpose it fortunately happens,
or rather it is part of the poet's business to make it
happen, that the truth or falsity matters not at all to
the acceptance. Provided that the attitude or feeling
is evoked the most important function of such language
is fulfilled, and any symbolic function that the words
may have is instrumental only and subsidiary to the
evocative function.

This subtle interweaving of the two functions is
the main reason why recognition of their difference is
not universal. The best test of whether our use of
words is essentially symbolic or emotive is the
question — “Is this true or false in the ordinary strict
scientific sense?” If this question is relevant then
the use is symbolic, if it is clearly irrelevant then we
have an emotive utterance.

But in applying this test we must beware of two
150dangers. There is a certain type of mind which
although it uses evocative language itself cannot on
reflection admit such a thing, and will regard the
question as relevant upon all occasions. For a larger
body of readers than is generally supposed poetry is
unreadable for this reason. The other danger is more
important. Corresponding in some degree to the strict
sense of true and false for symbolic statements (True^S),
there are senses which apply to emotive utterances
(True^E). Critics often use True^E of works of art, where
alternative symbols would be ‘convincing’ in some
cases, ‘sincere’ in others, ‘beautiful’ in others, and so
on. And this is commonly done without any awareness
that True^E and True^S are different symbols. Further
there is a purely evocative use of True — its use to
excite attitudes of acceptance or admiration; and a
purely evocative use of False — to excite attitudes of
distrust or disapprobation. When so used these
words, since they are evocative, cannot, except by
accident, be replaced by others; a fact which explains
the common reluctance to relinquish their employment
even when the inconvenience of having symbols so
alike superficially as True^S and True^E in use together
is fully recognized. In general that affection for a
word even when it is admitted to be ambiguous,
which is such a common feature of discussion, is very
often due to its emotive efficiency rather than to any
real difficulty in finding alternative symbols which will
support the same reference. It is, however, not always
the sole reason, as we shall see when we come in our
final chapter to consider the condition of word-dependence.

This disparity of function between words as
supports or vehicles of reference and words as expressions
or stimulants of attitudes has, in recent years,
begun to receive some attention, for the most part from
a purely grammatical standpoint. That neglect of
the effects of our linguistic procedure upon all our
151other activities which is so characteristic of linguists
has, however, deprived such studies as have been made
of most of their value. G. von der Gabelentz for
instance, though he declares that “Language serves a
man not only to express something but also to express
himself,” seems in no way to have considered what
extreme consequence this intermingling of functions
has for the theory as well as for the form of language.
And to take the most recent work upon the subject,
Vendryes, in his chapter upon Affective Language,
keeps equally strictly to the grammarian's standpoint.
“The logical element and the affective element,” he
says, “mingle constantly in language. Except for
technical languages, notably the scientific languages,
which are by definition outside life, the expression of
an idea is never exempt from a nuance of sentiment.”
“These sentiments have no interest for the linguist
unless they are expressed by linguistic means. But
they generally remain outside language; they are like
a light vapour which floats above the expression of the
thought without altering its grammatical form,” etc.
The two chief ways in which the affective side of
language concerns the linguist he finds, first in its
effect upon the order of words and secondly as determining
the vocabulary. Many words are dropped or
retained, for affective reasons. “It is by the action of
affectivity that the instability of grammars is to a great
extent to be explained. The logical ideal for a
grammar would be to have an expression for each
function and only one function for each expression.
This ideal supposes for its realization that the language
is fixed like an algebra, where a formula once established
remains without change in all the operations
in which it is used. But phrases are not algebraic
formulæ. Affectivity always envelops and colours the
logical expression of the thought. We never repeat
the same phrase twice; we never use the same word
twice with the same value; there are never two absolutely
152identical linguistic facts. This is due to the
circumstances which ceaselessly modify the conditions
of our affectivity.” 1154

It is perhaps unfair to ask from grammarians some
consideration of the wider aspects of language. They
have their own difficult and laborious subject to occupy
all their attention. Yet from a book the promise of
which was the cause of the abandonment by Couturat
of his projected “Manual of the logic of language”
a more searching inquiry might be expected. It still
remains true that linguists, of whom M. Vendryes is
one of the most distinguished, abound, but investigators
into the theory of language are curiously lacking. 2155

From the philosophical side also, the speculative
approach to this duality of the symbolic and evocative
functions has been made recently under various disguises.
All such terms as Intuition, Intellect, Emotion,
Freedom, Logic, Immediacy, are already famous for
their power to confuse and frustrate discussion.
In general, any term or phrase, ‘élan vital,’ ‘purely
logical analysis’… which is capable of being
used either as a banner 3156 or as a bludgeon, or as
both, needs, if it is to be handled without disaster, a
constant and conscious understanding of these two
functions of language. It is useless to try to sterilize
our instruments without studying the habits of the
bacteria. Not even mathematics is free as a whole
from emotive complications; parts of it seem to be,
but the ease with which mathematicians turn into
mystics (“Even were there no things at all, there
would still be the property of being divisible by 107” )
153when they consider its foundations, shows what the
true situation is.

One of the best known of these disguised discussions
of the emotive function of language centres about the
teaching of Bergson on the nature of knowledge. To
quote from a recent exposition: “The business of
philosophy, according to Bergson, is not to explain
reality, but to know it. For this a different kind of
mental effort is required. Analysis and classification,
instead of increasing our direct knowledge, tend rather
to diminish it.” 1157 As Bergson himself says: “From
the infinitely vast field of our virtual knowledge we
have selected, to turn into actual knowledge, whatever
concerns our action upon things; the rest we have
neglected.” 2158 And as his expositor continues: “The
attitude of mind required for explaining the facts
conflicts with that which is required for knowing them.
From the point of view simply of knowing, the facts
are all equally important and we cannot afford to
discriminate, but for explanation some facts are very
much more important than others. When we want to
explain, therefore, rather than simply to know, we tend
to concentrate our attention upon these practically
important facts and pass over the rest.” 3159

The processes of explanation as described by
Bergson bear a close resemblance to what we have
called reference when this is supported by symbolism.
Owing to his peculiar view of memory, however, he is
unable to make the use of mnemic phenomena which,
as we have seen, is essential if mysticism, even as
regards this kind of ‘knowledge’ is to be avoided.

The other kind of knowledge, ‘virtual knowledge,’
the knowledge which is ‘creative duration,’ the only
kind of knowledge of ‘really real reality’ Bergsonians
will allow, is, as he presents it, unavoidably mystical.
154Not only because any description of it must involve
the expositor in self-contradiction — as we have seen
any repudiation of orthodox symbolic machinery has
this consequence 1160 — but also because it requires an
initial act of faith in the existence of a vast world of
‘virtual knowledge’ which is actually unknown. None
the less, those who have no such faith, and merely
follow the advice of Bergsonians to neglect the actual
terms in the descriptions given and to perform instead
an ‘act of synthesis,’ can easily become persuaded that
they understand what ‘virtual knowledge’ is, and even
that they can possess it.

We have above (p. 81) insisted that knowledge in
the sense of reference is a highly indirect affair, and
hinted that though we often feel an objection to admitting
that our mental contact with the world is neither
close nor full, but on the contrary distant and schematic,
our reluctance might be diminished by a consideration
of our non-cognitive contacts. These, too, are for the
most part indirect, but they are capable of much greater
fullness. The more clear and discriminating reference
becomes, the slighter, relatively to similar but cruder
reference, is our link with what we are referring to —
the more specialized and exquisite the context involved.
With all that Bergson has to say about the tendency
for precise, discriminating, analytic attention to whittle
down our connection with what we are attending to, we
can agree. Bergson, moreover, has well emphasized
the part played by language in reinforcing and exaggerating
this tendency. Thinking casually of conies,
the context involved may be of immense complexity,
since a large part of our past experience with these
animals is operative. Thinking discriminatingly of
the same objects as ‘small deer,’ our context becomes
specialized, and only those features of conies need be
involved which they share with their co-members of the
155class in question. The others need not be lost, but we
can agree that there is a strong tendency for them to
disappear, and in any really difficult feats of discrimination
they will certainly be best omitted.

At the extreme of consciousness most removed from
analytic and abstract attention we have not one but a
variety of possible states, according to the kind and
extent of the contexts, to which the experience in
question belongs. The state may be comparatively
simple, as when we are engaged in some ordinary
perceptual activity, such as throwing dice; or it may
be predominantly emotional; or leaping for our lives
from the onrush of motor cyclists we may again
experience simple throbs of pure unsophisticated experience.
But certain of these concrete, immediate,
unintellectualized phases of life have in their own
right a complexity and richness which no intellectual
activities can equal. Amongst these æsthetic experiences
figure prominently. Many to whom Bergson's
recommendation of immediacy, and his insistence upon
the treasures awaiting those who regain it, make their
appeal will admit that this is because he seems to them
to be describing what happens when they are most
successful in artistic contemplation. We cannot enter
here into the details of what, from the standpoint of
more or less conventional psychology, may be supposed
to happen in these states of synæsthesis. 1161 What,
however, from this standpoint is indisputable is that
the more important of them derive their value from
the peculiar fashion in which impulses formed by and
representing the past experience of the contemplator
are set working.

Thus in a quite precise sense, though one which
can only be somewhat elaborately formulated, the states
of æsthetic contemplation owe their fullness and richness
to the action of memory; not memory narrowed
156down and specialized as is required in reference, but
memory operating in a freer fashion to widen and
amplify sensitiveness. In such conditions we are open
to a more diffused and more heterogeneous stimulation,
because the inhibitions which normally canalize our
responses are removed.

Partly because of certain of the felt characters of
the states we have been describing, a sense of repose
and satisfaction not unlike the repose which follows
a successful intellectual effort, though due to quite
different causes — partly for other reasons, it is not
surprising that these states should have been often
described as states of knowledge. The temptation to
a philosopher when concerned with a subject in which
he feels a passionate interest, to use all the words which
are most likely to attract attention and excite belief in
the importance of the subject is almost irresistible.
Thus, any state of mind in which anyone takes a great
interest is very likely to be called ‘knowledge,’ because
no other word in psychology has such evocative virtue.
If this state of mind is very unlike those usually so
called, the new “knowledge” will be set in opposition to
the old and praised as of a superior, more real, and more
essential nature. These periodic raids upon æsthetics
have been common in the history of philosophy. The
crowning instance of Kant, and the attempted annexation
of æsthetics by Idealism are recent examples.

The suggestion is reasonable, therefore, that when
the pseudo-problems due to cross vocabularies are
removed and the illusory promise of a new heaven
and a new earth, which Bergsonians somewhat weakly
hold out, has been dismissed, the point at issue in the
intuitionist-intellectualist controversy will be found to
be removable by an understanding of the dual function,
symbolic as well as emotive, of the word ‘knowledge.’
To deny that ‘virtual knowledge’ is in the symbolic
sense knowledge is in no way derogatory to the state
(according to the view here maintained, a state, or set
157of states, of specially free response to stimulation)
called by that name. It is merely to apply a rule which
all those who are aware of the functions of language
will support, namely, that in discussion, where
symbolic considerations are supposed to be prior to all
others, the evocative advantages of terms are only to
be exploited when it is certain that symbolically no
disadvantage can result.

But a more general consciousness of the nature of
the two functions is necessary if they are to be kept
from interfering with one another; and especially all
the verbal disguises, by which each at times endeavours
to pass itself off as the other, need to be exposed. It
ought to be impossible to pretend that any scientific
statement can give a more inspiring or a more profound
‘vision of reality’ than another. It can be more
general or more useful, and that is all. On the other
hand it ought to be impossible to talk about poetry
or religion as though they were capable of giving
‘knowledge,’ especially since ‘knowledge’ as a term
has been so overworked from both sides that it is
no longer of much service. A poem — or a religion,
though religions have so definitely exploited the confusion
of function which we are now considering, and
are so dependent upon it, as to be unmistakably pathological
growths — has no concern with limited and
directed reference. It tells us, or should tell us, nothing.
It has a different, though an equally important and a
far more vital function — to use an evocative term in
connection with an evocative matter. What it does,
or should do, is to induce a fitting 1162 attitude to experience.
158But such words as ‘fitting,’ ‘suitable’ or
‘appropriate’ are chilly, having little or no evocative
power. Therefore those who care most for poetry and
who best understand its central and crucial value, tend
to resent such language as unworthy of its subject.
From the evocative standpoint they are justified. But
once the proper separation of these functions is made
it will be plain that the purpose for which such terms
are used, namely to give a strictly symbolic description
of the function of poetry, for many reasons 1163 the
supreme form of emotive language, cannot conflict
with the poetic or evocative appraisal of poetry, with
which poets as poets are concerned.

Further, the exercise of one function need not, if
the functions are not confused, in any way interfere with
the exercise of the other. The sight of persons irritated
with science because they care for poetry (“Whatever
the sun may be, it is certainly not a ball of flaming gas,”
cries D. H. Lawrence), or of scientists totally immune
from the influences of civilization, becomes still more
regrettable when we realize how unnecessary it is.
As science frees itself from the emotional outlook, and
modern physics is becoming something in connection
with which attitudes seem rather de trop, so poetry
seems about to return to the conditions of its greatness,
by abandoning the obsession of knowledge and symbolic
truth. It is not necessary to know what things
are in order to take up fitting attitudes towards them,
and the peculiarity of the greatest attitudes which art
can evoke is their extraordinary width. The description
and ordering of such attitudes is the business of
æsthetics. The evaluation of them, needless to say,
must rest ultimately upon the opinions of those best
qualified to be judges by the range and delicacy of
their experience and their freedom from irrelevant
preoccupations.159

Chapter VIII
The meaning of philosophers

What do you read, my lord? — Polonius.
Words, words, words. — Hamlet.

“O wondrous power of words, by simple faith
Licensed to take the meaning that we love.”

Thus the poet; and observation does not invalidate
the perspicacious remark. It might, however, have
been supposed that logicians and psychologists would
have devoted special attention to meaning, since it is
so vital for all the issues with which they are concerned.
But that this is not the case will be evident 1164 to anyone
who studies the Symposium in Mind (October 1920 and
following numbers) on “The Meaning of ‘Meaning.’”

It is perhaps unnecessary to point out that such
brief extracts from lengthy philosophical disquisitions
as the limits of this chapter allow, cannot fairly represent
any given author's views upon that, whatever it may
be, if anything, for which he uses the word ‘meaning.’
Some quotations, however, do tell their own tale; but
even where no actual absurdity transpires, the resort
160to such a term in serious argument, as though it had
some accepted use, or as though the author's use were
at once obvious, is a practice to be discredited.

Dr Schiller began by announcing that the Greek
language is “so defective that it can hardly be said
to have a vocabulary for the notion” of meaning at all;
and in proceeding to state his own view that “meaning
is essentially personal…. what anything means
depends on who means it,” he found it necessary to
traverse Mr Russell's dictum that “the problem of the
meaning of words is reduced to the problem of the
meaning of images.” Mr Russell replied by endeavouring
“to give more precision to the definition
of meaning by introducing the notion of ‘mnemic
causation’” and succeeded thereby in evolving an
instructive description of metaphysics. “A word,” he
explained, “which aims at complete generality, such
as ‘entity,’ for example, will have to be devoid of
mnemic effects, and therefore of meaning. In practice,
this is not the case: such words have verbal associations,
the learning of which constitutes the study of
metaphysics.” Mr Joachim, who elected to stand aside
from the discussion, professed to find Mr Russell
“asserting that nobody can possibly think” and confined
himself to an analysis of the function of images,
drawing attention in a foot-note to the fact that for
Mr Russell meaning appeared (amongst other things)
as ‘a relation,’ that “a relation ‘constitutes’ meaning,
and that a word not only ‘has’ meaning, but is related
‘to its meaning.’”

This whole episode was characterized by Dr Schiller
six months later (April, 1921, p. 185) as presenting “the
usual features of a philosophic discussion. That is to
say, it reads like a triangular duel, in which each
participant aims at something different, and according
to the other misses it, and hits a phantom.” In dealing
with details he quotes Mr Russell's remark that “all
the words in which Dr Schiller endeavours to describe
161his unobservable entities imply that after all he can
observe them,” as a typical case of “the overriding of
actual meaning by verbal, which could hardly be
surpassed from the writings of Mr Bradley.”

In July Mr Alfred Sidgwick explained (p. 285) that
“meaning depends on consequences, and truth depends
on meaning;” and Professor Strong intervened (p. 313)
as a ‘critical realist’ to meet Dr Schiller's objections
to Mr Russell and to render the latter's theory intelligible
to Mr Joachim. He illustrated his rendering by
imagining an explosion. When we hear what we call
an explosion, “the sound has not so much acquired,
as become converted into a meaning… What is
non-concrete and non-sensuous is always a meaning,
a sense of that unfathomed beyond which we cannot
contemplate but only intend… To mean something
is to conceive or rather treat it as not wholly revealed
to the mind at the moment.”

To this Dr Schiller rejoins that Dr Strong always
confines his attention to the case “in which an ‘object’
is said to ‘mean so-and-so.’” This, he thinks, “imposes
on him the duties of deriving the personal meaning,
and of explaining the relativity of ‘the’ meaning
of an object to various cognitive purposes and personal
meanings” (p. 445). He concludes (p. 447) that “the
existence of personal meaning remains a pitfall in the
path of all intellectualism.” The controversy is presumably
still in progress.

Contemporaneously with the Symposium on Meaning
which appeared in Mind, an inquiry into the nature of
Aphasia was appearing in Brain 1165 and during the discussion
of Dr Head's views the question of meaning
came to the front. A special memorandum suggested
by the treatment of ‘semantic aphasia,’ was handed in
by Dr J. Herbert Parsons, 2166 and it throws interesting
162light on the degree of assistance which neurologists
can be expected to derive from the work of philosophers
in this field. According to Dr Parsons, at the lowest
biological level “it would be unwise to deny the
presence of a plus or minus affective tone — and this is
the primitive germ of ‘meaning.’” At the perceptual
level, however, “the relatively undifferentiated psychoplasm
is differentiated into specialized affective and
cognitive elements, which are reintegrated, thus undergoing
a synthesis which is the ‘meaning’ of the given
experience. Perceptual ‘meaning’ suffused with affective
tone, issues in instinctive conative activity.” Thus
at the end of the completed reaction “the ‘meaning’
has become enriched and complicated… This altered
‘meaning’ is stored up, and, though depressed below
the threshold of consciousness, is capable of being
revived… The integration and synthesis of the
already more plastic psychoplasm results in a higher,
more complex type of ‘meaning.’” Later the influence
of social environment makes itself felt, and in the complicated
process of social intercourse “the ultimate
results are equivalent to an interaction of old and new
‘meanings,’ resulting in an infinity of still newer, richer
and more refined ‘meanings.’” At this stage “the
creative activities assume a synergy at a higher level,”
and “show a projicience hitherto absent.” The child's
“gestures are no longer merely passive signs of his
mind's activities, but active indications of his feelings
and desires. This is the dawn of language.”

A detailed analysis of the Mind Symposium might
have been instructive as a preliminary to the framing
of a set of definitions, but its technique was unusually
disappointing, 1167 and since in any case the metaphysical
arena of the Old World inevitably suggests to many
an atmosphere of barren logomachy, we may more
163profitably deal with the confusions which arise as
occasion allows and cite here the procedure of the
latest co-operative product of the New. The Essays in
Critical Realism, which made their appearance in 1920,
are the work of seven American Professors who have
revised and redrafted their language until it met the
approval of all the other essayists. They are the fruits
of a decade of controversy in a limited controversial
field, where “our familiarity with one another's meaning
has enabled us to understand methods of expression
from which at first we were inclined to dissent.” The
main issues of the controversy had already been elaborated,
as the result of conferences begun in 1908-9, in a
similar co-operative volume by six Neo-realists. The
final outcome may be regarded as the clarification of
the life's work of a dozen specialists, all of whom have
been continuously improving their mutual terminology
in the full view of the public for over a decade.

Since that date, 1912, the word ‘meaning’ has not
ceased to play a decisive part in every dispute, and as
the Critical Realists have had such ample opportunity
of avoiding any ambiguities into which the Neo-Realists
164may have fallen, we may, as far as Realism is
concerned, confine ourselves to their efforts.

These two statements are used to lead up to the view
that perceptual data “cannot be the same existents as
their causes,” and that we “get back somewhere to
qualities.”

It would be a large undertaking, continues Professor
Lovejoy, to “analyse the meanings” of the formulations
of Pragmatism, which “began as a theory concerning
the conditions under which concepts and
propositions may be said to possess meaning, and
concerning the nature of that in which all meanings
must consist.” The pragmatist, he holds, ignores the
patent fact that “many of our meanings are retrospective.
… No logical hocus-pocus can transubstantiate
the meaning ‘yesterday’ into the meaning
‘to-morrow’…. It is, in very truth, a meaning
intrinsically incapable of directly-experienced fulfilment….
Without ever actually experiencing the
fulfilment of these meanings, we nevertheless have an
irresistible propensity to believe that some of them are
in fact valid meanings… A judgment is its own
master in deciding what it means, though not in
deciding as to the fulfilment of its meanings.”

According to Professor Pratt, the Neo-Realists
“performed a most fruitful piece of analysis in insisting
that the data presented to our thought consist of
meanings or natures,” but they did not distinguish
“between these meanings and the sensational part of
our mental states on the one hand and the existential
physical objects to which the meanings are attributed
on the other.” A number of people might describe
their conception of anything differently though all
165“meant, or thought of, the same thing.” He proceeds
to distinguish (p. 90) between the meaning which
one entertains in conception “and images which are
the ‘vehicle’ of our meaning. This meaning is that
which we find directly given to our thought,” and he
holds “this meaning or datum is often capable of
exact definition, i.e., it has, or rather is, a definable
nature.” Perception, equally with conception, “contains
not merely sensuous and revived images but a
large element of meaning as well.” Usually, “All the
sensed qualities are included within those meant.” As
regards outer reference (p. 92) “this may be regarded
as part of the datum or meaning of perception, but it
is an easily distinguishable part.” Thanks to past
reactions, the quality-group “of which one is aware,
directly means more than it is. As a result of all
one's past experience it has come to stand for an
active entity.” This quality-group “means, or immediately
implies to the individual the presence and,
to a considerable extent, the nature of some active
entity of which it is well for him to be aware. It is
in short the means of his perceiving the object.” In conclusion
he maintains that though Critical Realists
“do not pretend to an exhaustive knowledge of the
inner nature of physical entities, we have defined them
sufficiently to know what we mean by them, and to
make that meaning perfectly plain to every one but
the perversely blind.”

Professor Rogers of Yale, who deals with Error,
complains that Bosanquet failed to understand the
question of “degrees of truth” because of his “annoying
refusal to keep sharply separate the varying meanings
of terms. It is not a question whether the same
form of words means the same thing to different people.
It is a question whether any given meaning singly,
whatever it may be, is successful in corresponding to
the fact” (p. 123). Of Mr Joachim's account of things
in terms of systems, he remarks that “If we insist on
166defining the meaning of a fact in terms of its place in
a system, naturally it will cease to have that meaning
outside the system” (p. 125).

As regards identity “we naturally make a clear
distinction between the characters of things as embodied
in meanings which we attribute to them, and
the real existence of these characters in the things
themselves… The ‘identity of indiscernibles’ applies
to abstract logical meanings, not to existents.
Meanings we may call the same — provided we can
detect no difference in them — just because their ‘character’
is all there is to them; but things are not
necessarily the same when they are alike” (p. 131).
Professor Holt's analysis is, he thinks, an “approximately
correct account of what the critical realist intends
to refer to under the head of essences, or human meanings.
But for him the problem of knowledge consists,
not merely in the presence of these meanings or data,
but in their reference to the actual object” (p. 133).
Professor Perry's difficulties as regards error vanish
if we grant the distinction “between the something
as an existent about which I have a belief, and the something
(as an intellectual content or meaning or essence)
which I believe about it.” When in error, we have a
“meaning before the mind,” and wrongly suppose that
it characterizes a real object.

Dr Santayana urges that though without our animal
bodies “appearance would lose its seat and its focus,
and without an external object would lose its significance,”
we can yet take appearance absolutely and
“inhibit all reaction and understanding”; but since
even the passive and immediate data of appearance,
“its bare signals and language when stupidly gazed
at” have æsthetic reality, “the special and insidious
kind of reality opposed to appearance must mean an
underlying reality, a substance: and it had better be
called by that name.” And he introduces to us
Essences = Universals = Intuited æsthetic data — “symbols
167of sense or thought” (p. 165), which may be
identical with the essences embodied in the substance
though “the intention and the embodiment remain
different in existence, origin, date, place, substance,
function and duration.”

That the individual's field of experience “has a
certain structure, and is shot through with meanings
and affirmations,” seems to Professor Sellars of Michigan
“a matter of undeniable fact.” The chief error
of much contemporary thought is the refusal to recognize
“that thinghood and perception go together”; in
other words, in the percipient, “we have the content
of perception, and over against it in a qualifying way,
the motor complex of adjustment combined with the
realistic meanings and expectations which are characteristic
of perception.” What is needed is, he holds,
“a patient and persistent analysis which is able to go
forward step by step while doing justice to the structure
and meanings of the individual's experience” (p. 197).
And as regards knowledge of the past, “we can mean
a reality which no longer exists equally with a reality
which exists at the time of the intention” (p. 215).

Finally Professor Strong who examines the nature
of the ‘datum,’ which he replaces by Santayana's
‘essence’ (which, as we have seen, is regarded by
Critical Realism as also equivalent to ‘meaning’), concludes
that data are in their nature “not existences
but universals, the bare natures of the objects, in such
wise that the essence embodied and the essence given
may be the same.” 168

“What is given to us in sense-perception,” we learn
(p. 235), “is the sensation as a meaning, or to speak
more correctly what is given is the meaning and not
the sensation…. That this significance, or meaning,
or essence is not an existence and not in time and space,
but, like the meaning when we think of a universal,
a purely logical entity, is quite credible”; moreover,
the datum “is not properly a sensible fact. We cannot
actually find it as a feeling, we can only tend towards
it or mean it… A meaning here is not to be
understood as a peculiar kind of feeling, but as a
function which the feeling discharges” (p. 237).

We need not here attempt to correlate these different
uses of the term in what claims to be the last achievement
of co-ordinated symbolization. As might have
been expected this statement with its challenge to Neo-Realists,
Pragmatists, and Idealists aroused abundant
controversy, but the one inevitable source of misunderstanding
and disagreement, the omnipresence of
the term Meaning, was allowed to pass unchallenged.
It seems to have been accepted without question into
the vocabulary of American philosophy, for use on all
occasions of uncertainty, 1168 though to the English reader
it still happily sounds strange in most of its typical
contexts.

Part I is entitled ‘The Meaning of Values’ and on
the six pages 74-79 which reveal “the deciding fact”
the term ‘meaning’ appears no less than sixteen times.
The deciding fact is that we demand that things recur.
“We demand that there be a world; that means that
our experience be more than just passing experience.
Here is the original deed which gives eternal meaning
to our reality” (p. 75). “The world becomes a world
by its identical recurrence, and this identity means
fulfilment, means satisfaction, means value” (p. 79).

In passing it may be noted that identity does not
exclude change, for it is postulated that whatever
changes “must still present an identity in its changes
by showing that the change belongs to its own
170meaning.” Indeed “our question as to the validity
of pure values can have no other meaning except in
reference to this true world,” the world “of our experiences
in so far as they assert themselves;” and “it
would be meaningless to deny the question”.

To complete the argument with this accommodating
linguistic material, it would seem that since its identical
recurrence presumably is the ‘meaning’ of anything,
and since the ‘meaning’ of anything is presumably its
value, the statement above that “identical recurrence
means value” might equally well have appeared in the
form meaning means meaning.

And so, on the next page (430), the last of the book,
we conclude with the assurance that “To progress in
the sense of the self-assertion of the will in will-enhancement
remains for mankind, too, the ultimate
meaning of duty.”

A study of these extracts in the German version of
Münsterberg's work is an interesting exercise in comparative
linguistic, and the contribution of the term
‘meaning’ to the cogency of the argument is considerable.
There may be those who find it hard to believe
that any writer responsible for such a verbal exploit
could also enjoy a reputation as a thinker of the first
rank. There is, however, another ambitious modern
attempt by an American theorist to deal specifically
with the fundamentals of psychology; and in the
preface to this work 1169 we find a reference to Münsterberg's
“illuminating work on the great problems of
philosophy and of natural and mental science…
It may be truthfully said that in his death America has
lost its one great theoretical psychologist.” Professor
Moore has no occasion to quote largely from the particular
work selected above, but his extracts (pp. 107-110)
from Münsterberg's Psychology General and Applied,
and Psychotherapy are equally bespattered with the term.
And as might be expected Professor Moore's own
treatment is also vitiated at its most crucial points by
his too hospitable attitude to this plausible nomad.173

To understand the nature of psychology as a science
we must, he holds, carefully distinguish Science from
Metaphysics, and “the key-word of the problem of
metaphysics is Interpretation. To interpret anything is
to determine its meaning. If the fundamental presupposition
of all science is that every fact has a cause;
the fundamental presupposition of metaphysics is that
every fact has a meaning” (p. 97). In other words, in
philosophy as opposed to science, “each fact is treated
not as the effect of some antecedent cause, but as the
expression of a Meaning.” Science must precede
metaphysics — “We cannot know what facts mean until
we know what the facts are, we cannot interpret the
facts until we have described them.”

But, objects the critic (p. 100), “is it not true that
the very essence of a mental process is its meaning?”
No. Titchener has given six good reasons why mental
processes are “not intrinsically meaningful” (p. 101).
But, the critic insists (p. 102), Do not all our experiences
“in their inmost nature mean something. Do we ever
experience a ‘meaningless’ sensation?” We have
no reason, the reply runs, to believe that the mind
“began with meaningless sensations, and progressed
to meaningful perceptions. On the contrary we must
suppose that the mind was meaningful from the very
outset.”

And here we pause at the very pertinent question:
“What then from the psychological point of view is
this meaning?” The answer is given without hesitation
and in italics — “From the psychological point of
view, meaning is context.” To explain: In every perception,
or group of sensations and images, “the
associated images form as it were a context or ‘fringe’
which binds together the whole and gives it a definite
meaning,” and it is this “fringe of meaning that
makes the sensations not ‘mere’ sensations but
symbols of a physical object.” So when we see an
orange it is the contextual images of smell and taste
174“which enable us to ‘recognize’ the object — i.e., give
a meaning to the sensations” of colour and brightness.
Similarly (p. 103) “every idea has a core or nucleus of
images, and a fringe of associated images… which
give meaning to the nuclear images.”

It is a curious approach to the problems of sign-interpretation,
this account of Meaning which (psychologically)
is context, which is carried by context, which
is much more than context, which is expressed by facts,
with which psychology is not concerned — and yet is concerned,
so far as it can be represented by contextual
imagery. 1170175

But there are stranger things to follow, for here
True Meaning makes its appearance — in connection
with a bell. “The true meaning of the percept of the
bell is its reference to the real objective bell,” and this
reference is represented in the mind by contextual
images which “constitute its meaning ‘translated into
the language of’ psychology. So the true meaning of
an idea lies in its logical reference to an objective
system of ideas” (p. 104); and a little later (p. 111) we
find that “all experiences are expressions of the inner
meanings of the self.”

It is hard to believe that Professor Moore would
have been satisfied with such a vocabulary had he
attempted to investigate the psychology of signs and
symbols; and this investigation could not but have
shown him how much of his present work had its
origin in an unfortunate choice of, and attitude towards,
symbols. As it is, the constant appeal to an esoteric
Doctrine of Meaning is reminiscent of the dialectical
devices of mediæval theologians, and we may conclude
by noting that the Doctrine is specifically invoked in
relation to Religion.

“Psychology may discuss as freely the mental processes involved
in religious experience as it does those concerned in our
experience of physical things; but in neither case can its decisions
affect the question of the meaning… of those experiences.
The question of the nature of the processes undergone by the
human mind in any spheres of activity is a question of fact,
calling for analytical description and explanation in causal terms:
the problem of the validity or truth-value of these processes is a
question of meaning, calling for interpretation” (p. 122).

For those who regard interpretation as a purely
causal process, and consider that when the meaning of
anything is interpreted it is but explained in causal
terms (while at the same time recognizing a totally
176distinct sense of meaning in which the ‘meaning’ of a
poem or a religion would be the emotion or attitude
evoked through it), the extent to which this symbol
can change places with its other selves should provide
material for reflection.

Our object here, however, is rather to provide
instances of its use in current constructive and controversial
literature, and it remains only to group together
a few further typical examples.

“We have direct acquaintance with the ideas or
meanings about which we have thoughts and which
we may be said to understand,” declared the late Lord
Keynes; and again, “We are able to pass from direct
acquaintance with things to a knowledge of propositions
about the things of which we have sensations
or understand the meaning.” 1175 So helpful a term is
equally in demand as a carminative in ecclesiastical
controversy, 2176 as a vade mecum in musical criticism, 3177
as an indication of the precise point where doctors
differ, 4178 and as a lubricant for the spinning-wheel of
the absolute relativist. 5179 “If education cannot be
178identified with mere instruction, what is it? What
does the term mean?” asks the educationist. “I
answer, it must mean a gradual adjustment to the
spiritual possession of the race.” 1180 Meaning is therefore
just the sort of word with which we may attempt
to probe the obscure depths of the souls of fishes.
“Let us fix attention on the state of the mind of the
goldfish… Suddenly comes a new element into
consciousness — the conscious counterpart of the stimuli
of the eye caused by the bread falling into the water…
The food is an object in space and time for the fish and
has its meaning, but when the food is eaten both percept
and meaning disappear… This is an instance
of percept and meaning tied.” 2181

As a specimen of the language of Psycho-analysts,
on the other hand, the following by the late Professor
J. J. Putnam 4188 of Harvard may be considered: —

“It seems, and is, a small matter to walk in the country
without one's coat, but a similar insufficiency of costume, if
occurring in a dream, may be a circumstance of far wider meaning…
It will be obvious from the foregoing that the term
‘sexual’ as defined in the psycho-analytic vocabulary, is of far
wider meaning than is ordinarily conceived… The next point
has reference to ‘sublimation.’ This outcome of individual
evolution, as defined by Freud, has a strictly social meaning…

The logical end of a psycho-analytic treatment is the recovery
of a full sense of the bearings and meanings of one's life.

A man's sense of pride of his family may be a symptom of
narcistic self-adulation; but like all other signs and symbols,
this is a case where two opposing meanings meet…”

The Pragmatists made a bold attempt to simplify
the issue. “That which is suggested is meaning,”
wrote Professor Miller, 5189 and Professor Bawden 6190 is
equally simple — “Feeling is the vague appreciation
of the value of a situation, while cognition is a clear
and distinct perception of its meaning.” The trouble
begins, however, with the first attempts at elaboration.
180“An experience is cognitional,” says Professor Dewey 1191
“which is contemporaneously aware of meaning something
beyond itself. Both the meaning and the thing
meant are elements in the same situation… One
is present as not-present-in-the-same-way-in-which-the-other-is.
… We may say that the smell of a rose,
when involving conscious meaning or intention, is
mental.”

Nor is it only in Ethics that important philosophical
positions are based on this arbitrary foundation.
“Things, as we know, are largely constructions,” says
one modern metaphysician 2197 — “a synthesis of sense
elements and meanings… The concept is no mere
word, because it has meaning… A universal, as
the object of a meaning, is not a mental act.” It is
impossible, urges another, 3198 who also speaks of “analysing
the meaning of a process of change from a conceptual
point of view,” to imagine “that we ourselves
can be analysed into sense-data, for sense-data are
‘given’ or ‘presented’ by the very meaning of the
term.” And again, “It is doubtless true that ‘body’
and ‘mind’ are used with more than one meaning to
which a reasonable significance may be attached.” 4199
Meanings to which significance is attached have also
the authority of Lotze, 5200 who held that “historical
persons and events, in spite of all the significance
attached to their meaning, are often very insignificant
182in the external form of their appearance,” and who
also informs us that in Moorish architecture “the
soaring pointed bow of horse-shoe shape has no
properly constructive meaning, but rather recalls the
mighty opening of a cleft” (p. 66), while a landscape
in pictorial composition “has a meaning only as a part
of the actual world” (p. 82).

Æsthetics, however, has always flourished on loose
usage, and non-philosophic writers have here been
more than usually persistent in their invocation of the
word at all critical points. “Colour as colour,” writes
Van Gogh, “means something; this should not be
ignored, but rather turned to account.” 1201 The poet, too,
we read, “said what he meant, but his meaning seems
to beckon away beyond itself, or rather to expand into
something boundless which is only focussed in it.” 2202

Chapter IX
The meaning of meaning

Father! these are terrible words, but I have no time
now but for Meanings. — Melmoth the Wanderer.

A study of the utterances of Philosophers suggests
that they are not to be trusted in their dealings with
Meaning. With the material which they have provided
before us, let us see whether more creditable results can
be achieved by the technique which we have already
elaborated.

To begin with it is not difficult to frame two definitions
corresponding to those of Group A in the case of
Beautiful. In two ways it has been easy and natural
for philosophers to hypostatize their definiendum;
either by inventing a peculiar stuff, an intrinsic
property, and then saying let everything which
possesses this be said to possess meaning, or by
inventing a special unanalysable relation, and saying
let everything related by this relation to something
else be said to have a meaning.

With the second of these two definitions a grammatical
alternative is opened up which reappears in
all the other suggested definitions and tends very
greatly to confuse the discussion. We may either take
Meaning as standing for the relation between A and B,
when A means B, or as standing for B. In the first
case the meaning of A will be its relation to B, in the
second it will be B. This ambiguity once it is understood
gives rise to little difficulty, but the avoidance of
it by the symbols ‘reference’ and ‘referent’ is one of
the distinct advantages of that vocabulary.185

The other definitions show again a similarity with
those of Beautiful in that they are preponderantly
psychological definitions. It should not however, be
concluded from these two examples that all definition
problems develop into psychology. If we were attempting
to define ‘bathing’ or ‘absorption,’ let us say,
we should find the emphasis upon quite different
defining routes. ‘Meaning’ evidently is a symbol
some of whose elucidations must rest upon psychology,
and the example of Beauty was chosen because that
symbol, too, lies though less deeply in the same
predicament.

With Group A we need be no further concerned.
Let us consider Group B. The first (III) Dictionary
meaning, or the philologist's signification, is, in spite
of its comical appearance as formulated above, very
widely used; and in the domain of philology it has
undoubted value, as will be shown when we come to
discuss, in the light of definition XIV, the kindred
questions of Good Use and Communication.

Connotation (IV) the ‘meaning’ of traditional logic,
and Essence (V) the ‘meaning’ of the Critical Realists
who follow Dr Santayana as quoted above, may be
considered together, for ‘Essences’ by those who do
not let their realism overpower their criticism may
best be regarded as Connotation hypostatized.

The term Connotation has been adopted by those
logicians who follow Mill in the practice of discussing
as though they were primary and paramount two senses
in which a symbol may be said to mean: (1) It means
the set of things to which it can be correctly applied;
and the members of this set are said to be denoted
or indicated by the word, or to be its denotation. (2) It
means the properties used in determining the application
of a symbol, the properties in virtue of which anything
187is a member of the set which is the denotation; these
properties are said to be the connotation of a symbol,
or sometimes simply its meaning. The relation of
denotation to connotation has been conveniently summed
up as follows: The connotation of a word determines
its denotation which in turn determines its comprehension,
i.e., the properties common to the things to
which it can be applied. The term connotation is,
however, often used with the same sense as comprehension.

It will be plain to all who consider how words are
used that this account is highly artificial. Neither
denoting nor connoting can be used as if it were either
a simple or a fundamental relation. To take denotation
first, no word has any denotation apart from some
reference which it symbolizes. The relations between
a word and the things for which it stands are indirect
(cf. diagram, Chapter I., p. 2), and, we have urged,
causal. When we add the further complications introduced
by correct usage, we get a result so artificial
that the attempt to use ‘denoting’ as the name of a
simple logical relation becomes ludicrous. The case
is still worse with ‘connoting.’ The connotation is
a selection of properties or adjectives; but properties
are not to be found by themselves anywhere; they are
fictitious or nominal entities which we are led to feign
through the influence of the bad analogy by which
we treat certain parts of our symbols as though they
were self-complete symbols. We have no justification,
beyond this bad analogy, for treating adjectives as
though they were nouns. The sole entities in the
real world are propertied things which are only
symbolically distinguishable into properties and things.
This does not, of course, make symbolization, which
proceeds as though properties and things were separable,
any less desirable upon occasion. No convenient
symbolic device is objectionable so long as we know
that it is a device and do not suppose it to be an addition
188to our knowledge. To let a convenience turn into an
argument, and decide for us as to the nature of the
universe in the fashion of Dr Santayana's ‘Essences’
is a gratuitous tactic. On the other hand as linguistic
machinery there is no harm and much service in
universals. For instance, in expounding the causal
or contextual theory of reference we made free use of
the terms ‘character’ and ‘relation’ as though these
might stand for independent and respectable elements
in the real world. There is a linguistic necessity for
such procedure but to exalt this into a logical necessity
for the ‘subsistence’ of such elements is to forget what
the world is like.

Thus, to begin with, the connotation of a word is
a set of nominal entities, but we have still to decide
which these shall be. One method would be by linguistic
usage; “a knowledge of the usage of language alone
is sufficient to know what a phrase means,” says Mr
Johnson (Logic, p. 92). According to this method, if
strictly followed, the connotation of a word would become
indistinguishable from its meaning in the sense
of “the other words annexed to a word in the dictionary”
(III). But another method is possible, the
consideration of which will show more plainly still the
artificiality of connotation and the little reliance which
can be placed in it for logical purposes; for instance,
in definition. We can in part translate the convenient
formula given above as follows: The reference employing
(or symbolized by) a word determines its referents
(i.e., denotation) which in turn determine what different
references may be made to them. Two symbols would
then have the same connotation when they symbolize
similar references. But in our account of reference
anything becomes a referent for a given process or
act of referring only in virtue of certain characters
through which it becomes a completing member of the
context including the sign for the process. Thus the
connotation of a reference (and derivatively of the words
189symbolizing it) would be those characters of its referent
in virtue of which this is what is referred to. Bearing
in mind that these characters are but nominal entities
we can now see how easy it has been for logicians
through the formidable shorthand of ‘denotation’ and
‘connotation’ as applied to words to overlook the
causal nature of the relations they were unwittingly
discussing. It is not surprising that the attempt to
explain the relation of meaning to denotation for phrases
like ‘The King of France’ by such shorthand methods
should have been found difficult. 1209

One further point amusingly shows the artificiality
of the traditional account, namely, the impossibility of
applying it to names, which without undue rashness
may be regarded as the simplest symbols out of which
all our other symbolic machinery has developed. Mill
concluded that proper names are non-connotative.
Mr Johnson in agreeing with him (and “all the best
logicians” ) makes a reservation: 2210 —

“This does not amount to saying that the proper
name is non-significant or has no meaning; rather we
find, negatively, that the proper name does not mean
the same as anything that could be meant by a
descriptive or connotative phrase; and positively, that
it does precisely mean what could be indicated by some
appropriate descriptive phrase.” Further shifts 3211 are
190then necessary, but serve only to destroy ‘meaning’ as
a useful symbol.

VI, though it appeals to Empathists, Croceans and
Solipsists, is most charitably regarded as a metaphor,
in which case it is a strange and striking way of
phrasing views closely similar to XIII. Dr Schiller's
way of putting it, “Meaning is an activity taken up
towards objects and energetically projected into them
like an α particle,” obscures his actual agreement with
the mnemic causation which he is combating; since
when he speaks of “a demand we make upon our
experience” as “selecting the objects of attention,” he
appears to be describing in activist language the very
processes (cf. XIII (a) infra) which he is so unwilling
to admit. The dispute between ‘act’ and ‘process’ as
fundamental psychological terms is obviously subsequent
to a full discussion of the problem of Meaning.
As is also indicated by Professor Strong's contribution 1212
we presumably have here an instance of a common
controversial predicament, the use for the same referents
of symbols taken out of different, but to a large extent
translatable, symbol systems.

We pass to VII, which arises from the study of such
remarks as

They meant no harm.

He means well.

I meant to go.

What I meant was what I said.

A mechanistic universe is without meaning.

If, as is usually the case when these phrases are
used, we can substitute the word ‘intend’ for ‘mean’
it will be clear that we have a quite different kind of
‘meaning’ from any involved when ‘intention’ cannot
191be so substituted. 1213 My ‘meaning’ or ‘intention,’ as
that which I endeavour to promote, is something wished,
as distinguished from something known or referred to
(‘intended,’ or ‘tended towards,’ in the terminology of
certain American writers). Thus between this sense
and that with which we have to deal in such sentences
as “‘ Chien’ and ‘Dog,’ both mean the same thing,”
there is no contradiction. There is, however, a pun,
and thanks to the practice of disputants who compound
the sense of reference with the sense of intention in the
phrase “What I meant was” (= “What I intended to
refer to was” or “what I intended you to refer to was” )
— we have a dangerous source of confusion. The
difficulty of making a close examination of the matter
under discussion is greatly increased, for what I
intended to refer to may be quite other than what I did
refer to, a fact which it is important to remember if it
is hoped to reach mutual comprehension, and eventually
agreement or disagreement.

The intention of the speaker may very naturally be
used in conjunction with reference in order to provide
complex definitions of meaning for special purposes.
To quote from a recent article: “Is the meaning of a
sentence that which is in the mind of the speaker at the
moment of utterance or that which is in the mind of the
listener at the moment of audition? Neither, I think.
192Certainly not that which is in the mind of the listener,
for he may utterly misconstrue the speaker's purpose.
But also not that which is in the mind of the speaker,
for he may intentionally veil in his utterance the
thoughts which are in his brain, and this, of course,
he could not do if the meaning of the utterance were
precisely that which he held in his brain. I think the
following formulation will meet the case: The meaning
of any sentence is what the speaker intends to be understood
from it by the listener.” 1214

‘To be understood’ is here a contraction. It stands
for: (a) to be referred to + (b) to be responded with + (c)
to be felt towards referent + (d) to be felt towards speaker
+ (e) to be supposed that the speaker is referring to +
(f) that the speaker is desiring, etc., etc.

These complexities are mentioned here to show how
vague are most of the terms which are commonly
thought satisfactory in this topic. Such a word as
‘understand’ is, unless specially treated, far too vague
to serve except provisionally or at levels of discourse
where a real understanding of the matter (in the reference
sense) is not possible. The multiple functions of speech
will be classified and discussed in the following chapter.
There it will be seen that the expression of the speaker's
intention is one of the five regular language functions.
It should not be stressed unduly, and it should be
remembered that as with the other functions its importance
varies immensely from person to person and
from occasion to occasion.

The realization of the multiplicity of the normal
language function is vital to a serious approach to the
problem of meaning. Here it is only desirable to point
out that ‘meaning,’ in the sense of ‘that which the
speaker intends the listener to refer to,’ and ‘meaning,’
in the sense of ‘that which the speaker intends the
listener to feel and to do,’ etc., are clearly distinguishable.
193In many of the more subtle speech situations these
distinctions must be recognized and used.

The first of these is particularly concerned in those
cases of misdirection which we saw in our first chapter
to be so universal. In the case of a successful lie the
person deceived makes the reference which the deceiver
intends he shall, and if we define ‘meaning’ as ‘that
which the speaker intends the listener to refer to,’ the
victim will have interpreted the speaker aright. He will
have grasped his meaning. But let us consider a more
astute interpreter, who, by applying a further interpretative
process (based, say, upon his knowledge of
business methods) arrives either at a mere rejection of
the intended reference or at another reference quite
different from that intended. In the latter case, if he
has hit upon the reference from which the suggested
false reference was designed to divert him, he would
often be said to have understood the speaker, or to have
divined his ‘true meaning.’ This last meaning, it
should be observed, is non-symbolic. The sagacious
listener merely takes the speaker's behaviour, including
the words he utters, as a set of signs whence to interpret
to an intention and a reference in the speaker which no
words passing on the occasion symbolize. The batsman
who correctly plays a ‘googly’ is making exactly the
same kind of interpretation. He guesses the ‘meaning’
of the bowler's action by discounting certain of
the signs exhibited.

All cases of ‘duplicity,’ whether deliberate (intentional)
or not, may be analysed in the same manner; 1215
the special instance of self-deception as it concerns
introspective judgments, which are discussed below,
being of most importance for the general theory. Here
great care is required in avoiding any confusion between
the speaker's intended or professed references and his
actual references.194

This particular ambiguity is indeed one of the most
undesirable of those with which we have to deal. Unless
the referential and the affective-volitional aspects of
mental process are clearly distinguished, no discussion
of their relation is possible; and the confusion of reference,
with one very special form of the latter aspect,
namely ‘intending,’ is disastrous. To bring the point
out by a play of words, we very often mean what we do
not mean; i.e., we refer to what we do not intend, and
we are constantly thinking of things which we do not
want to think of. ‘Mean’ as shorthand for ‘intend to
refer to,’ is, in fact, one of the unluckiest symbolic
devices possible.

The distinction between the two aspects of mental
process from the standpoint of the context theory may
be briefly and therefore vaguely indicated as follows:
Given the psychological context to which a sign belongs,
then the reference made by the interpretation of the sign
is fixed also. But it is possible for the same sign (or for
signs with very similar characters) to belong to different
psychological contexts. Certain geometrical figures,
that may be seen, more or less ‘at will,’ either as receding
or as extruding from the plane upon which they are
drawn offer well-known and convenient examples. If
now we raise the question, How does the sign come to
belong to the context to which it does belong, or how
does it pass from one context to another? we are raising
questions as to the affective-volitional aspect. The facts,
concerning habit-formation, desire, affective tone, upon
the basis of which these questions must be answered,
are to some extent ascertained; but pending the discovery
of further facts and an hypothesis by which they
can be interpreted and arranged, it remains possible to
speculate upon the matter either in activist or in automatist
language. Which kind of language gives
scientifically the most adequate symbolism, or whether
a neutral symbolism is not possible, are matters as to
which it is premature to decide. Meanwhile there is no
195excuse for making a confused statement of an unsolved
and difficult problem into a chief instrument of all our
inquiries, which is what we should be doing if we
admitted ‘meaning’ in the sense here discussed as a
fundamental conception.

As regards VII (b) those who are not clear as to the
scope of the equation, “His meaning is certain,” = “He
has definite wishes,” often find themselves led to the
conclusion that ‘meaning’ = ‘wishes’ = ‘volition’ (a
mental event), i.e., is entirely psychological, or as they
are often pleased to say, purely personal. 1216 The same
linguistic ambiguity often arises again when the Universe
is regarded as showing evidence of a will or design,
and if ‘meaning’ is substituted for the ‘intention’ or
‘purpose’ of such a will, then the meaning of anything
will be its purpose — as conceived by the speaker qua
interpreter of the divine plan; or, for biological teleologists
with a partiality for the élan vital — its function.
Such a phrase as the Meaning of Life (cf., for example,
Professor Münsterberg's treatment above) usually implies
such a view, but there is sometimes another possible
interpretation when Meaning is equated with ‘Significance’
(VIII). Here the notion of purpose is not always
implied, and the meaning of anything is said to have
been grasped when it has been understood as related
to other things or as having its place in some system as
a whole.

Good examples of both these uses are provided by
Mr Russell, and it is hardly necessary to add that, as
here used by him, both are innocuous and convenient
locutions. At the close of the immortal account by
Mephistopheles of the history of our cosmos, we read:
“Such, in outline, but even more purposeless, more
void of meaning, is the world which Science presents
196for our belief.” And again, in relation to the haphazard
treatment of mathematics in text-books: “The love of
system can find free play in mathematics as nowhere
else. The learner who feels this impulse must not be
repelled by an array of meaningless examples or distracted
by amusing oddities.” 1217

The kind of system within which the thing, said in
this sense to have ‘meaning,’ is taken as fitting is not
important. Designs or intentions, human or other,
form one sub-class of such systems, but there are many
others. For example, some people were said to be
slow in grasping the ‘meaning’ of the declaration of
war; in other words, they did not easily think of the
consequences of all kinds which were causally linked
with that event. Similarly we may ask what is the
‘meaning’ of unemployment.

The theologian will elucidate the ‘meaning’ of sin
by explaining the circumstances of Adam's fall and
the history and destiny of the soul. Similarly the
‘meaning’ of top hats may flash across the mind of a
sociologist when he recognizes them as part of the
phenomena of conspicuous ostentation.

“I doubt,” says Mr Stanley Leathes, “if numerical
dates have any meaning to the majority of children.
I once asked a Sunday school boy: How long ago
Our Lord had lived? He replied: ‘Forty days.’” 2218
The complaint is not that the dates do not ‘suggest’
anything, but presumably that their ‘significance’ in
the general measurement of time has not been grasped
by the puerile mind. The figures for the distances of
remote stars are similarly said to be without ‘meaning’
for us all.

But ‘meaning’ in this sense is too vague to be of
much service even to orators. Is the meaning of
unemployment its causes or its effects, its effects taken
sociologically, or as the unemployed individual suffers
197them? Accordingly various restrictions are commonly
introduced by aid of which more specific senses of
‘meaning,’ as place within some system, are obtained.
Two of them are sufficiently important to rank as
independent definitions of meaning, since each has
been made the keystone of a metaphysical edifice,
namely ‘meaning’ as the practical and as the theoretical
consequences. In both cases the ‘meaning’ is the rest
of the system within which whatever has the ‘meaning’
is taken. We shall find another narrower and a more
scientific variety of this ‘meaning’ in use when we
come to consider natural signs.

The account of meaning in terms of Practical
Consequences (IX) is chiefly associated with the
pragmatists. William James himself considers that
“the meaning of any proposition can always be
brought down to some particular consequence in our
future practical experience, whether passive or active,” 1219
or as he puts it in Pragmatism (p. 201): “True ideas
are those that we can assimilate, validate, corroborate,
verify. False ideas are those that we can not. That is
the practical difference it makes to us to have true
ideas; that, therefore, is the meaning of truth, for it is
all that truth is known as.”

Correspondingly there are those who introduce the
word ‘means’ into their prose as a synonym for
‘involves’ or ‘logically implies’ (X). All or any of
the theoretical consequences of a view or statement are
thus included in common philosophic parlance in its
‘meaning,’ as when we are told (Mind, 1908, p. 491)
that “while to Spinoza insistence on ends alone means
ignorance of causes, to Prof. Laurie insistence on
causes alone means ignorance of ends.”

XI (Emotion) requires little comment. It is a
definite sense of meaning which except amongst men
of letters is not likely to be brought in to confuse other
issues. A separate treatment of the emotional use of
198language will be found in the following chapter, where
what has already been said on this subject receives
application. Some typical instances of the emotional
use of meaning were provided in the preceding chapter.
The word is often purely emotive (cf. ‘Good’ p. 125),
and on these occasions, if the writer is what is known
as a stylist, will have no substitute nor will a sensible
reader attempt a symbolic definition.

The detailed examination of this sense of meaning
is almost equivalent to an investigation of Values, such
as has been attempted by Professor W. M. Urban in
his formidable treatise on the subject, where ‘worth-predicates’
appear as ‘funded affective-volitional
meanings.’ “The words ‘God,’ ‘love,’ ‘liberty,’ have
a real emotional connotation, leave a trail of affective
meaning… We may quite properly speak of the
emotional connotation of such words as the funded
meaning of previous emotional reactions and the
affective abstracts which constitute the psychical
correlates of this meaning as the survivals of former
judgment-feelings.” 1220 It is regrettable that Urban's
taste for the collocation of forbidding technicalities
should have prevented a more general acquaintance
with views for the most part so sound and so carefully
expounded.

Proceeding then to the third group we have first
(XII) the definition which embodies the doctrine of
natural signs. Any one event will, it is generally
assumed, be connected with other events in a variety
of ways. Any one event will be actually related
causally or temporally or in some other way to other
events so that, taking this event as a sign in respect
of some one such relation, there will be another event
which is its meaning, i.e., the relatum so related.
Thus the effect of the striking of a match is either a
flame, or smoke, or the head falling off, or merely a
scraping noise or an exclamation. In this case the
199actual effect is the meaning of the scrape, if treated as
a sign in this respect, and vice versa.

It is in this sense that the Psycho-analyst often
speaks of the meaning of dreams. When he discovers
the ‘meaning’ of some mental phenomenon, what he
has found is usually a conspicuous part of the cause,
and he rarely makes any other actual use of the word.
But by introducing theories of unconscious wishes,
‘meaning’ in the sense of something unconsciously
intended, and by introducing ‘universal symbols,’
kings, queens, etc., ‘meaning’ in the sense of some
intrinsic property of the symbol, may easily come to be
what he believes himself to be discussing. In other
words, for him as for all natural scientists the causal
sign-relations are those which have the greatest interest.

In passing from this sense of ‘meaning’ to XIII,
which must be carefully distinguished, we have to
recall the account of interpretation given above. All
thinking, all reference, it was maintained, is adaptation
due to psychological contexts which link together
elements in external contexts. However ‘universal’
or however ‘abstract’ our adaptation, the general
account of what is happening is the same. In this
fashion we arrive at a clear and definite sense of
‘meaning.’ According to this the meaning of A is
that to which the mental process interpreting A is
adapted. 1221 This is the most important sense in which
words have meaning.

In the case of simple interpretations, such as the
recognition of a sound, this adaptation is not difficult
to explain. In more complex interpretations such as
the reader is attempting to carry out at this moment,
a detailed account is more difficult, partly because such
interpretations go by stages, partly because few important
psychological laws have as yet been ascertained
and these but vaguely. To take an analogous case,
before Newton's time scientists were in much doubt
200as to the ‘meaning’ of tidal phenomena, and peculiar
‘sympathy’ and ‘affinity’ relations used to be postulated
in order to connect them with the phases of the
moon ‘the ruler of the waters.’ Further knowledge of
more general uniformities made it possible to dispense
with such phantom relations. Similarly more accurate
knowledge of psychological laws will enable relations
such as ‘meaning,’ ‘knowing,’ ‘being the object of,’
‘awareness’ and ‘cognition’ to be treated as linguistic
phantoms also, their place being taken by observable
correlations.

The most usual objections to such a view as this
derive from undue reliance upon introspection. Introspective
judgments like other judgments are interpretations.
Whether we judge ‘I am thinking of rain,’ or,
after looking at the barometer, judge ‘It is going to
rain’; we are equally engaged in a sign situation. In
both cases we are making a secondary adaptation to a
previous adaptation as sign, or more usually to some
part or concomitant of the adaptation; such as, for
instance, the words symbolizing the reference about
which we are attempting to judge in introspection, or,
failing words, some non-verbal symbol, or, failing even
that, the obscure feelings accompanying the reference.
It is possible of course to respond directly to our own
responses. We do this constantly in long trains of
habitual and perceptual actions; but such responses
being themselves non-conscious, i.e., conscious of nothing,
do not lead to introspection judgments of the kind
which provide evidence for or against any view as to
the nature of thinking. Such judgments, since they
must appear to rest upon the reflective scrutiny of consciousness
itself, are interpretations whose signs are
taken from whatever conscious elements accompany the
references they are about. It is certain that these signs
are unreliable and difficult to interpret; often they are
no more than dim, vague feelings. We therefore tend
to introduce symbolization, hoping so to gain additional
201and clearer signs. When, for instance, we attempt
what is called the analysis of a judgment by direct
introspection our procedure leads as a rule to the provision
of alternative symbols which we endeavour to
convince ourselves symbolize the same reference. We
then say that one symbol is what we mean by the other.
In most modern arguments concerning fundamentals
some positive or negative assertion of this form can be
found as an essential step. It is thus very important
to consider what kind of evidence is available for such
assertions.

The usual answer would be that it is a matter not
of evidence but of immediate conviction. But these
direct certainties notoriously vary from hour to hour
and are different in different persons. They are in fact
feelings, and as such their causes, if they can be
investigated, will be found not irrelevant to the question
of their validity. Now the main cause of any conviction
as to one symbol being the correct analysis of
another, i.e., as to the identity of the references symbolized
by both, is to be found in the similarity of any
other signs of the references in question which may be
obtainable. These, since imagery is admittedly often
irrelevant, will be feelings again: — feelings accompanying
the references, feelings of fitness or unfitness, due
to the causal connections of symbols to references, and
feelings due to the mere superficial similarities and
dissimilarities of the symbols. Thus it is this tangled
and obscure network of feelings which is the ground
of our introspective certainties. It is not surprising
that the task of clarifying our opinions by the method
of direct inspection and analysis should be found
difficult, or that the results obtained should give rise
to controversy.

Those who have attempted to decide what precisely
they are judging when they make the commonest judgments,
such as ‘I am thinking,’ ‘That is a chair,’
‘This is good’ will not be in haste to dispute this.
202It is indeed very likely that we more often make mistakes
in these secondary judgments than in most others,
for the obvious reason that verification is so difficult.
Nobody's certainty as to his reference, his ‘meaning,’
is of any value in the absence of corroborative 1222 evidence,
though this kind of self-confidence dies hard.

It is because the non-verbal sensations and images
which accompany references are such unreliable signs
that symbols are so important. We usually take our
symbolization as our guide to our meaning, and the
accompanying sign feelings become indistinguishably
merged in the feelings of our symbols. The fact, however,
that on some occasions all the available symbols
can be felt to be inappropriate to the reference which
they are required to symbolize, shows that other feeling-signs
are attainable. We are thus not completely at the
mercy of our symbols.

None the less, there are obvious reasons for that
prodigious trustfulness in symbols as indications of
what we are meaning which is characteristic of mathematical
and other abstract thinkers. Symbols properly
used are for such subjects indispensable substitutes for
feeling accompaniments not so easily distinguished.
The feeling accompanying, for instance, a reference
to 102 apples is not easily distinguishable from that
accompanying a reference to 103, and without the
symbols we should be unable to make either reference
as distinct from the other. In abstract thought as a
rule and for most thinkers, instead of our references
determining our symbols, the linkage and interconnection
of the symbols determines our reference.
203We merely watch that no violation of certain rules of
procedure is brought about. Some of these rules are
of no great importance, those recorded in the parts of
grammar which deal with literary usage and the conventions
of sentence formation. Others however are
of quite a different standing and are due to nothing
less than the nature of things in general. In other
words these rules are logical laws in the sense that any
symbol system which does not obey them must break
down as a means of recording references, no matter to
what the references be made. These fundamental
necessities of a symbol system and the mere rules of
polite speech above mentioned have historically been
subjected to some confusion. We had occasion to
discuss some of the former in Chapter V.; some of the
latter will receive mention and comment when we come
to deal with Symbol Situations in our final chapter.

Subject to these logical requirements we are able,
largely by means of symbols defined in terms of one
another, to compound references, or, in other words, to
abstract common parts of different references — to distinguish,
to compare and to connect references in, to,
and at, various levels of generality. The compounding
of these diverse modes of adaptation into a specific
judgment is the process generally alluded to as Thinking,
this activity being commonly maintained through
any long train by the use of symbols. These, as
substitutes for stimuli not available at any given instant,
as retaining the product of elaborate concatenations of
adjustments, and as affording means for the rearrangement
of these adjustments, have become so powerful,
so mechanical and so intricately interconnected as to
conceal from us almost entirely what is taking place.
We come to regard ourselves as related to a variety of
entities, properties, propositions, numbers, functions,
universals and so forth — by the unique relation of knowledge.
Recognized for what they are, i.e., symbolic
devices, these entities may be of great use. The attempt
204to investigate them as referents leads, as we have seen,
to Philosophy, and constitutes the unchallenged domain
of philosophers.

It will be noticed that definitions (XII) and (Xlllb)
for the case of true interpretations have the same effect.
The meaning (Xlllb) of a sign adequately interpreted
will be that to which it is actually related by the sign
relation. But for the case of false interpretations the
two ‘meanings’ will be different. Another point of
interest is that this account removes the necessity for
any ‘Correspondence Theory of Truth’ since an
adequate reference has as its referent not something
which corresponds to the fact or event which is the meaning
of a sign by definition (XII) but something which
is identical with it. We may if we please say that a
reference corresponds with its referent, but this would
be merely shorthand for the fuller account of reference
which we have given.

With these considerations before us we can now
understand the peculiarities of Symbols with their twofold
‘meaning’ for speaker and hearer. A symbol as
we have defined it (cf. pp. 11, 12 supra) symbolizes an
act of reference; that is to say, among its causes in the
speaker, together no doubt with desires to record and
to communicate, and with attitudes assumed towards
hearers, are acts of referring. Thus a symbol becomes
when uttered, in virtue of being so caused, a sign to a
hearer of an act of reference. But this act, except where
difficulty in understanding occurs, is of little interest
in itself, and the symbol is usually taken as a sign of
what it stands for, namely that to which the reference
which it symbolizes refers. When this interpretation is
successful it follows that the hearer makes a reference
similar in all relevant respects to that made by the
speaker. It is this which gives symbols their peculiarity
as signs. Thus a language transaction or a communication
may be defined as a use of symbols in such a way
that acts of reference occur in a hearer which are similar
205in all relevant respects to those which are symbolized
by them in the speaker.

From this point of view it is evident that the problem
for the theory of communication is the delimitation and
analysis of psychological contexts, an inductive problem
exactly the same in form as the problems of the other
sciences. Owing, however, to the difficulty of observing
psychological events and the superficial nature of
the uniformities hitherto observed, the methods employed
in testing whether communication has or has not taken
place are indirect. Since we are unable to observe
references directly we have to study them through
signs, either through accompanying feelings or through
symbols. Feelings are plainly insufficient and symbols
afford a far more sensitive indication. 1223 But symbols
also mislead and some method of control has to be
devised; hence the importance of definition. Where
there is reason to rely upon the indicative power of
symbols, no doubt a language purged of all alternative
locutions is scientifically desirable. But in most matters
the possible treachery of words can only be controlled
through definitions, and the greater the number of such
alternative locutions available the less is the risk of
discrepancy, provided that we do not suppose symbols
to have ‘meaning’ on their own account, and so people
the world with fictitious entities.

The question of synonyms leads us naturally to the
consideration of (XIV) Good Use. We have already
seen what correctness of symbolization involves. A
symbol is correct when it causes a reference similar to
that which it symbolizes in any suitable interpreter.
Thus for any given group of symbol users there will
arise a certain fixity of something which will be called
206proper meaning or Good Use. This something tends
to be spoken of as the meaning of the words in question.
What is fixed is the reference which any member of this
group will make in interpreting a symbol on any occasion
within the relevant universe of discourse. It is
no doubt very important that these meanings should
not vary beyond narrow limits. But we may be
legitimately anxious to maintain uniform standards
of comparison without finding it necessary to suppose
them supernaturally established or in their own nature
immutable. The belief which is so common that
words necessarily mean what they do derives from the
ambiguity of the term ‘necessary,’ which may stand
either for the fact that this is a requisite of communication
or for the supposed possession by words of intrinsic
‘meanings.’ Thus it has been argued that such a word
as Good has no synonym and is irreplaceable, so that
persons making good use of this word will have an
idea which they cannot otherwise symbolize — from
which it is held to follow that, since the word is
certainly used, there must be a unique and simple
ethical idea, or, as is sometimes said, a unique property
or predicate, whether possessed by anything or not.
In a precisely similar fashion mathematicians are apt
to aver that if nothing whatever existed, there would
yet be the property of ‘being 107 in number.’

These fixities in references are for the most part
supported and maintained by the use of Dictionaries,
and for many purposes ‘dictionary-meaning’ and ‘good
use’ would be equivalents. But a more refined sense
of dictionary-meaning may be indicated. The dictionary
is a list of substitute symbols. It says in effect: “This
can be substituted for that in such and such circumstances.”
It can do this because in these circumstances
and for suitable interpreters the references caused by
the two symbols will be sufficiently alike. The Dictionary
thus serves to mark the overlaps between the
references of symbols rather than to define their fields.207

The two remaining definitions of our list (XV.,
XVI.) arise through this difficulty in the control of
symbols as indications of reference. As we have seen,
the reference which the user of a symbol believes himself,
thanks to his trust in the symbol, to be making
may be quite different from that which he is actually
making; a fact which careful comparison of locutions
often reveals. Similarly the reference made by a hearer
will often be quite unlike that made by the speaker.
The final case, in which the meaning of a symbol is
what the hearer believes the speaker to be referring to,
is perhaps the richest of all in opportunities of misunderstanding.208

Chapter X
Symbol situations

For one word a man is often deemed to be wise
and for one word he is often deemed to be foolish.
We ought to be careful indeed what we say. —
Confucius.

Abba Ammon asked Abba Sisoes, saying, “When
I read in the Book my mind wisheth to arrange the
words so that there may be an answer to any question.”
The old man said unto him, “This is unnecessary,
for only purity of heart is required. From this
it ariseth that a man should speak without overmuch
care.” — Palladius, “The Book of Paradise.”

The context theory of interpretation as applied to the
use of words may now be sketched in outline. Let us
consider first the hearer's side of the matter, returning
later to the more difficult case of the speaker. As a
preliminary to any understanding of words, we necessarily
have a very simple kind of interpretation which
may be called sensory discrimination, or sensory recognition.
At this level 1224 we can be said to be discriminating
between sounds as sounds (the case where what
is discriminated is a movement of the organs of articulation,
or an image of this or of a sound, is quite
parallel); and thus we are here interpreting an initial
sign. Clearly unless one sound or image be distinguished,
consciously or unconsciously, from another
no use of words is possible. Usually the discrimination
209is unconscious, our use of words being habitual; it can,
however, become conscious, as in learning a foreign
tongue. One of the chief distinctions also between
poetry and strict scientific prose is that in poetry we
must consciously attend to the sensory characters of
the words, whereas in prose we need not do so. This
conscious attention to words as sounds does, however,
tend to impede our further interpretations.

The next stage of interpretation takes us from the
mere recognition of the initial sign as sound of a
certain kind to the recognition of it as a word. The
change is due to a change in the psychological context
of the sign. To recognize it as a sound with a distinctive
character we require a context consisting of the
sign and of other past sound sensations more and less
similar. To recognize it as a word requires that it form
a context with further experiences 1225 other than sounds.
In what precise fashion we first come to know that
there are words, or to take some sounds as words but
not others, is still to be experimentally investigated,
but as infants we do not make this step by guessing
straight off that people are talking to us. Long before
this surmise could become possible we have developed
an extensive private language through the fact that
certain sounds have come into contexts with certain
other experiences in such a way that the occurrence of
the sound is a sign interpreted by a response similar to
that aroused by the other associated experience. This
interpretation also may be conscious or unconscious.
Normally it is unconscious, but again if difficulty
arises it tends to become conscious. When we understand
with ease we are as a rule less aware of the words
used than when, through unfamiliarity of diction or the
strangeness of the referent, we are checked in our
interpretation.

These considerations are of importance in education.
210Many children appear more stupid than they are, not
through misinterpreting words but through failure to
recognize them first as sounds; and adults also differ
greatly in their ability to distinguish vocal sounds
when spoken rapidly or with an ‘accent.’ This ability
greatly affects the ease with which languages are
acquired.

With the recognition of the sound as a word the
importance of the prior recognition of the sound appears
to be decreased. This cannot actually be the case.
It is true that we can recognize a word whether it be
pronounced high or low, quickly or slowly, with a
rising intonation or a falling and so on. But however
different two utterances of one word may be as sounds,
they must yet have a common character; 1226 otherwise
they could not be recognized as the same word. It is
only in virtue of this character that the two sounds are
in similar psychological contexts and so interpreted
alike. We may be unable consciously to detect this
common character, but this need not surprise us. In
general it seems plausible to assume that simpler stages
of interpretation tend to lapse out of consciousness as
more elaborate developments grow out of them, provided
that they are successfully and easily carried out.
Difficulty or failure at any level of interpretation leads
in most cases to the re-emergence of the lower levels
into consciousness and to a kind of preoccupation with
them which is often an adverse condition for the
higher interpretations whose instability has led to their
emergence.

So far we have reached the level of the understanding
of simple names and statements, and a considerable
range of reference can be recorded and communicated
by this means alone. A symbol system of this simple
type is adequate for simple referents or aggregates of
simple referents, but it fails at once for complex
211referents, or groups of referents which have a structure
more intricate than mere togetherness. To symbolize
references to such complex referents complex symbols
with specialized structures are required; although it
does not appear to be necessary that the symbol should
in any very close way reflect or correspond to the
complexity of the referent. Possibly in primitive
languages this correspondence is closer. In highly
developed languages the means by which complex
symbols are formed, by which they receive their
structure as symbols, are very many and various.
Complex symbols with the same referent may be given
alternative forms even when the simple symbols, the
names, contained remain unaltered. The study of these
forms is a part of grammar, but a more genuine
interest in, and awareness of, psychological problems
than it is usual for grammarians to possess is required
if they are to be fruitfully discussed.

We may now consider a few of the easier cases of
these complex symbols. Let us begin with the contrast
between proper names and descriptive phrases. We
saw above that particular references require contexts of
a much simpler form than general references, and any
descriptive phrase involves for its understanding a context
of the more complicated form. To use such a
symbol as the name of an individual — let us call him
Thomas — we need merely that the name shall be in a
context with Thomas-experiences. A few such experiences
are usually sufficient to establish this conjunction;
for every such experience, since we rarely
encounter an acquaintance without realizing that he has
a name and what that name is, will help to form the
context. Contrast with this the understanding of such
a descriptive name as ‘my relatives.’ Here the experiences
required will not be in all cases the same.
At one time a grandfather, at another a niece will present
themselves; but not upon all occasions will their relationship
to us be in any degree a dominant feature, nor
212is the relationship which they agree in bearing to their
grandson and uncle respectively an obvious one. Thus
a range of experiences differing very widely one from
another is necessary if the required context is to be
built up.

‘Relatives’ is in fact an abstraction, in the sense
that the reference which it symbolizes cannot be formed
simply and directly by one grouping of experience, but
is the result of varied groupings of experiences whose
very difference enables their common elements to survive
in isolation. This process of selection and elimination
is always at work in the acquisition of a vocabulary
and the development of thought. It is rare for words
to be formed into contexts with non-symbolic experience
directly, for as a rule they are learnt only through other
words. We early begin to use language in order to
learn language, but since it is no mere matter of the
acquisition of synonyms or alternative locutions, the
same stressing of similarities between references and
elimination of their differences through conflict is required.
By these means we develop references of
greater and greater abstractness, and metaphor, the
primitive symbolization of abstraction, becomes possible
Metaphor, in the most general sense, is the use of one
reference to a group of things between which a given
relation holds, for the purpose of facilitating the discrimination
of an analogous relation in another group. 1227
In the understanding of metaphorical language one
reference borrows part of the context of another in an
abstract form.

There are two ways in which one reference may
appropriate part of the context of another. Thus a
reference to man may be joined with a reference to sea,
the result being a reference to seamen. No metaphor
is involved in this. When, on the other hand, we take
arms against a sea of troubles, that part of the context
213of the reference to sea which is combined with the other
references appears in an abstract form, i.e., the relevant
characters of the sea will not include attraction by the
moon or being the resort of fishes. The poetic value of
the metaphor depends in this case chiefly on the way
in which the ceaseless recurrence of the waves accentuates
the sense of hopelessness already present — as the
Cuchulain legend well shows.

In fact the use of metaphor involves the same kind
of contexts as abstract thought, the important point
being that the members shall only possess the relevant
feature in common, and that irrelevant or accidental
features shall cancel one another. All use of adjectives,
prepositions, verbs, etc., depends on this principle.
The prepositions are particularly interesting, the kinds of
contexts upon which they depend being plainly different
in extent and diversity of members. ‘Inside’ and
‘outside,’ it would appear, are the least complicated in
context, and consequently, as might be expected, are
easily retained in cases of disturbance of the speech
functions. The metaphorical aspects of the greater part
of language, and the ease with which any word may
be used metaphorically, further indicate the degree to
which, especially for educated persons, words have
gained contexts through other words. Very simple folk
with small and concrete vocabularies do on the other
hand in some degree approximate to the account given
above (p. 211), since the majority of their words have
naturally been acquired in direct connection with
experience. Their language has throughout many of
the characteristics of proper names. Hence in part their
comparative freedom from confusions, but hence also
the naive or magical attitude to words. Such linguists
may perhaps be said to be beneath the level at which
confusion, the penalty we pay for our power of abstraction,
becomes possible.

In what has been said hitherto we have dealt chiefly
214with the listener, who interprets symbols as they are
given to him. We have yet to examine the processes
by which references, as they proceed in a speaker, are
symbolized. This in some respects is the reverse of the
preceding case, but in others what happens is entirely
different. For the listener the word is the sign, and
without it the required reference does not occur. Possibly
for some mental types an exactly similar process occurs
in the speaker, with the sole difference that the words
are not given from without, but arise through some sort
of internal causation. Here there are not two distinct
processes, reference and symbolization, but only one —
reference through symbols; the situation being such
that the reference is governed by the symbol.

With most thinkers, however, the symbol seems to
be less essential. It can be dispensed with, altered
within limits and is subordinate to the reference for
which it is a symbol. For such people, for the normal
case that is to say, the symbol is only occasionally part
of the psychological context required for the reference.
No doubt for us all there are references which we can
only make by the aid of words, i.e., by contexts of which
words are members, but these are not necessarily the
same for people of different mental types and levels;
and further, even for one individual a reference which
may be able to dispense with a word on one occasion
may require it, in the sense of being impossible without
it, on another. On different occasions quite different
contexts may be determinative in respect of similar
references. It will be remembered that two references,
which are sufficiently similar in essentials to be regarded
as the same for practical purposes, may yet differ very
widely in their minor features. The contexts operative
may include additional supernumerary members. But
any one of these minor features may, through a change
in the wider contexts upon which these narrower
contexts depend, become an essential element instead
of a mere accompaniment. This appears to happen
215in the change from word-freedom, when the word is not
an essential member of the context of the reference, to
word-dependence, when it is.

The practical consequences of these differences
between individuals, and between occasions for the same
individual, are important. In discussion we have constantly
to distinguish between those who are unable to
modify their vocabularies without extensive disorganization
of their references, and those who are free to vary
their symbolism to suit the occasion. At all levels of
intellectual performance there are persons to be found
to whom any suggestion that they should change their
symbols comes, and must come, as a suggestion that
they should recant their beliefs. For such people to
talk differently is to think differently, because their words
are essential members of the contexts of their references.
To those who are not so tied by their symbolism this
inability to renounce for the moment favourite modes of
expression usually appears as a peculiar localized
stupidity. 1228 But it need not necessarily betoken a crude
and superstitious view of the relations of words to things,
for we should be ready to recognize that such adherence
to special words as though they had sovereign and
talismanic virtue, may be a symptom that for the speaker
the word is a necessary part of the reference context:
either because it was so when the reference was first
made, or because non-verbal signs alone would be
insufficient to avoid confusion. On the other hand, too
great a readiness to use any and every suggested symbol
216may also be a symptom of a low power of discrimination
between references; suggesting to the observer that the
speaker is making no fixed reference whatever.

But the symptomatology of language behaviour is
an intricate matter and little trust can be put in observations
which are not able to be checked by a wide knowledge
of the general behaviour of the subject. These
instances are here outlined merely to indicate the kind
of work which is still necessary. It is the sort of work
at which many people are by nature very successful;
they can often readily decide merely from the way in
which words proceed out of the mouth of a speaker, and
quite apart from the particular words, whether he is
worth listening to. A study of the mannerisms of
politicians and preachers is, however, useful as a check
upon too hasty conclusions. In general, the distinction
between those for whom reference governs symbol and
those for whom symbol governs reference, is constantly
required, although as we have already pointed out the
two conditions, word-independence and word-dependence
as they may be called, can rarely be found in
isolation, and most speakers alternate from one condition
to the other. In spite of this practical difficulty
the distinction between word-dependence and freedom
is one of the starting-points for linguistic investigation,
because the symptoms of nonsense-speech, verbiage,
psittacism or whatever we may elect to call the devastating
disease from which so much of the communicative
activity of man suffers, are quite different
for the two conditions, and, indeed, without the distinction,
are conflicting and ambiguous. Most writers
or speakers will agree from their own experience that
on some occasions their speech proceeds slowly, heavily
and importantly, because, while they are word-dependents,
the necessary words without which nothing whatever
would happen occur slowly and have to be waited
for, whereas on other occasions the words are emitted
in the same fashion because, being word-free for the
217moment, they are choosing the symbolism most suited
to the reference and to the occasion, with a view to some
finality of statement.

Neither of these speech processes can be dogmatically
established as the only right or proper process.
Word-dependence, for instance, must on no account be
identified with psittacism, or be regarded as necessarily
tending thereto. Psittacism is the use of words without
reference; and the fact that a word is necessary to a
reference is, as will easily be seen, in no way an indication
of an absence of reference. None the less if we
consider those other activities, such as eating or
bicycling, which are similar to speech in that they are
subject to a variable degree of control, there is reason
perhaps to decide in favour of a speech procedure which
should be a mingling of the two extremes of word-dependence
and word-freedom. At certain points in
serious utterances, the degree of deliberate control
should be at its maximum, i.e., the psychological context
into which the word fits and to which the reference
is due should contain as many varied members as
possible. The rest of the symbolization should be left
to the guidance of those systems of narrow contexts
which are called verbal habits, speech-mechanisms, or
the linguistic senses.

Considerable light upon the use of symbols is
thrown, as is always the case in psychological investigation,
by pathology. Much may be expected from
the work now in progress on aphasia. 1229 Meanwhile it
218is interesting to consider some of the difficulties which
occur in the normal use of language. Corresponding
to the hierarchies of interpretations described above we
have as many levels of possible failure. We may fail
to recognize a word qua sound, both when the word is
spoken to us and when we are about to utter it ourselves.
Secondly, although we are successful in this,
the context required for the understanding of a word
may lapse. This disturbance may be due either to
physiological, or, as the psycho-analysts have shown,
to emotional, interference. The failure may occur over
a name, and in such cases there is reason to suspect
emotional influence; or it may occur over a descriptive
phrase, or indeed any abstract symbol, in which case,
since many delicate adaptations to widely differing
experiences having only a slender common part are
involved, failure to discriminate this part is likely to be
accompanied by failure over the general abstract field. 1230
219Those periodical moments of stupidity to which we are
all prone, in which all abstract remarks appear pedantic
and incomprehensible, seem very often to be physiologically
determined.

Passing again to a higher level, there may be no
inability to understand those symbols which are components
of a complex symbol, and yet we may fail to
interpret the whole sentence. In this case we should
be said not to appreciate the logical form of the symbol.
Logical form might here be defined as what is common
to such complex symbols as “Crusoe landed from the
wreck,” and “Quixote fell off Rosinante,” where the
components 1231 may be subjected to a one for one substitution.
We have suggested above that the problem
of logical form requires further attention which it is not
likely to receive on current logical assumptions. It is
fatal to regard it as an ultimate notion, for what is
involved in interpreting a complex symbol is that the
contexts of the component symbols should, together
with the whole symbol, form a context of higher type.
All discursive symbolization involves this weaving
together of contexts into higher contexts, and interpretation
of such complex symbols is of the same nature
as that of simple symbols, with the difference only that
the members of these higher contexts are themselves
contexts. The same mechanisms of abstraction, metaphor,
etc., occur, and the same levels at which failure
is possible repeat themselves. Thus many people are
able to understand such a symbol as “The fire is hot,”
though baffled by predicative facts or if called upon to
consider relational attributes.

The study of the form or structure of complex references
220together with the form or structure of their
symbols is fundamental both for Logic and for what
is usually called grammar, which may be regarded
as the Natural History of symbol systems. This
science has, for obvious reasons, occupied the attention
of educationists and students of language, to the
detriment of more far-reaching inquiries. As normative,
grammar tends to confine itself to a verbal analysis
of How the King Talks, and, though sometimes
suggestive, applies no real critical apparatus. In particular
it is not realized that a Usage is only Good for
a given universe of discourse, and the ordering of these
different classes of occasions on which words may be
used has never been seriously approached.

A science which can justify itself as a discipline
imparting insight into the nature of the language
medium has at present no such status either with
instructors or instructed. The appointment, fashionable
in philological circles, of Standing Joint Committees,
to deal with the preliminaries of the science,
is an indication that it is still in the state which led
Smart to exclaim in 1831, “God help the poor children
who are set to learn the definitions in elementary
grammar.” But indeed the traditional problems of
grammar, the establishment of usage, the analysis of
sentences, the classification of the parts of speech, are
secondary problems of minor importance. They are
not open to investigation until the primary problem
of the nature of the language medium to which Symbolism
addresses itself has been explored. If this
fundamental investigation can be carried a very little
further it is probable that these later problems upon
which grammarians have lavished the treasures of
human industry and acumen, will be seen in some
cases to be purely artificial, in others to be concerned
with points of detail. 1232

The wider educational problems which concern the
221acquisition of language in infancy have frequently
received attention, and much useful material has been
amassed by Sully, Meumann, O'Shea, and Piaget;
but psychologists still make assumptions which prevent
any advantage accruing from the investigation.
“The Infant begins by imitating spoken words without
understanding them and then understands them,” says
Münsterberg. Fortunate infant to reach the second
stage! But unluckily the ingenuous little one does
nothing of the kind. Far more accurate is Rousseau's
view in his Thoughts on Education — “Inattention on
our part to the real way in which words are understood
by children appears to me the cause of their first errors;
and these, even when removed, have a great influence
on their turn of mind the remainder of their lives.”
The whole question of the acquisition and use of
language requires a fresh foundation, and must be
treated concretely with a view to the free development
of the interpretative faculties.

As an example of the kind of procedure which is
desirable, we may instance the ordering of the levels
at which, as we saw in Chapter IV. (p. 86), ‘chair,’
‘wood,’ ‘fibres,’ and so forth become correct symbols
for what we are perched upon. It was there pointed
out in what way the set of confusions known as metaphysics
has arisen through lack of this true grammatical
approach, the critical scrutiny of symbolic procedure. In
the same manner our analyses of Beauty and Meaning
are typical instances of what grammar might long ago
have achieved had grammarians only possessed a better
insight into the necessities of intelligent intercourse,
and a livelier sense of the practical importance of their
science. Preoccupied as is natural by the intricate
details of a vast subject-matter, and master of an imposing
technique and an elaborate semi-philosophic
nomenclature, the grammarian has unwittingly come
to stand somewhat fixedly in the way of those who
wish to approach the questions — How are words used?
222and, How should they be used? The grammarian also
is studying questions somewhat similar at first sight,
namely — Which words are used when? and, Which
should be used when? He resents the suggestion that
his work may be of small importance through his having
mistaken his question. In short, a normative examination
of words cannot be begun without a normative
examination of thinking, and no important question
of verbal usage can be considered without raising
questions as to the rank or level and the truth or falsity
of the actual references which may employ it. Symbols
cannot be studied apart from the references which they
symbolize and, this being admitted, there is no point
at which our examination of these references may stop
with safety, short of the fullest possible investigation.

Returning now to complexities in references and in
their symbols, the attempt to trace correspondence leads
to the adoption of two distinct sets of considerations as
guiding principles. With one of these, with the study
of reference, we have here been throughout concerned.
Symbolic form varies with variation of reference. But
there are other causes for its variation upon which we
have said something above (pp. 148-9). Besides symbolizing
a reference, our words also are signs of emotions,
attitudes, moods, the temper, interest or set of the mind
in which the references occur. They are signs in this
fashion because they are grouped with these attitudes
and interests in certain looser and tighter contexts.
Thus, in speaking a sentence we are giving rise to, as
in hearing it we are confronted by, at least two sign-situations.
One is interpreted from symbols to reference
and so to referent; the other is interpreted from
verbal signs to the attitude, mood, interest, purpose,
desire, and so forth of the speaker, and thence to the
situation, circumstances and conditions in which the
utterance is made.

The first of these is a symbol situation as this has
been described above, the second is merely a verbal
223sign-situation like the sign-situations involved in all
ordinary perception, weather prediction, etc. Confusion
between the two must be avoided, though they are
often hard to distinguish. Thus we may interpret
from a symbol to a reference and then take this reference
as a sign of an attitude in the speaker, either the
same or not the same as that to which we should
interpret directly from his utterance as a verbal sign.

The ordering of verbal sign-situations is a large
subject in which various branches may be distinguished.
The following seem, together with strict symbolization,
which it will be convenient to number as (i) , to
cover the main functions of language as a means of
communication.

(ii) There are the situations which derive from
attitudes, such as amity or hostility, of the speaker to
his audience. In written language many of the most
obvious signs for these attitudes 1233 are necessarily lost.
Manner and tone of voice have to be replaced by the
various devices, conventional formulæ, exaggerations,
under-statements, figures of speech, underlining, and
the rest familiar in the technique of letter-writing.
Word order is plainly of especial importance in this
connection, but, as we shall see, no general literary
device can be appropriated to any one of the functions
of speech, it is sure to be borrowed on occasion by the
others. Thus for this function almost any symbolic
224transformations can be brought in. For instance
telescoped or highly summarized phraseology is often
used, even where on referential grounds it is unsuitable,
as a mark of courtesy or respect to the hearer, or to
avoid the appearance of pedantry or condescension
which an expanded statement might produce. A speaker
will naturally address a large audience in terms different
from those which he employs in familiar conversation;
his attitude has changed.

(iii) In a similar fashion our attitude to our referent
in part determines the symbols we use. Here again
complicated cases occur in which it may be uncertain
whether our attitude is itself stated, or merely indicated
through verbal signs. Æsthetic judgments in particular
present this difficulty, and often the speaker himself
would be unable to decide which was taking place.
Emphasis, redundance, and all forms of reinforcement
can be, and are commonly, used for these reasons,
though equally they are used for the sake of their
effects upon hearers (iv); or as rallying-points, rests or
supports in case of difficulty of reference (v).

(iv) The structure of our symbols is often determined
by our Intention, the effects which we endeavour to
promote by our utterance. If we desire a hearer to
commit suicide we may, on occasion, make the same
remarks to him whether our reason for desiring such
action is benevolent interest in his career or a dislike of
his personal characteristics. Thus the symbol modification
due to the effect intended must not be confused
with that due to the attitude assumed towards an interlocutor,
although often, of course, they will coincide.

(v) Besides their truth, or falsity, references have a
character which may be called, from the accompanying
feelings, Ease or Difficulty. Two references to the
same referent may be true but differ widely in this ease,
a fact which may be reflected in their symbols. The
two symbols, “I seem to remember ascending Mount
Everest,” and “I went up Everest,” may, on occasion,
225stand for no difference in reference and thus owe their
dissimilarity solely to degrees of difficulty in recalling
this uncommon experience. On the other hand this
may, of course, be a real symbolic difference which does
not merely indicate difference of difficulty but states it.
This ease or difficulty should not be confused with
certainty or doubt, or degree of belief or disbelief, which
come most naturally under the heading (iii) of attitude
to the referent. Each of these non-symbolic functions
may employ words either in a symbolic capacity, to
attain the required end through the references produced
in the listener, or in a non-symbolic capacity when the
end is gained through the direct effects of the words.

If the reader will experiment with almost any
sentence he will find that the divergence which it shows
from a purely symbolic notation governed solely by
the nature of the reference which it symbolizes, will be
due to disturbing factors from one or more of the above
four groups. Further, what appears to be the same
difference will sometimes be due to one factor, at other
times to another. In other words, the plasticity of
speech material under symbolic conditions is less than
the plasticity of human attitudes, ends and endeavours,
i.e., of the affective-volitional system; and therefore the
same modifications in language are required for quite
different reasons and may be due to quite different
causes. Hence the importance of considering the
sentence in the paragraph, the paragraph in the chapter,
and the chapter in the volume, if our interpretations
are not to be misleading, and our analysis arbitrary.

It is, of course, not difficult to mention other factors
which modify the form or structure of symbols. A
hiccup, for instance, may do this, or laryngitis or
brachydactyly; so will the distance of the audience,
and more seriously the character of the occasion; or if
the speaker is excited or irritated for some extraneous
reason, his diction may show traces of this affect. The
whole past linguistic history both of the individual and
of the race to which he belongs obviously exercise
enormous influence; the Scot does not naturally talk
Yiddish. But all these influences upon linguistic form,
though the last is of paramount importance to the
comparative linguist, are not language functions in the
sense here considered. 1234 The state of the diaphragm,
of the throat, or of the fingers, the acoustics of a church
or a parade-ground are no concern of the Theory of
language; and although Comparative Philology has
often been regarded as in itself comprising the whole
field of the science, it is clear that this study belongs
essentially to history. In saying this we do not
minimize the interest and importance of the data which
227it provides. The functions we are examining are those
necessarily operative in all communication, the ways in
which the work of speech is performed, the essential
uses which speech serves.

Whether our list is exhaustive or not, it is at any
rate certain these functions cannot be reduced in
number without great loss of clarity and the omission
of considerations in many cases vital to the understanding
of the detail of language behaviour.

In translation, for example, the lack of such an
analysis of the ways in which words are used has led to
much confusion. Faced by the unaccountable failure
of apparently accurate renderings, linguists have been
too ready to accept the dicta of philosophers on this
point, as well as their vague vocabulary. Thus, according
to Sapir, “all the effects of the literary artist have
been calculated, or intuitively felt, with reference to the
formal ‘genius’ of his own language; they cannot be
carried over without loss or modification. Croce is
therefore perfectly right in saying that a work of literary
art can never be translated. Nevertheless, literature
does get itself translated, sometimes with astonishing
adequacy.” 1235 So a problem appears to arise, and as a
solution it is suggested that “in literature there are
intertwined two distinct kinds or levels of art — a generalized,
non-linguistic art, which can be transferred
without loss into an alien linguistic medium, and a
specifically linguistic art that is not transferable. I
believe the distinction is entirely valid, though we
never get the two levels pure in practice. Literature
moves in language as a medium, but that medium
comprises two layers, the latent content of language —
our intuitive record of experience — and the particular
conformation of a given language — the specific how of
our record of experience. Literature that draws its
sustenance mainly — never entirely — from the lower
level, say a play of Shakespeare's, is translatable
228without too great a loss of character. If it moves in
the upper rather than in the lower level — a fair example
is a lyric of Swinburne's — it is as good as untranslatable.”
And to illustrate this distinction, literature is
compared with science; a scientific truth is said to be
impersonal, “in its essence untinctured by the linguistic
medium in which it finds expression… Nevertheless
it must have some expression, and that expression
must needs be a linguistic one. Indeed, the apprehension
of the scientific truth is itself a linguistic process,
for thought is nothing but language denuded of its
outward garb.” Literature, on the other hand, is
“personal and concrete… The artist's intuition, to
use Croce's term, is immediately fashioned out of a
generalized human experience… Certain artists
whose spirit moves largely in the non-linguistic (better,
in the generalized linguistic layer), even find a certain
difficulty in getting themselves expressed.” Whitman
and others are supposed to be, as it were, “striving for
a generalized art language, a literary algebra…
Their art expression is frequently strained, it sounds
at times like a translation from an unknown original —
which indeed is precisely what it is.”

If we attempt to deal with the difficulties of translation
in terms of the ‘formal genius’ and ‘latent
content’ of the linguistic medium, and of the ‘non-linguistic
layer’ in which ‘intuition’ moves, mysteries
are inevitable. But a recognition of the richness of
the means at the disposal of poetry, with which we
shall shortly be concerned, allows us to dispense with
the doubtful assistance of the Neapolitan dialectic.
Translation, in fact, may succeed or fail for several
quite intelligible reasons. Any purely symbolic use of
words can be reproduced if in the two vocabularies
similar symbolic distinctions have been developed.
Otherwise periphrases or new symbols will be required,
and the degree of possible correspondence is a matter
which can be simply investigated. On the other hand,
229the more the emotive functions are involved the less
easy will be the task of blending several of these in two
vocabularies. And further, the greater the use made
in the original of the direct effects of words through
rhythm, vowel-quality, etc., the more difficult will it be
to secure similar effects in the same way in a different
sound-medium. Thus some equivalent method has to
be introduced, and this tends to disturb the other
functions so that what is called the ‘success’ of a translation
is often due chiefly to its own intrinsic merits.
With an understanding both of the functions of
language and of its technical resources the criticism of
translations provides a particularly fascinating and
instructive method of language study.

The view that speech on almost all occasions presents
a multiple, not a single, sign-situation throws a fresh
light upon many problems of traditional grammar. In
particular the treatment of sentence formation and syntax
will have to be undertaken afresh. From this point
of view we may note as typical a philologist 1236 content
with merely a dual language function in his definitions
of the word and the sentence.

A word is an articulate sound symbol in its aspect of
denoting something which is spoken about.

A sentence is an articulate sound symbol in its aspect
of embodying some volitional attitude of the speaker to the
listener.

Dr Gardiner's ‘volitional attitude’ would appear to
be included in No. IV of our list of functions. It will
be generally agreed that no use of speech can be admitted
to be an attempt at communication unless this
function is concerned.

The utility to grammarians of the terms so defined
is not obvious. What is of importance is the heterogeneity
which the author rightly insists upon between
the two functions of speech mentioned. The other
230functions which need to be considered in any comprehensive
analysis of Language are not less heterogeneous.
The charge is sometimes brought against writers on
psychology that they have neglected the side of the
listener. It is certainly true that preoccupation with
‘expression’ as the chief function of language 1237 has been
disastrous. But this is not so much because of the
neglect of the listener thereby induced as because of the
curiously narcotic effect of the word ‘expression’ itself.
There are certain terms in scientific discussion which
seem to make any advance impossible. They stupefy
and bewilder, yet in a way satisfy, the inquiring mind,
and though the despair of those who like to know what
they have said, are the delight of all whose main
concern with words is the avoidance of trouble. ‘Expression’
is such an one, ‘embody’ is another, and we
have just been concerned with ‘meaning’ in detail.
What is wanted is a searching inquiry into the processes
concealed by such terms, and as our analysis shows the
introduction of the listener does little to throw light upon
the matter. Moreover, psychologists and others, when
they have been concerned with the fact that Speech does
imply a listener, have not failed to insist upon the point.
Thus Dittrich, the holder of one of the few recognized
Chairs of the subject, wrote in 1900: “For linguistic
science it is fundamental that language is an affair not
merely of expression but also of impression, that communication
is of its essence, and that in its definition this
must not be overlooked.” He accordingly includes in
his own definition the words, “in so far as understanding
could be attempted by at least one other individual.” 2238
What such additional words contribute to a science
may be doubted; but it is certain that von Humboldt
went too far in this direction when he said: 3239 “Man
only understands himself when he has experimentally
tested the intelligibility of his words on others.”
231Steinthal's insistence on the part played by the listener
in the origin and development of language is also well
known; 1240 and de Saussure in his standard treatment of
speech functions which, as we saw in our first chapter,
was otherwise unsatisfactory, goes so far as to draw
pictures of the listener attending to the speaker and so
completing the ‘language circuit.’ 2241 A similar circuit
for volitional signs is diagrammatically completed by
Martinak through the fulfilment of the wish by the
listener; 3242 while Baldwin devotes over seventy pages of
the second volume of his Thought and Things to language
as affected by its functions in intercourse, and the
relations of speaker and listener in what he calls
“predication as elucidation” and “predication as
proposal.” 4243

But the most important practical recognition of the
fact that language has many functions is to be found in
Brunot's massive onslaught on current grammatical
procedure. 5244 Already, in 1903, the doyen of French
scholarship had convinced himself of the necessity of
abandoning the so-called ‘parts of speech,’either as a
method of approach or in actual teaching; and in 1908,
as Professor at the Sorbonne, he recorded this conviction
with clarity and vigour. For fifteen years, in ten
revisions, he worked over the debated ground: “After
each revision I returned to the same conclusion — that
no tinkering with the old scheme, no re-grouping of the
facts of language would be satisfactory so long as the
classification by parts of speech was retained. We must
make up our minds to devise methods of language-study
no longer drawn up on the basis of signs but on
the basis of ideas.” Unlike the majority of linguists,
Professor Brunot is fully aware that a purely psychological
analysis of the speech situation lies behind this
232functional approach to language, and it is interesting to
find that his exhaustive account of French idiom is in
accordance with the fivefold division proposed above.

We may now state the connection of reference to
symbol, subject to these disturbing factors, more
accurately. The reference of a symbol we see now,
is only one of a number of terms which are relevant
to the form of a symbol. It is not even the dominant
factor in most cases, and the more primitive the speech
which we investigate, the less important does it appear
to be. None the less, since, for all our finer dealings
with things not immediately present — i.e., not in very
close and simple contexts with our present experience —
since for all our more complicated or refined reference
we need supports and distinguishing marks, this strictly
symbolic function of words easily becomes more important
than any other. It is thus natural in an account
of the functions of words in ordinary usage to begin
with strict symbolization.

In the normal case not one, but a variety of symbol
forms is possible so far as the reference which they have
to accompany is concerned. The reference could be
accompanied let us say by A, or by B, or by C, or by
D; these being symbols of different forms or structure.
Any one of these is a possible member of the context
upon which the reference depends, in the sense that its
inclusion would not alter the reference. It is this range
of possible forms which enables the symbol to perform
so many different services, to be a sign in so many
distinct though contemporary situations.

Suppose now the speaker, in addition to referring,
assumes some attitude towards his audience, let us say
amity. Then among these symbolic forms A, B, C, D,
there may be one, say D, which is more suitable to the
special shade of this attitude than the others, in the
sense that it is a possible member of the context of the
attitude, one of that group of symbols whose utterance
would not alter the attitude. If this were all that were
233involved D would be uttered, since any other suitable
remark would presumably involve some change in the
reference.

Suppose further that the speaker feels, let us say,
disgust, towards his referent. This will lead in similar
fashion to further modification of the symbol. So again
will the speaker's hopes, desires and intentions with
regard to the effects of his remarks. Often the same
modifications will satisfy both these conditions, but
sometimes, when for instance the speaker's own attitude
and that which he wishes to promote are for any
reason discrepant, the natural word-attitude contexts
must lapse, and judicious symbolization becomes for
some people more difficult. In a similar fashion the
speaker's own clearness or vagueness in reference has
often to be disguised or to submit to compromise. His
certainty or uncertainty, his doubt or degree of belief
may as we have above remarked, be best ranked with
general attitudes to referents.

Most writing or speech then which is of the mixed
or rhetorical kind as opposed to the pure, or scientific,
or strictly symbolic, use of words, will take its form as
the result of compromise. Only occasionally will a
symbolization be available which, without loss of its
symbolic accuracy, is also suitable (to the author's attitude
to his public), appropriate (to his referent), judicious
(likely to produce the desired effects) and personal (indicative
of the stability or instability of his references).
The odds are very strongly against there being many
symbols able to do so much. As a consequence in most
speech some of these functions are sacrificed. In ‘good
morning’ and ‘good-bye’ the referential function lapses,
i.e., these verbal signs are not symbols, it is enough
if they are suitable. Exclamations and oaths similarly
are not symbols; they have only to satisfy the condition
of appropriateness, one of the easiest of conditions at
the low-level of subtlety to which these emotional signs
are developed. The only contexts required here would
234seem to be of the simplest order possible in psychology,
as simple as the toothache-groan context. Orders or
commands must satisfy reference and purpose conditions,
but may, indeed often must, avoid both suitability and
appropriateness in the senses used above, as for instance
in many military orders. Threats on the other hand
can easily dispense with reference, i.e., be meaningless,
and may be governed only by the purpose intended.
Questions and requests are similar to commands in the
respects above mentioned and differ from them merely
in the means through which the effects desired are
sought.

These instances of the dropping of one or more of
the language functions lead us naturally to the most
remarkable and most discussed case of such variation,
the distinction, namely, between the prose and the
poetic uses of language. In these terms the distinction
is not happily symbolized, poetry being best defined
for the most general and most important purposes by
relation to the state or states of mind produced by the
‘poem’ in suitable readers and without any relation to
the precise verbal means. Instead therefore of an
antithesis of prose and poetry we may substitute that
of symbolic and emotive uses of language. In strict
symbolic language the emotional effects of the words
whether direct or indirect are irrelevant to their employment.
In evocative language on the other hand all
the means by which attitudes, moods, desires, feelings,
emotions can be verbally incited in an audience are
concerned. We have already discussed at some length
(p. 159) the importance of distinguishing between these
two uses of language, and we may here add a few
further considerations dealing with the means by which
evocative languages secure their effects.

These accessory effects of words have often been
described by men of letters, without much having been
done towards their detailed study. Lafcadio Hearn, for
instance, writes that for him “words have colour, form
235character. They have faces, ports, manners, gesticulations
: they have moods, humours, eccentricities: they
have tints, tones, personalities. I write for beloved
friends who can see colour in words, can smell the
perfume of syllables in blossom, can be shocked with
the fine elfish eccentricity of words. And in the
eternal order of things, words will eventually have their
rights recognized by the people.”

Words or arrangements of words evoke attitudes both
directly as sounds, and less directly in several different
ways through what are called loosely ‘associations.’
The effects of the words due directly (i.e., physiologically)
to their sound qualities are probably slight and
only become important through such cumulative and
hypnotic effects as are produced through rhythm and
rhyme. More important are the immediate emotional
accompaniments due to past experience of them in their
typical connections. To get these, there is no need for
the connections themselves to be recalled. Thirdly
there are the effects ordinarily alluded to as the emotions
due to associations, which arise through the recall of
whole situations. So far we have confined our attention
to verbal languages, but the same distinction and the
same diversity of function arise with non-verbal
languages. When we look at a picture, as when we
read a poem, we can take up one or both of two attitudes.
We can submit to it as a stimulus, letting its colour-qualities
and form-qualities work upon us emotionally.
Or with a different attitude we can interpret its forms
and colours (its words). The first of these attitudes
is not an indispensable preliminary to the second.
To suppose so would be to mistake the distinction.
Mr Clive Bell has performed a useful service in pointing
out that many people are accustomed to pass, in
the case of pictures, to the second of these attitudes,
omitting the first entirely. Such omission, of course,
deprives the picture of its chief part. Professor Saintsbury
has performed a similar service for hasty readers.236

But although the first of these attitudes, submission
to the work of art as a stimulus, is in need of encouragement,
the second attitude, that of interpretation, is
equally necessary. At this point both critics become
over zealous for an aspect of the truth. After allowing
pure forms to affect us, we must, in most cases, go on
to interpret if we are to allow the picture or poem to
produce its full result. In so doing, there are two
dangers which good sense will avoid. One is the
danger of personal associations, concerning which
nothing need be said. The other is the danger of confusing
the evocation of an attitude towards a situation
with the scientific description of it. The difference
between these very different uses of language is most
clearly apparent in the case of words. But all that we
have said will apply equally to the contrast between
art and photography. It is the difference between the
presentation of an object which makes use of the direct
emotional disturbances produced by certain arrangements,
to reinstate the whole situation of seeing, or
hearing, the object, together with the emotions felt
towards it, and on the other hand, a presentation which
is purely scientific, i.e., symbolic. The attitude evoked
need not necessarily be directed towards the objects
stated as means of evoking it, but is often a more
general adjustment. It will make these distinctions
more plain if we consider them in the closely analogous
field of painting, where emotions do not enter in different
ways but only with an increased difference and distinction
between them in accordance with the ways by
which they enter. Exactly as we may distinguish the
direct emotional effects of sound qualities and stresses,
so we can distinguish the similar direct effects of colour
and form. Just as, for instance, vowel and consonantal
quality may conflict with rhythm, so colour may conflict
with form: that is, they may evoke incongruous
emotions. Similarly, it is admitted that colours acquire
emotional effects through experience, emotional effects
237which are not the emotional effects of their associations.
An Eskimo and a Moor, for instance, are differently
affected by English colouring, because different
selections of it are familiar, quite apart from association.

The complexities and ambiguities in the use of
238language for purposes of evocation are admittedly not
less than those from which scientific language suffers.
But when two people differ in what they are in ordinary
usage perfectly correct in calling “their interpretations”
of a poem or a picture, the procedure to be adopted is
quite other than that advisable should they differ in
their interpretations of a physicist's remarks. None the
less, there is, in the two cases, an underlying similarity
due to the fact that both are sign-situations though only
the second is symbolic in the strict sense of the term.

The difference between the two uses may be more
exactly characterized as follows: In symbolic speech
the essential considerations are the correctness of the
symbolization and the truth of the references. In
evocative speech the essential consideration is the
character of the attitude aroused. Symbolic statements
may indeed be used as a means of evoking attitudes,
but when this use is occurring it will be noticed that the
truth or falsity of the statements is of no consequence
provided that they are accepted by the hearer.

The means by which words may evoke feelings and
attitudes are many and offer an alluring field of study
to the literary psychologist. As sounds, and again as
movements of articulation, and also through many
subtle networks of association, the contexts of their
occurrences in the past, they can play very directly upon
the organized impulses of the affective-volitional systems.
But above all these in importance, heightening and
controlling and uniting these subordinate influences, are
the rhythmic and metrical effects of word arrangements.
If, as may reasonably be supposed, rhythms and especially
metres have to a small degree an hypnotic effect,
the very marked difference in evocative power between
words so arranged and words without recurrent system
is readily accounted for. Some degree of hyperæsthesia
would be a convenient assumption to explain further
the greater sensitiveness to vowel and consonantal
characters which accompanies metrical reading, and the
239flat or tinny effect of the same syllables occurring in
vers libres. Emotionality, exaggeration or belief-feelings,
the occulting of the critical faculties, the suppression of
the questioning — ‘Is this so as a matter of fact?’ —
attitude, all these are characteristics of metrical experiences
and fit in well with a hypnosis assumption. When
we add to these effects of metre, its powers of indirect
representation (as the words ‘swinging’, ‘rolling’,
‘heavy’, ‘rushing’, ‘broken’, applied to rhythms
indicate) its powers of directly controlling emotions (as
the words ‘lulling’, ‘stirring’, ‘solemn’, ‘gay’ indicate)
and its powers of unification (as at a low level its use as
a mere mnemonic shows), we shall not be surprised to
find it so extensively present in the evocative use of
speech.

The indirect means of arousal which are possible
through words need not be dwelt upon here at length.
Through statement; through the excitement of imagery
(often effected at low levels of refinement by the use of
metaphor); through metaphor itself — used not, as in
strict symbolizing, to bring out or stress a structural
feature in a reference, but rather to provide, often under
cover of a pretence of this elucidation, new sudden and
striking collocations of references for the sake of the
compound effects of contrast, conflict, harmony, interinanimation
and equilibrium which may be so attained,
or used more simply to modify and adjust emotional
tone; through association; through revival; and
through many subtle linkings of mnemic situations,
words are capable of exerting profound influence quite
apart from any assistance from the particular passions,
needs, desires or circumstances of the hearer. With
the further aid of these there is, as has often been illusrated
in history, no limit to their evocative range.

The characteristic feature of these forms of evocation
which occur in the arts, where severance from such
personal particular circumstances is necessary for the
sake of universality, is the constant mingling of direct
240and indirect means. The neglect or underestimation
of the direct means available in poetry is, however,
common in those who do not use the medium, and has
often led to attempts to exclude poetry from the arts
on the ground that its appeal is indirect only, through
ideas, and not sensory in character. This contention
is due merely to ignorance.

It is unfortunately very necessary to insist upon
the importance of the distinction between these two
functions of speech. Confusion between them leads
to wrangles in which Intellect and Emotion, Reason
and Feeling, Logic and Intuition, are set in artificial
opposition to one another: though as is easily perceived,
these two functions need not in any way trespass
upon one another's provinces. 1245 None the less, analogous
sets of recording symbols have developed for each use —
a Truth, Reality and Universality for symbolic speech
and a Truth, Reality and Universality for evocative
speech. This formal parallelism is very misleading,
since the words Truth^S and Truth^E are totally distinct
as symbols, the first being defined in terms of reference,
while the second is equivalent to appropriate and
genuine, and does not involve reference. It is unfortunate
that devotees of literature should so often pass
their whole active mental existence under the impression
that through their antitheses of Intuition and Logic in
this field they are contemplating a fundamental issue.

The chaos to which uncritical reliance upon speech
has reduced this topic, together with so many others
which rightly arouse intense interest, is by itself a
powerful argument for the prosecution of the inquiry
into Symbolism. When we remember the fruitless
questionings and bewilderment caused by the irrelevancies
and the intrinsic peculiarities of words, not only
to children but to all who endeavour to pass beyond the
mere exchange of accepted and familiar references, we
241shall not be tempted to think that the proposal seriously
to investigate language must be either a joke or a form
of pedantry — as those do who, having never been
troubled by thought, have never found any difficulty
in expressing it. The view that language gives rise
to no such difficulties can be dispelled for all intelligent
people either by observation or by personal experience.
The opposite view that the difficulties are too formidable
to be overcome, though more worthy of the human
mind, must be rejected for similar reasons. What
language already does, is the ground for hope that
it may in time be made fully to perform its functions.
To this end the Theory of Signs and Education must
co-operate. No formal apparatus of Canons and Rules,
no demands that abuses of language shall be reformed,
will take effect, unless the habits which will enable
language to be freely used are developed. What is
required is not only strictness of definition and rigidity
of expression, but also plasticity, ease and freedom in
rapid expansion when expansion is needed. These
abilities can only be developed through the training
which at present is devoted to matters for whose
understanding an adequate language is a prerequisite.

A new Science, the Science of Symbolism, is now
ready to emerge, and with it will come a new educational
technique. Language is the most important
instrument we possess. At present we attempt to
acquire and to impart a knowledge of its use by
mimicry, by intuition, or by rule of thumb, in contented
ignorance of its nature. It is not by his own efforts
that the modern child is in so many ways better equipped
than Aristotle; for such improvement must be the result
of co-operative endeavour. Those who are not satisfied
by the solutions of linguistic problems offered in these
pages will, it is to be hoped, discover better. If, however,
our claim to have provided a new orientation is
a just one, the far-reaching practical results which we
have discussed are already capable of attainment.242

Summary

At the close of a long discussion involving the
detailed examination of many separate problems,
elaborate examples of the application of method,
historical illustrations and special criticisms of vicious
tendencies, a brief outline of the main topics dealt with
is desirable in order to give a general impression of
the scope and task of the Science of Symbolism. Only
by excluding all allusion to many subjects not less important
than those here mentioned, can we avoid the
loss of perspective inevitably entailed by the list of
Contents to which the reader is referred.

1. — Thoughts, Words and Things.

The influence of language upon Thought is of the
utmost importance. Symbolism is the study of this
influence, which is as powerful in connection with everyday
life as in the most abstruse speculation.

The chief of these assumptions derives from the
magical theory of the name as part of the thing, the
theory of an inherent connection between symbols and
243referents. This legacy leads in practice to the search
for the meaning of words. The eradication of this
habit can only be achieved by a study of Signs in general,
leading up to a referential theory of Definition by which
the phantom problems resulting from such superstitions
may be avoided. When these have been disposed of,
all subjects become more accessible and more interesting.

2. — The Power of Words.

The magic of words has a special place in general
magic. Unless we realize what have been the natural
attitudes towards words until recent years we shall fail
to understand much in the behaviour of logicians and
others among modern mystics, for these same attitudes
still persist in underground and unavowed fashion. At
the same time the theory of signs can throw light upon
the origins of these magical beliefs and their persistence.

3. — Sign-situations.

In all thinking we are interpreting signs.
In obvious cases this is readily admitted. In the
more complex cases of mathematics and grammar more
complicated forms of the same activity only are involved.
This is hidden from us by an uncritical use of symbols,
favouring analyses of ‘meaning’ and ‘thinking’ which
are mainly occupied with mirages due to ‘linguistic
refraction.’

We must begin therefore with Interpretation.

Our Interpretation of any sign is our psychological
reaction to it, as determined by our past
experience in similar situations, and by our present
experience.

If this is stated with due care in terms of causal
contexts or correlated groups we get an account of
judgment, belief and interpretation which places the
psychology of thinking on the same level as the other
244inductive sciences, and incidentally disposes of the
‘Problem of Truth.’

A theory of thinking which discards mystical relations
between the knower and the known and treats
knowledge as a causal affair open to ordinary scientific
investigation, is one which will appeal to commonsense
inquirers.

Sign-situations are always linked in chains and the
simplest case of such a sign-chain is best studied in
Perception.

4. — Signs in Perception.

The certainty of our knowledge of the external
world has suffered much at the hands of philosophers
through the lack of a theory, of signs, and through
conundrums made possible by our habit of naming
things in haste without providing methods of identification.

The paradoxes of really round pennies which appear
elliptical, and so forth, are due to misuses of symbols;
principally of the symbol ‘datum.’

What we ‘see’ when we look at a table is first,
modifications of our retinas. These are our initial signs.
We interpret them and arrive at fields of vision, bounded
by surfaces of tables and the like. By taking beliefs in
these as second order signs and so on, we can proceed
with our interpretation, reaching as results tables, wood,
fibres, cells, molecules, atoms, electrons, etc. The later
stages of this interpretative effort are physics. Thus
there is no study called ‘philosophy’ which can add to
or correct physics, though symbolism may contribute
to a systematization of the levels of discourse at which
‘table’ and ‘system of molecules’ are the appropriate
symbols.

The method by which confusions are to be extirpated
in this field is required wherever philosophy has been
245applied. It rests partly upon the theory of signs, partly
upon the Rules of Symbolization discussed in the next
chapter.

5. — The Canons of Symbolism.

Underlying all communication, and equally fundamental
for any account of scientific method, are the
rules or conventions of symbolism.

Some of these are obvious enough when stated, but,
perhaps for this reason, have been generally neglected.
Others have been cursorily stated by logicians, concerned
hitherto with a narrow range of traditional
problems. When, however, all are fully set forth in
the forms implied by systematic discourse, the solutions
of many long-standing problems are found to be de facto
provided.

Examples of such problems are those of Truth,
Reality, Universals, Abstractions, Negative Facts,
Virtuous Triangles, Round-squares and so forth.

The rules or postulates in question which most need
formulation are Six in number, and appear as the
Canons of Symbolism. They derive from the nature of
mental processes, but, being required for the control
of symbolization, are stated in terms of symbols and
referents.

The observance of these Canons ensures a clear
prose style, though not necessarily one intelligible to
men of letters.

6. — Definition.

In any discussion or interpretation of symbols we
need a means of identifying referents. The reply to
the question what any word or symbol refers to consists
in the substitution of a symbol or symbols which can
be better understood.

Such substitution is Definition. It involves the
selection of known referents as starting-points, and
246the identification of the definiendum by its connection
with these.

The defining routes, the relations most commonly
used for this purpose, are few in number, though
specialists in abstract thought can employ others. In
fact they may be pragmatically generalized under eight
headings. Familiarity with these defining routes not
only conduces to ease of deportment in reasoning and
argument, but offers a means of escape from the maze
of verbal cross-classifications which the great variety
of possible view-points has produced.

7. — The Meaning of Beauty.

The application of this procedure in practice may be
demonstrated by taking one of the most bewildering
subjects of discussion, namely Esthetics.

Beauty has been very often and very differently
defined — and as often declared to be indefinable. If,
however, we look for the characteristic defining relations,
we find that the definitions hitherto suggested reduce
conveniently to sixteen.

Each of these then provides a distinct range of
referents, and any such range may be studied by those
whom it attracts. If in spite of the disconcerting
ambiguity thus revealed (and all freely-used terms
are liable to similar ambiguity) we elect to continue
to employ the term Beauty as a shorthand substitute
for the definition we favour, we shall do so only on
grounds of ethics and expediency and at the risk of all
the confusions to which such behaviour must give rise.

In addition to its symbolic uses ‘Beauty’ has also
its emotive uses. These have often been responsible
for the view that Beauty is indefinable, since as an
emotive term it allows of no satisfactory verbal substitute.
Failure to distinguish between the symbolic
and emotive uses is the source of much confusion in
discussion and research.247

8. — The Meaning of Philosophers.

Proceeding on the same principles to ‘Meaning’
itself, we find a widely divergent set of opinions in the
writings of the best philosophers. The recent discussions
in Mind and in Brain show the helplessness
of expert disputants in dealing with the resultant
ambiguities of the term. The procedure of the ables
and most practical group of American thinkers, the
Critical Realists of 1921, reveals the same incompetence,
while the use made of the word by so influential
an authority as Professor Münsterberg is equally open
to objection. In fact, a careful study of the practice
of prominent writers of all schools leads to the conclusion
that in spite of a tacit assumption that the
term is sufficiently understood, no principle governs
its usage, nor does any technique exist whereby
confusion may be avoided.

9. — The Meaning of Meaning.

When, however, the problem is scientifically approached,
we find that no less than sixteen groups o
definitions may be profitably distinguished in a field
where the most rigid accuracy is desirable.

In other cases ambiguity may be fatal to the particular
topic in which it occurs, but here such ambiguity
even renders it doubtful what discussion itself is. For
some view of ‘meaning’ is presupposed by every
opinion upon anything, and an actual change of view
on this point will for a consistent thinker involve change
in all his views.

The definitions of Meaning may be dealt with under
three headings. The first comprises Phantoms linguistically
generated; the second groups and distinguishes
Occasional and erratic usages; the third
covers Sign and Symbol situations generally.

One interesting effect of such an exposition is that
248it forces us for the time being to abandon the term
‘meaning’ itself, and to substitute either other terms,
such as ‘intention,’ ‘value,’ ‘referent,’ ‘emotion’ for
which it is being used as a synonym, or the expanded
symbol which, contrary to expectation, emerges after
a little trouble.

A careful study of these expansions leaves little
room for doubt that what philosophers and metaphysicians
have long regarded as an abstruse and
ultimate notion, falling entirely within their peculiar
domain and that of such descriptive psychologists as
had agreed to adopt a similar terminology, has been
the subject of detailed study and analysis by various
special sciences for over half a century. During the
last few years advances of biology, and the physiological
investigation of memory and heredity have placed the
‘meaning’ of signs in general beyond doubt, and it
is here shown that thought and language are to be
treated in the same manner.

10. — Symbol situations.

The first stage of the Development of Symbolism as
a Science is thus complete, and it is seen to be the
essential preliminary to all other sciences. Together
with such portions of grammar and logic as it does
not render superfluous it must provide both what has
been covered by the title Philosophy of Mathematics,
and what has hitherto been regarded as Meta-physics — supplementing
the work of the scientist at either end
of his inquiry.

All critical interpretation of Symbols requires an
understanding of the Symbol situation, and here the
main distinction is that between the condition in which
reference is made possible only by symbols (Worddependence)
and that for which a free choice of symbols
can be made (Word-freedom). The examination of
language processes in their perfection or in their
249degeneration must also start from this distinction. It
is further important to notice that words have further
functions in addition to that of strict symbolization.
The study of these evocative aspects leads naturally to
an account of the resources of poetical language and
of the means by which it may be distinguished, from
symbolic or scientific statement. Thus the technique
of Symbolism is one of the essential instruments of the
æsthetics of literature.

Its practical importance will be found in its application
to Education and to Discussion in general; for
when the Influence of Language upon Thought is
understood, and the Phantoms due to linguistic misconception
have been removed, the way is open to
more fruitful methods of Interpretation and to an
Art of Conversation by which the communicants can
enjoy something more than the customary stones and
scorpions.250

Appendices

Appendix A
On Grammar

“Incomprehensible abstractions, pretentious yet for the most
part empty definitions, false rules, indigestible lists of forms,
one has only to turn over a few pages of any text book to find
variegated specimens of these sins against reason, truth and
education.” These are strong words in which to condemn
the bulk of modern grammatical teaching, but, as we have seen
above in Chapter X. (p. 232), Professor Brunot, after fifteen years
of further work on linguistic analysis since their publication, 1246
found no reason to modify them. Considering the medley of
verbal superstition, obsolete philosophy, and ill-comprehended
logic, which we have found in the course of these pages doing
duty for a theory of verbal function, it is not surprising that the
best-informed philologists should feel that no words can be too
strong for the grammatical fare on which the twentieth-century
child is still nourished.

His final comment is: “A profound pity overcomes one in
thinking of the hundreds of thousands of children compelled to
undergo an education composed of such aberrations.” 1247

It is with a view to the elimination of the most patent of these
absurdities that the various Committees on Grammatical Terminology
have been labouring in various countries since the
1906 conferences at the Musée pédagogique in Paris. The
Recommendations of the English Committee were issued in
1911, and efforts are now being made by the various Language
Associations to have them applied. In such an application,
however, two distinct problems are involved. One is the elimination
of outstanding absurdities in a grammatical terminology
for any one language; and as to the desirability of a reformed
terminology and the value of the work, of the Committee in this
respect, as far as it goes, there is little room for controversy. The
other concerns “the importance of adopting from the first, in
all grammar teaching, a terminology which should be capable of
being employed, with the minimum of variation, for the purposes
of any other language that is subsequently learnt.” 2248 It is true
that “a uniform terminology brings into relief the principles of
structure common to all allied languages; needless variation of
terms conceals the substantial unity,” 3249 but it must be remembered
that insistence on supposed similarities of structure by Indo-European
grammarians has been a chief hindrance to ethnologists
in their study of primitive speech, that most vitally important
branch of their subject. Within such a group of languages as that
to which English belongs it is useful to have a system to mark
similarities, 4250 but there is always the risk that the uniformity
252thus stressed may come to be regarded as a necessity of all
language, and indeed, of thought itself. It is then natural for
these alleged necessities of expression to appear as reflections of
the actual nature of the things spoken about themselves.

It is with the last of these four questions that we are here concerned
and, whether with a full sense of its implications or not,
the procedure of grammarians — in their treatment of subject
and predicate, for instance — has often seemed tacitly to assume
Wittgenstein's answer: “To the configuration of the simple
signs in the propositional sign corresponds the configuration of
the objects in the state of affairs.” 1 251This unplausible conclusion
rests on the arbitrary identification of the indirect relation
‘standing for,’ discussed in our first chapter, with representation.
“In order to be a picture a fact must have something in common
with what it pictures” runs Prop. 2.16, and further 2.171, “The
picture can represent every reality whose form it has… 2.182,
Every picture is also a logical picture… 3, The logical
picture of the facts is the thought… 3.1, In the proposition
the thought is expressed perceptibly through the senses…
3.12, The sign through which we express the thought I call the
253propositional sign. 3.2, In propositions thoughts can be so expressed
that to the objects of the thoughts correspond the elements
of the propositional sign.” If every word must here be
understood in a special sense, such an account of a symbol
situation resembles the pronouncements of the Pre-Socratic
aphorists; yet to call it a ‘logical’ and not a psychological account
is, on the whole, an unconvincing apologetic.

Two steps are made in this argument. The first purports to
secure a common structure in thoughts and things in order to
explain how a thought can be ‘of’ a thing. But on a causal
theory this assumption of correspondence in structure is unnecessary
and highly improbable. 1252 The second step, the assertion
of correspondence between the structure of the propositional
sign and the structure of the facts is even more bold and baseless.
In a simple case, as when we make diagrams and in such notations
as those of chemistry and music, we can no doubt secure some
degree of correspondence, because, as was pointed out in the
chapter cited, the elements of such mimetic language approximate
to simple signs. In the case of notations, it has been the deliberate
effort of generations of scientists to force their symbols into
simple correspondence with the things for which they are to
stand. Again, in any primitive tongue there may come a time
when, through the simplicity of the distinctions made by the
race amongst the things surrounding them, their language will
show an analogous set of distinctions. Here, however, the correspondence
is through the correspondence of references to
things and of kinds of words to kinds of references. But it is
plain that such a language cannot keep pace with the additional
distinctions in their thought and with its growing complexity.
New kinds of words and new verbal structures would be desirable
for new aspects and structures which they wish to distinguish.
The old machinery, therefore, has to be strained and recourse is
254had to fictitious entities, due to linguistic elements and structures
no longer fulfilling their proper function but inadequately serving
purposes for which they were not originally developed. Thus
‘Energy’ in modern physics seems to be the wrong kind of word
for the referents concerned, and no other word belonging to any
of the recognized grammatical categories is likely to be better
fitted. Hence some difficulties of the Quantum theory.

The attempt to generalize from the exceptional cases in which
symbols and referents partially correspond, to a necessity for
such correspondence in all communication is invalid. The extent
of the correspondence in any given case can only be settled by an
empirical inquiry; but the result of such an inquiry is not
doubtful. Such a correspondence may give to scientific symbol
systems vastly increased scope and accuracy, and render them
amenable to deductive processes; but it can only be imposed
when limited to the simplest and most schematic features, such
as number or spatial relations. Ordinary language, as a rule,
dispenses with it, losing in accuracy but gaining in plasticity,
facility, and convenience. Nor is the loss so great as is sometimes
supposed, for by straining language we are able to make and
communicate references successfully, in spite of the misleading
character of our symbols if taken literally. 1253 For some, such as
Wittgenstein himself, the possibility of this correspondence and
the impossibility of doing more leads to a dissatisfaction with
language; and to an anti-metaphysical mysticism. For others,
such as Bergson, 2254 the alleged impossibility of this correspondence
255based upon the assumed nature of reality, leads to a different
kind of dissatisfaction; and to a mystical metaphysics.

For the grammarian these ultimate issues may appear to be
remote, but none the less he cannot have a view upon the relations
of language with fact, or a basis for the discussion of true linguistic
function in the sense denned in Chapter X. (which is, of
course, different from the functions of words in sentence formation)
without raising these issues.

We may consider, as a typical instance of a language function
which has been supposed to be derived from a fundamental
feature of reality, and to be capable of direct treatment by common
sense without resort to a theory of reference, the problem
of the proposition and the subject-predicate relation… Since
all traditional views on this matter derive from Aristotle it is
worth while to recall the way in which it was first approached.
What is signified for Aristotle by words (whether single or in
combination), says his clearest modern exponent, is some variety
of mental affections 1255 “or of the facts which they represent.
But the signification of a term is distinguished in an important
point from the signification of that conjunction of terms which we
call a Proposition. A noun, or a verb, belonging to the aggregate
called a language, is associated with one and the same phantasm
or notion, without any conscious act of conjunction or disjunction,
in the minds of speakers and hearers: when pronounced, it
arrests for a certain time the flow of associated ideas, and determines
the mind to dwell upon that particular group which is
called its meaning. But neither the noun nor the verb, singly
taken, does more than this; neither one of them affirms, or denies,
or communicates any information true or false. For this last
purpose, we must conjoin the two together in a certain way, and
make a Proposition. The signification of the Proposition is
thus specifically distinct from that of either of its two component
elements. It communicates what purports to be matter of fact,
which may be either true or false; in other words, it implies in
the speaker, and raises in the hearer, the. state of belief or disbelief,
which does not attach either to the noun or to the verb
separately. Herein the Proposition is discriminated from other
significant arrangements of words (precative, interrogative,
which convey no truth or falsehood), as well as from its own component
256parts. Each of these parts, noun and verb, has a significance
of its own; but these are the ultimate elements of speech,
for the parts of the noun or of the verb have no significance at all.” 1256

In this statement may be found all the uncertainty and hesitation
which since Aristotle's time have beset both grammarians
and logicians. Notably the doubt whether words signify ‘mental
affections’ or the facts which these ‘represent,’ and the confusion
between the assertive character of the proposition (which is here
used as equivalent to sentence) and the states of belief and disbelief
which may occur in connection with it.

With the first source of confusion we have dealt at length, but
the second demands further attention if it is to be avoided.
Recent psychological research, especially into the nature of suggestion
and into the effects of drugs upon the feelings, has done
nothing to invalidate William James' view as to the relation of
belief to reference. “In its inner nature, belief, or the sense of
reality, is a sort of feeling more allied to the emotions than to
anything else.” Belief and disbelief as opposed to doubt are
“characterized by repose on the purely intellectual side,” and
“intimately connected with subsequent practical activity.” 2257
Belief and disbelief, doubt and questioning, seem to be what
nowadays would be called affective-volitional characteristics of
mental states, and thus theoretically separable from the states to
which they attach. The same reference, that is to say, may at one
time be accompanied by belief and at another by disbelief or
doubt. For this reason, so far as language is modified by the
nature of the belief-feelings present, these modifications come
under the heading of expression of attitude to referent, the third
of the language functions distinguished in Chapter X.

This separation greatly assists a clear analysis of the most
important character of the proposition, namely, the way in which
it seems to symbolize assertion, to stand for a complete object of
thought, a character lacking to the parts of a simple sentence.
A noun by itself or a verb by itself somehow differs from the
whole which is made up when they are suitably juxtaposed,
and this difference has been the pivotal point upon which not
merely grammatical analysis, but logic and philosophy have also
turned ever since Aristotle s time.

The confusion has been further aggravated by the introduction
of the problem of truth in an unsolved condition. Propositions
have been almost universally regarded as the only objects to
257which the words ‘true’ and ‘false’ are properly applied; though
this unanimity has been somewhat disguised by differences of
view as to whether true propositions are those which express
true beliefs or whether true beliefs are those whose objects are
true propositions. In these controversies the various shifts of
the symbol ‘proposition’, standing, as it does, at one time for
a sentence, at another for a referent, and yet another for a relational
character of a mental act or process, provide a fascinating
field for the Science of Symbolism to explore. But in view of
what has been said above in Chapter III. on the analysis of the
differences which distinguish a complex symbol such as ‘Snow
chills’ from the single symbols such as ‘snow’ and ‘chills’
which compose it, the apparent complications due to the introduction
of truth raise no difficulty. They are merely a bewildering,
because imperfectly parallel, re-naming of the problem.

According to the theory of signs all references, no matter how
simple they may be, are either true or false, and no difference in
this respect is to be found between the reference symbolized by
‘snow’ and that symbolized by ‘snow chills.’ This statement
requires to be guarded from over-hasty interpretation. It is
easy to use single words in such a way that they are not symbols,
and so do not stand for anything. When this is done no doubt
some stray images and other mental goings-on may be aroused,
and if we are not careful in our use of ‘meaning’ we may then
suppose that non-symbolic words so considered have meaning
just as much and in the same sense as they do when present
symbolically in a proposition. The single word, whether noun
or verb, only has meaning in the sense here required, when taken
in such a way that it enters a reference contest of the normal
kind; and only so taken is it a symbolic (as distinguished from
an emotive) component of a proposition. Any word so considered
comes to be, qua symbol of a reference to some state of affairs,
capable of truth and falsehood; and in this respect it differs in
no way from a sentence used symbolically for purposes of statement.

We have yet to see, therefore, in what the marked difference
between single words and sentences consists; and, as we should
expect from the nature of the symbol situation, we find the
difference to be not one but several, none of which is always or
necessarily present although some may be said to be normally
involved. 1258 In the first place the references of the symbols will
258often differ structurally. Thus the reference of ‘larks sing,’
since it has two components, will differ from that of ‘larks’
just as do ‘soaring larks’ or ‘lark pie,’ being also dual references.
This difference is therefore unessential, though most complex
references do in fact use the prepositional form. One reason
for the use of this form is because it is the normal means by
which the togetherness of the component references is symbolized
in cases where ambiguity is possible. Thus the sentence is the
chief, but not the only symbolic device by which the togetherness
of references is made plain. It is this which is usually described
as the ‘synthetic’ function of the proposition, 1259 an unsatisfactory
term, since verbal arrangements which are not of the prepositional
form, such as ‘lark pie,’ or ‘this lark pie’ 2260 — are equally synthetic.
In logic the translation of all propositions into the subject-copula-predicate
form has been a convention to avoid ambiguity,
though modern logicians have found that more elaborate conventions
are desirable for relational propositions.

But the sentence also serves emotively in various ways. 3261 It is
the conventional mode of Address, since listeners expect some
special signal that a reference is occurring before they incline
their ears cognitively. Further, it is the conventional verbal
sign of the presence of Belief, of feelings of acceptance, rejection
or doubt, in the speaker; and a stimulus to similar feelings in the
259hearer. It may, of course, also express intentions, desires, and
so forth, on the part of the speaker that these attitudes shall be
adopted by the hearer.

With this account of the sentence before us we may consider
the traditional view both as to the distinction between noun and
verb and as to the necessity of combining them in all assertion.
There is some reason to suppose that in primitive languages the
separation of verbs and nouns reflected the distinction between the
actions of the speaker and the objects which surrounded him. At
a later stage, by a natural formal analogy, this division in linguistic
material was extensively used to mark the distinction between
things or particulars and the states, qualities, and changes which
‘belong’ or ‘happen’ to these particulars. As has been argued,
these supposed entities are in all cases of linguistic provenance,
but this did not prevent the antithesis between particular and
universal, thing and property, subject and predicate, substantive
and adjective, noun and verb, confusedly named in all these forms,
from appearing as the most fundamental with which thought could
be concerned. 1262 For Aristotle neither particular nor universal was
separately conceivable, and it is not too much to see in his doctrine
of the proposition an application of this metaphysic. On his
assumption that words ‘correspond’ to reality, neither the noun
alone, standing for a particular, nor the verb alone, standing
for a universal could in itself have a complete ‘meaning.’
There could be no better instance of the influence both of the
belief that different words and word-arrangements must stand
for different kinds of referents, and of the belief that different
kinds of referents require different kinds of words. Both these
assumptions we have seen to be unfounded.

But even should the truth of the above contentions be granted,
the moral, it may be said, is surely that grammarians should
avoid all commerce with fundamentals and confine themselves
to so-called ‘common sense’ classifications. It must, however,
be remembered that ‘common sense’ in matters of linguistic is
itself only an elaborate and confused theory, some of whose
tenets figure in our second chapter. Moreover, the current
distinctions as well as the terminology which the grammarian
proposes to employ are the legacy not only of Aristotelian dogma,
260but of that “Century of metaphysical syntax,” which, as Professor
Hale has pointed out, 1263 followed on the application of the
Kantian theory of Categories to Grammar by Hermann in 1801.
Since, therefore, a searching inquiry into the psychology of
language cannot in any case be avoided, if more is to result
from an ancient and honoured science than the mere standardization
of a score or so of convenient names for groups of words,
it is important that the issue should be squarely faced. We would
by no means belittle the serious endeavour of grammarians to
produce a certain order out of the present chaos, or underestimate
the time and energy which go to achieve this end. The
division of opinion between two of the first authorities in Europe
manifested recently 2264 as to the legitimacy of the terms ‘subjunctive-equivalent’
and ‘future in the past’ (recommended in
the Report of the Committee on Grammatical Terminology, pp. 35-6)
in elucidation of the sentence ‘I should write to him if I knew
his address,’ is, however, a good instance of the kind of nomenclature
which is being evolved. But granted that a respectable
nomenclature can be extracted from the litter of scholastic
vocables now in use, what would have been achieved? We should
not have done more than name the principal forms of speech, and
this clearly would not justify the present restriction of Grammar to
the learning of these names and to the acquisition of respect
for the standard usage of the locutions named. What is wrong
with Grammar is not its defective terminology but the lack of
interest displayed by Grammarians in the less arid and familiar
portions of the field which it professes to cover. It is for this
reason that dissatisfaction with Grammar is so prevalent, and if
as a ‘subject’ it is not to disappear from the curriculum, and with
it all theoretical study of language as an instrument of communication,
its reform must not be delayed too long. 3265

The understanding of the functions of language, of the many
261ways in which words serve us and mislead us, must be an essential
aim of all true education. Through language all our intellectual
and much of our social heritage comes to us. Our whole outlook
on life, our behaviour, our character, are profoundly influenced
by the use we are able to make of this, our chief means of contact
with reality. A loose and insincere use of language leads not
only to intellectual confusion but to the shirking of vital issues or
the acceptance of spurious formulæ. Words were never a more
common means than they are to-day of concealing ignorance
and persuading even ourselves that we possess opinions when we
are merely vibrating with verbal reverberations.

How many grammarians still regard their science as holding
the keys of knowledge? It has become for them too often merely
a technical exercise of strictly limited scope, instead of the inspiring
study of the means by which truth is acquired and preserved.
No doubt the founders of the science sufficiently misconceived
the actual powers of language, but they realized its importance.
We have examined in the course of our study the means by which
we may be put on our guard against the pitfalls and illusions due
to words. It should be the task of Grammar to prepare every
user of symbols for the detection of these. Training in translation
(p. 107), and above all in expansion (p. 93); in the technique
of substitution (p. 113), and the methods of preventing and
removing misunderstanding at different levels (p. 222); in the
discrimination of symbolic from emotive words and locutions
(p. 149); and in the recognition of the five main functions of
language (p. 224) — all are amongst the indispensable preliminaries
for the right use of language as a means of communication,
and consequently the business of Grammar.262

Appendix B
On Contexts

Now if there be anything (say/) which forms with s in virtue
of SF a Proximity context determinative in respect of s, then f
is said to be the Referent of i qua interpretation of s in this respect.
It will be noticed that f has by definition the character F and is
in proximity with s.

If there be something having these properties, then i is said
to be a true interpretation of s in respect of S; but if there be
nothing with the required properties, then i is said to be a false
interpretation in the same respect.

In more informal language, when, as a result of hearing a match
scrape, we expect a flame sensation, our belief is a process which
is a member of a psychological context united by a multiple
mnemic relation, among whose other members are past sensations
263of scrapes and flames, themselves united in dual contexts by the
relation of proximity. If now the scrape is related by this relation
to a flame, our belief is true; this sensation is the referent of our
belief. If there is no flame to which the scrape is so related our
belief is false. We have discussed (p. 71) what, if anything,
may be said to be the referent in this case.

It will be noticed that the above account deals merely with
contexts whose members are sensations. In the diagram
‘stimulus-sensation’ contexts are also included. Any actual
instances of interpretation will naturally be far more complicated
than any account or diagram which can be put on paper. The
bracket including other psychological contexts indicates one
reason for this. There must be some sense in which one context
can be said to be dependent upon others. To take a concrete
instance, the action of a penny-in-the-slot machine may be
treated as a simple dual context (insertion of penny — appearance
of chocolate) provided that certain vast multiple contexts involving
264the growth of the cocoa-tree, the specific gravity of copper, and
the regular inspection of the contrivance recur uniformly.
Psychology is throughout concerned with similar situations, but
it is less easy to analyse the contexts involved in this fashion.
It is peculiarly difficult, indeed, in Psychology to discover contexts
whose members are few in number. Even a stimulussensation
context, in order to be determinative in respect of the
character of the sensation, must ordinarily include other psychological
members, amongst which will be other sensations and
the conditions to which we allude when we use the word
‘attention.’265

Appendix C
Aenesidemus' Theory of Signs

What we know of the views of Aenesidemus is derived chiefly
from brief references in the writings of Sextus Empiricus; but
the fourth book of his lost work Πυῤῥωνίων λόγοι was devoted
to the Theory of Signs. The main arguments are summarized
by Sextus in §§ 97-134 of his Hypotheses, though it is not always
clear how much has been added by Sextus himself.

Sextus, however, is not content with disproving the Epicurean
account of signs as sensible objects. He goes on to attack the
view of the Stoics, and to show that they cannot be apprehended by
reason or intellect. Aenesidemus himself may not have gone
beyond the demonstration that (in the words of Photius) “there
are no signs, manifest and obvious, of what is obscure and latent,”
and there are those who think it probable that Sextus himself
was chiefly responsible for the distinction familiar to the later
Sceptics between two classes of signs — signs ‘commemorative’
and signs ‘demonstrative.’ 2269 According to this distinction
“there are signs which act, as we should say, by the law of
association, reminding us that in past experience two phenomena
were conjoined, as smoke with fire, a scar with a wound, a stab
to the heart with subsequent death. If afterward one of the two
phenomena is temporarily obscured and passes out of immediate
consciousness, the other, if present, may serve to recall it; we
are justified in calling the one which is present a sign, and the
other, which is temporarily absent, the thing signified. With the
term ‘sign,’ as thus understood, the sign commemorative or
reminiscent, Sextus has no quarrel. By its aid prediction is
justified; we can infer fire from smoke, the wound from the scar,
approaching death from the fatal stab, for in all these cases we
proceed upon past experience. Sextus reserves his hostility for
another class of signs which we may call the sign demonstrative.
When one of two phenomena assumed to be the thing signified
never has occurred in actual experience but belongs wholly, by
its own nature, to the region of the unknown, the dogmatists
nevertheless maintained that, if certain conditions were fulfilled,
its existence was indicated and demonstrated by the other phenomenon,
which they called the sign. For instance, according to
the dogmatists, the movements of the body indicate and demonstrate
267the existence of the soul; they are its sign. It is ‘sign’
then, in this latter sense, the indicative or demonstrative sign,
whose existence Sextus disputes and undertakes to refute.”

If such an interpretation of their views is correct it is clear
that with their account of reminiscent signs the Sceptics came
very near to formulating a modern theory of scientific induction,
while their scepticism about demonstrative signs amounts to a
denial of the possibility of inferring to the transcendental. Given
a fact, or as the Stoics called it, a ‘sign,’ we cannot determine a
priori the nature of the thing signified. That the main terms in
which the discussion was conducted suffered from confusions
which still haunt their modern equivalents, is not surprising;
there can be no signs of things to which we cannot refer, but
things can be referred to which are not experienced.

When the excavation of Herculaneum is accomplished, the
lost treatise of Philodemus on the Epicurean theory of signs and
inference which is likely to become available, together with other
similar documents relative to this remarkable controversy, may
throw more light on the progress which had been made in these
early times towards a rational account of the universe; and so
enable us to realize something of what a healthy scepticism might
have achieved had theological interests not so completely dominated
the next fifteen hundred years.268

Appendix D
Some Moderns

Those unfamiliar with the literature of Meaning will find it
difficult to realize how strange and conflicting are the languages
which the most distinguished thinkers have thought fit to adopt
in their attempts to deal with Signs, Symbols, Thoughts and
Things. In our eighth chapter sundry examples were given with
a brevity which, though necessary, may have inclined the fairminded
to question whether there has not been an occasional
injustice. We therefore append more lengthy examples, which
can be judged on their merits, from the pens of the most eminent
specialists who have dealt with the question in recent years.
It is hoped by this means to justify the assertion made at the outset
that a fresh approach was necessary.

§1. Husserl

We may begin with what is perhaps the best known modern
attempt to deal comprehensively with the problem of Signs and
Meaning, that of Professor Edmund Husserl. And it is important
for the understanding of Husserl's terminology to realize that
everything he writes is developed out of the “Phenomenological
Method and Phenomenological Philosophy” which he has been
elaborating since 1910, as Professor of Philosophy, first at Göttingen
and later at Freiburg. In June, 1922, in a course of lectures
at London University, he gave an exposition of his system to a
large English audience, and the following sentences are taken
from the explanatory Syllabus in which he, or his official translator,
endeavoured to indicate both his method and his vocabulary.

And as one of his conclusions Husserl explains that “the
transcendental monadism, which necessarily results from the
retrospective reference to absolute subjectivity, carries with it a
peculiar a priori character over against the constituted objectivities,
that of the essence-requirements of the individual
monads and of the conditions of possibility for a universe of
‘compossible’ monads. To this ‘metaphysical’ inquiry there
thus belongs the essence-necessity of the ‘harmonious accord’
of the monads through their relation to an objective world mutually
constituted in them, the problems of teleology, of the meaning
of the world and of the world's history, the problem of God.”

Such are the formulæ through which Husserl d< sired his
system to be approached, and in the narrower field of Meaning
the selection of essentials has similarly been undertaken by his
disciple, Professor J. Geyser, of the University of Münster, in
his Neue und alte Wege der Philosophic, which is devoted to a
summary of Husserl's main contributions to the theory of knowledge
in the Logische Untersuchungen, and Ideen zu einer reinen
Phenomenologie.

According to Husserl, the function of expression is only
directly and immediately adapted to what is usually described as
the meaning (Bedeutung) or the sense (Sinn) of the speech or parts
of speech. Only because the meaning associated with a wordsound
expresses something, is that word-sound called ‘expression’
(Ideen, p. 256 f). “Between the meaning and the what is
meant, or what it expresses, there exists an essential relation,
because the meaning is the expression of the meant through its
own content (Gehalt). What is meant (dieses Bedeutete) lies
in the ‘object’ of the thought or speech. We must therefore
distinguish these three — Word, Meaning, Object.” 1270270

The object is that about which the expression says something,
the meaning is what it says about it. The statement is therefore
related to the object by means of its meaning. But Husserl maintains
explicitly: “The object never coincides (zusammenfällt)
with the meaning” (L.U., II., i., p. 46). He bases this assertion
on the fact “that several expressions can have the same meaning,
but different objects, and again, different meanings, but the same
object” (Ibid., p. 47). “The two expressions ‘equiangular and
equilateral triangle’ have for example a different meaning, but
name the same object. Conversely a different object but the
same meaning is signified when Bucephalus and a carthorse are
described as ‘horse.’ The meaning of an expression becomes an
object only when an act of thought turns towards it reflectively.” 1271

The sense of the expression ‘meaning’ which, according to
Geyser (p. 33) is as a rule synonymous with ‘concept’ (was meist
als Begriff bezeichnet wird), Husserl illustrates by the comparison
of two cases. In the perception of a white object, we can be
satisfied by perceiving it and eventually distinguishing something
or other in it. For this function, expression and meaning are not
necessary. But we can also pass on to the thought: “This is
white.” The perceiver has now added to the perceiving a mental
act, which expresses, means the thing perceived and the thing
distinguished in what is perceived, that is to say, the objective.
The expression is therefore, to state the matter generally, a form
which raises the sense “into the kingdom of the ‘Logos’ of the
conceptual and thereby the ‘general’” (Ideen, p. 257). The
function of the expression, of this peculiar intention, “exhausts
itself in expressing, and that with this newly entering form of the
conceptual” (Ibid., p. 258). Further, ‘expressing’ is an imitative,
and not a productive function.

By the words ‘expression’ and ‘meaning,’ Husserl describes
in the first place concepts, but also judgments and conclusions:
“Pure logic, wherever it deals with concepts, judgments, conclusions,
has in fact to do exclusively with these ideal unities,
which we here call meanings” (L.U., II., i., p. 916). In general,
it is “evident that logic must be knowledge of meanings as such,
of their essential kinds and differences, as well as of the laws
purely grounded in them (that is to say ideal). For to these
essential distinctions belong also those between meanings, with
and without objects, true and false…” (Ibid., p. 92). All
thought has a certain appropriate range of acts of expressing or
meaning, which are neither identical with the sensory word nor
271with the objects of cognition. “It is not easy to realize clearly
that actually after abstraction of the sensory word-sound stratum,
a stratification is found of the kind that we suppose here; that is
to say in every case — even in that of unclear, empty, and merely
verbal thought — a stratum of expressing meaning and a substratum
of expressed. Still less easy is the understanding of the
essential connections of these strata” (Ideen, p. 259).

Husserl proceeds to distinguish between what he calls ‘meaning-intentions’
(Bedeutungsintentionen) and ‘realized meanings’
(erfiillte Bedeutungen); between ‘meaning-conferring’ and
‘meaning-realizing’ acts (L.U., i., p. 38); and between the
psychological and objective-phenomenological treatment of
meaning. 1272 Phenomenologically, when we ask the meaning of
the expression ‘prime-number’ we refer to (meinen) this expression
in itself and as such, not in its particularity (Besonderheit),
as it is spoken by a given individual in a lecture, or as it is found in
such and such a book written in such and such a way. Rather
we simply ask: What does the expression ‘Prime-number’
mean? Similarly we do not ask what at this or that moment was
the meaning of the expression thought and experienced by such
and such a man; we ask in general about its meaning as such and
in itself. Husserl expresses this state of affairs by saying that in
such questions it is a matter of the expression and the meaning
‘in specie,’ ‘as species,’ ‘as idea,’ ‘as ideal unity’; for what is
referred to is one and the same meaning, and one and the same
expression, however these may be thought or spoken (L.U., II.,
i., p. 42 f). Hence: Meanings, ideal objects, must have being,
since we predicate truly of them — as when we say that four is
an even number (Ibid., p. 125); but their existence does not
depend on their being thought. They have eternal, ideal, existence. 2273
“What Meaning is can be given to us as immediately
as colour and tone. It cannot be further defined; it is a descriptive
ultimate. Whenever we complete or understand an expression,
it means something to us and we are actually conscious of
its sense.” Distinctions between meanings are also directly given,
and we can classify these in the Phenomenology of meaning, as
‘symbolic-empty’ ‘intuitively realized,’ etc.; such operations
as identification and distinction, relating, and generalizing
abstraction, give us “the fundamental logical concepts, which
are nothing but ideal conceptions of the primitive distinctions of
meaning” (Ibid., p. 183).272

§2. Bertrand Russell

Mr Russell's best known view (which must now, however,
be read in connection with his more acceptable psychological
account discussed in our third chapter, and with his Monist
articles 1918-1919) is to be found at page 47 of his Principles of
Mathematics. He is there concerned with the connection of his
doctrine of adjectives with certain traditional views on the nature
of propositions, and with the theory of Bradley 1274 “that all words
stand for ideas having what he calls meaning, and that in every
judgment there is a something, the true subject of the judgment,”
which is not an idea and does not have meaning. “To have meaning,”
says Mr Russell, “is a notion confusedly compounded of
logical and psychological elements. Words all have meaning,
in the simple sense that they are symbols which stand for something
other than themselves. But a proposition, unless it happens
to be linguistic, does not itself contain words: it contains the entities
indicated by words. Thus meaning, in the sense in which
words have meaning, is irrelevant to logic. But such concepts
as a man have meaning in another sense: they are, so to speak,
symbolic in their own logical nature, because they have the property
which I call denoting. That is to say, when a man occurs
in a proposition (e.g., ‘I met a man in the street’), the proposition
is not about the concept a man, but about something quite
different, some actual biped denoted by the concept. Thus
concepts of this kind have meaning in a non-psychological sense.
And in this sense, when we say ‘this is a man,’ we are making a
proposition in which a concept is in some- sense attached to what
is not a concept. But when meaning is thus understood, the
entity indicated by John does not have meaning, as Mr Bradley
contends; and even among concepts, it is only those that denote
that have meaning. The confusion is largely due, I believe, to
the notion that words occur in propositions, which in turn is due
to the notion that propositions are essentially mental and are to
be identified with cognitions.”

§3. Frege

Frege's theory of Meaning is given in his Begriffsschrift,
Grundlagen der Aritkmetik and his articles on “Begriff und
Gegenstand,” and “Sinn und Bedeutung.” A convenient summary,
which we here follow, is given at p. 502 of his Principles
by Mr Russell, who holds that Frege's work “abounds in subtle
273distinctions, and avoids all the usual fallacies which beset writers
on Logic.” The distinction which Frege makes between meaning
(Sinn) and indication (Bedeutung) is roughly, though not exactly,
equivalent to Mr Russell's distinction between a concept as such
and what the concept denotes (Principles, §96). Frege did not
possess this distinction in the first two of the works under consideration
; but it makes its appearance in B.u.G., and is specially
dealt with in S.u.B. Before making the distinction, he thought
that identity had to do with the names of objects (Bs., p. 13):
“A is identical with B” means, he says, that the sign A and the
sign B have the same signification (Bs., p. 15) — a definition which,
in Mr Russell's view, “verbally at least, suffers from circularity.”
But later he explains identity in much the same way as Mr
Russell did in the Principles, §64. “Identity,” he says, “calls
for reflection owing to questions which attach to it and are not
quite easy to answer. Is it a relation? A relation between
Gegenstande or between names or signs of Gegenstande?”
(S.u.B., p. 25). We must distinguish, he adds, the meaning, in
which is contained the way of being given, from what is indicated
(from the Bedeutung). Thus ‘the evening star’ and ‘the
morning star’ have the same indication, but not the same meaning.
A word ordinarily stands for its indication; if we wish to
speak of its meaning, we must use inverted commas or some
such device. The indication of a proper name is the object which
it indicates; the presentation which goes with it is quite subjective;
between the two lies the meaning, which is not subjective and yet
is not the object. A proper name expresses its meaning, and
indicates its indication.

“This theory of indication,” adds Mr Russell, “is more
sweeping and general than mine, as appears from the fact that
every proper name is supposed to have the two sides. It seems
to me that only such proper names as are derived from concepts
by means of the can be said to have meaning, and that
such words as John merely indicate without meaning. If one
allows, as I do, that concepts can be objects and have proper
names, it seems fairly evident that their proper names, as a
rule, will indicate them without having any distinct meaning;
but the opposite view, though it leads to an endless regress,
does not appear to be logically impossible.”

§4. Gomperz

From all this, says Dittrich, the peculiarly relational character
of that element of the statement named meaning results. Meaning
cannot be identified with mere designation (Bezeichnung). One
and the same sound, e.g., ‘top’ can, he urges, designate very
different foundations; and if, with Martinak, we confine meaning
to the relation between the sign and what is designated, we
cannot reach a satisfactory definition. Interpretation (Auffassung)
may similarly be a many-one relation; moreover to use
the term meaning for that relation would omit the linguistic
element. Nor can meaning be identified with the relation of
expression (Ausdruck). Finally, Meaning appears as a definite
but complex relation, based on the theory of ‘total-impressions’
(Totalimpression) and common emotional experiences which
distinguishes the pathempiricists. 1276 “Any sound whatever can
276designate any foundation; but it can only mean when it becomes
a statement through the constitution of a general-typical import,
and that becomes the foundation (Grundlage) for a fact (Sachverhalt).” 1277

§5. Baldwin

Professor Baldwin's mode of treating the problem of Meaning
is best studied in his Thought and Things. Vol. II. of this work
deals with what he calls ‘Experimental Logic,’ and Chapter VII.
is devoted to the Development of Logical Meaning. “Our most
promising method of procedure would seem to be to take the
various modes or stages in the development of predication, and
to ask of each in turn as to its structural or recognitive meaning,
its ‘what’ — that is, what it now means, as an item of contextuated
and socially available information. The ‘what’ is the subjectmatter
of judgment. Having determined this, we may then
inquire into the instrumental use of such a meaning: the ‘proposal’
that the meaning when considered instrumentally suggests
or intends. This latter we may call the question of the ‘why’
of a meaning: the for-what-purpose or end, personal or social,
the meaning is available for experimental treatment. If we use
the phrase ‘selective thinking,’ as we have above, for the entire
process whereby meanings grow in the logical mode — the process
of ‘systematic determination’ sketched in the preceding chapter
— then we may say that every given meaning is both predication
as elucidation of a proposal, and predication as a proposal for
elucidation. It is as his elucidation that the believer proposes
it to another; it is as proposal that the questioner brings it to the
hearer for his elucidation. We may then go forward by this
method…”

Later (p. 299) the question arises:“In what sense can a meaning
that is universal as respects community still be singular?” And
the answer is as follows: “That it does banish singular meaning
from the logical, if by the singular we mean a type of meaning
that lacks community. For when a meaning of singularity is
rendered in a judgment then precisely the marks that served to
make it singular are generalized in one of the modes of community
— as recurring in different experiences either for the same or for
different persons. The intent of singularity which admits of no
generalization has then retreated into the domain of direct
appreciation or immediate experience.” This, he says, may be
illustrated without difficulty. “Suppose I submit the statement
‘this is the only orange of this colour.’ By so doing I give the
orange a meaning in community in two ways. I mean that you
can find it the only one with me, or that I myself can find it the
same one by repeating my experience of it.”

Finally (p. 423), in replying to Professor A. W. Moore's
difficulties as regards his terminology, Baldwin explains himself
thus: — “Our relativisms are contrast-meanings, dualisms,
instrumentalities one to another, and the mediation and abolishing
of these contrasts, dualisms, means to ends, removes the
relativities and gives the only tenable ‘absolute.’ This is the
‘absolute’ that experience is competent to reach. If you ask
why this does not develop again into new relativities, I answer,
in fact it does; but in meaning it does not. For the meaning is the
universal of all such cases of mediation. If the mediation effected
in the æsthetic is one of typical meaning everywhere in the progression
of mental ‘dynamic,’ then it is just its value that it discounts
in advance any new demands for mediation which new
dualisms may make. The æsthetic is absolute then in the only
sense that the term can mean anything: it is universal progressionwise,
as well as content or relation-wise. It mediates the genetic
dynamogenies as well as the static dualisms.” And then he turns
to meaning.

“As to ‘meaning,’ I hold that after meaning arises as
over against mere present content, then the content of necessity
278and by contrast also becomes meaning; since consciousness
may then intend or mean both, either,, or the difference
between the two. As I put it in Vol. I., with the rise of
meaning there arise meanings (in the plural). To hold a
content to just its bare presence is to make it a meaning —
after consciousness is once able to mean ‘that only and not
anything else.’ Consequently the use of ‘meaning’ for
what is had in mind (as in the phrase ‘I mean so and so’)
supersedes the use of it for that merely which is attached to
a content (as in ‘it means much’). When I say (in the former
sense) ‘I mean chickens,’ I do not intend to restrict ‘meaning’
to what the chicken suggests beyond the bare image.
On the contrary, I intend the whole bird.”

It should be added that C. S. Peirce, to whom we now turn,
wrote very highly of Professor Baldwin's terminology.

§6. C. S. Peirce

By far the most elaborate and determined attempt to give an
account of signs and their meaning is that of the American
logician C. S. Peirce, from whom William James took the idea
and the term Pragmatism, and whose Algebra of Dyadic Relations
was developed by Schroeder. Unfortunately his terminology
was so formidable that few have been willing to devote time to
its mastery, and the work was never completed. “I am now
working desperately to get written before I die a book on Logic
that shall attract some minds through whom I may do some real
good,” he wrote to Lady Welby in December, 1908, and by the
kindness of Sir Charles Welby such portions of the correspondence
as serve to throw light on his published articles on Signs are here
reproduced.

In a paper dated 1867, May 14th (Proc. Am. Acad. Arts and Sci.
(Boston), VII (1868), 295), Peirce defined logic as the doctrine of
the formal conditions of the truth of symbols; i.e., of the reference
of symbols to their objects. Later, when he “recognized
that science consists in inquiry not in ‘doctrine’ — the history of
words, not their etymology, being the key to their meanings,
especially with a word so saturated with the idea of progress as
science,” he came to realize, as he wrote in 1908, that for a long
time those who devoted themselves to discussing “the general
reference of symbols to their objects would be obliged to make
researches into the references to their interpretants, too, as well
as into other characters of symbols, and not of symbols alone
279but of all sorts of signs. So that for the present, the man who
makes researches into the reference of symbols to their objects
will be forced to make original studies into all branches of the
general theory of signs.” This theory he called Semeiotic, and
its essentials are developed in an article in the Monist, 1906,
under the title, “Prolegomena to an Apology for Pragmaticism.”

A sign, it is there stated, “has an Object and an Interpretant,
the latter being that which the Sign produces in the Quasi-mind
that is the Interpreter by determining the latter to a feeling, to
an exertion, or to a Sign, which determination is the Interpretant.
But it remains to point out that there are usually two Objects,
and more than two Interpretants. Namely, we have to distinguish
the Immediate Object, which is the object as the Sign itself
represents it, and whose Being is thus dependent upon the
Representation of it in the sign, from the Dynamical Object,
which is the Reality which by some means contrives to determine
the Sign to its Representation. In regard to the Interpretant
we have equally to distinguish in the first place, the
Immediate Interpretant, which is the interpretant as it is revealed
in the right understanding of the Sign itself, and is ordinarily
called the ‘meaning’ of the sign; while, in the second place, we
have to take note of the Dynamical Interpretant, which is the
actual effect which the Sign, as a Sign, really determines. Finally,
there is what I provisionally term the Final Interpretant, which
refers to the manner in which the Sign tends to represent itself
to be related to its Object. I confess that my own conception of
this third interpretant is not yet quite free from mist.”

The special interest to Peirce of the distinctions thus christened
was their application in explaining and developing a system of
‘Existential Graphs,’ whereby diagrams are furnished “upon
which to experiment, in the solution of the most difficult problems
of logic.” A diagram, he notes, “though it will ordinarily have
Symbolide features, is in the main an Icon of the forms of relations
in the constitution of its Object.” And in the same terminology
it could be said that the footprint which Crusoe found in the sand
“was an Index to him of some creature, while as a Symbol it
called up the idea of a man.” In the material here reproduced we
are not concerned with the special applications which its author
made of his theory, but in view of his constant insistence on the
logical nature of his inquiry and his desire to avoid psychology,
a further trichotomy 1278 of general interest may here be mentioned.
Logic he defined in an article in the Monist (Vol. VII., 1896-7,
p. 25) as dealing with the problem, “to what conditions an
assertion must conform in order that it may correspond to the
‘reality’”; Speculative Grammar was the name given also by
Duns Scotus to “the study of properties of beliefs which belong
to them as beliefs”; and thirdly, “the study of those general
conditions under which a problem presents itself for solution,
and then under which one question leads on to another,” appears
as Universal Rhetoric. In writing to Lady Welby, he remarks that
‘Signifies,’ the term which she used for the study of Meaning,
“would appear from its name to be that part of Semeiotic which
inquires into the relation of Signs to the Interpretants (for which,
281as limited to Symbols, I proposed in 1867 the name Universal
Rhetoric).” He strongly urges her to make a scientific study of
Semeiotic, as well as of his Graphs (“I wish you would study
my Existential Graphs; for in my opinion it quite wonderfully
opens up the true nature and method of logical analysis — that is
to say, of definition; though how it does so is not easy to make out,
until I shall have written my exposition of that art” ); and in a
letter written in 1904, shortly before the publication of his chief
Monist article, he deals with the classification of Signs at some
length.

He prefaces his remarks by insisting that “a sign has two
Objects, its object as it is represented and its object in itself. It
has also three Interpretants, its interpretant as represented or
meant to be understood, its interpretant as it is produced, and its
interpretant in itself.” Signs may be divided as to their own
material nature, as to their relations to their objects, and as to
their relations to their interpretants.

“As it is in itself a sign is either of the nature of an appearance,
when I call it a qualisign; or secondly, it is an individual
object or event, when I may call it a sinsign (the syllable sin
being the first syllable of Semel, simul, singular, etc.); or
thirdly, it is of the nature of a general type, which I call a
legisign. As we use the term ‘word’ in most cases, saying that
‘the’ is one ‘word’ and ‘an’ is a second ‘word,’ ‘word’
is a legisign. But when we say of a page in a book that it has
250 ‘words’ upon it, of which twenty are ‘the s,’ the ‘word’
is a sinsign. A sinsign so embodying a legisign, I term a
replica of the legisign. The difference between a legisign and
a qualisign, neither of which is an individual thing, is that a
legisign has a definite identity, though usually admitting a
great variety of appearances. Thus &, and, and the sound are
all one word. The qualisign, on the other hand, has no identity.
It is the mere quality of an appearance, and is not exactly
the same throughout a second. Instead of identity it has
great similarity, and cannot differ much without being called
quite a different qualisign.”

With regard to the other main divisions of signs he explains
that “in respect to their relations to their dynamic objects, I
divide signs into Icons, Indices and Symbols (a division I gave in
1867). I define an Icon as a sign which is determined by its
dynamic object by virtue of its own internal nature. Such is
any qualisign like a vision, or the sentiment excited by a piece of
282music considered as representing what the composer intended.
Such may be a sinsign like an individual diagram; say a curve
of the distribution of errors. I define an Index as a sign determined
by its dynamic object by virtue of being in a real relation
to it. Such is a Proper Name (a legisign), such is the occurrence
of a symptom of a disease (the symptom itself is a legisign, a
general type of a different character. The occurrence in a particular
case is a sinsign). I define a Symbol as a sign which is
determined by its dynamic object only in the sense that it will be
so interpreted. It thus depends either upon a convention, a
habit 1279 or a natural disposition of its interpretant or of the field of
its interpretant (that of which the interpretant is a determination).
Every symbol is necessarily a legisign; for it is inaccurate to call
a replica of a legisign a symbol.”

In respect of its immediate object a sign may either be a sign
of quality, of an existent or of a law; while in regard to its relation
to its signified interpretant, it is said to be either a Rheme, a
Dicent, or an Argument. “This corresponds to the old division
Term, Proposition, and Argument, modified so as to be applicable
to signs generally. A Term is simply a class-name or Propername.
I do not regard the common noun as an essentially necessary
part of speech. Indeed, it is only fully developed as a
separate part of speech in the Aryan languages and the Basque —
possibly in some other out of the way tongues. In the Semitic
languages it is generally in form a verbal affair, and usually is so in
substance too. As well as I can make out, such it is in most
languages. In my universal algebra of logic there is no common
noun.”

A Rheme is defined as “a sign which is represented in its
signified interpretant as if it were a character or mark (or as being
so).” It is any sign that is neither true nor false, like most single
words except ‘yes’ and ‘no,’ which are almost peculiar to
modern languages.

A Dicent is defined as “a sign represented in its signified
interpretant as if it were in a real relation to its object (or as being
so if it is asserted).” A proposition, he was careful to point out
in the Monist (1905, p. 172). is for him not the German Satz, but
“that which is related to any assertion, whether mental and self-addressed
or outwardly expressed, just as any possibility is
related to its actualization.” It is here defined as a dicent symbol.
283“A dicent is not an assertion, but a sign capable of being
asserted. But an assertion is a dicent. According to my
present view (I may see more light in future) the act of assertion
is not a pure act of signification. It is an exhibition of
the fact that one subjects oneself to the penalties visited on
a liar if the proposition asserted is not true. An act of
judgment is the self-recognition of a belief; and a belief
consists in the deliberate acceptance of a proposition as a
basis of conduct. But I think this position is open to doubt.
It is simply a question of which view gives the simplest view of
the nature of the proposition. Holding then that a Dicent does
not assert, I naturally hold that the Argument need not be
actually submitted or urged. I therefore define an Argument
as a sign which is represented in its signified interpretant not
as a Sign of that interpretant, the conclusion, but as if it were
a Sign of the Interpretant, or perhaps as if it were a Sign of the
state of the Universe to which it refers in which the premises
are taken for granted.”

This treatment of the familiar logical distinction between
Term, Proposition, and Argument is somewhat different from the
284account given in the Monist (1906) article, where it is explained
that “the first two members have to be much widened,” and
where we are introduced to Semes, Phemes, and Delomes.

“By a Seme I should mean anything which serves for any
purpose as a substitute for an object of which it is, in some
sense, a representative or Sign. The logical Term, which is a
class-name is a Seme. Thus the term ‘The Mortality of man’
is a Seme. By a Pheme I mean a sign which is equivalent
to a grammatical sentence, whether it be Interrogative, Imperative
or Assertory. In any case, such a Sign is intended to have
some sort of compulsive effect on the Interpreter of it. As
the third member of the triplet, I sometimes use the word
Delome (pronounced deeloam, from δήλομα), though Argument
would answer well enough. It is a sign which has the Form
of tending to act upon the Interpreter through his own selfcontrol,
representing a process of change in thoughts or signs,
as if to induce this change in the Interpreter.”

A Graph, he says, is a Pheme, “and in my use hitherto, at
least, a Proposition. An Argument is represented by a series of
Graphs.”

There follows a discussion of.“the Percept, in the last analysis
the immediate object of all knowledge and all thought.”

“This doctrine in nowise conflicts with Pragmaticism, which
holds that the Immediate Interpretant of all thought proper
is Conduct. Nothing is more indispensable to a sound epistemology
than a crystal-clear discrimination between the object
and the Interpretant of knowledge; very much as nothing is
more indispensable to sound notions of geography than a crystal-clear
discrimination between north latitude and south latitude;
and the one discrimination is not more rudimentary than the
other. That we are conscious of our Precepts is a theory which
seems to me to be beyond dispute; but it is not a fact of Immediate
Perception. A fact of Immediate Perception is not a
Percept, nor any part of a Percept; a Percept is a Seme, while a
fact of Immediate Perception or rather the Perceptual Judgment
of which such fact is the immediate Interpretant is a Pheme
that is the direct Dynamical Interpretant of the Percept, and
of which the Percept is the Dynamical Object, and is with some
considerable difficulty (as the history of psychology shows)
distinguished from the Immediate Object, though the distinction
is highly significant. But not to interrupt our train of
thought, let us go on to note that while the Immediate Object
285of a Percept is excessively vague, yet natural thought makes up
for that lack (as it almost amounts to) as follows: — A late
Dynamical Interpretant of the whole complex of Percepts is
the Seme of a Perpetual Universe that is represented in instinctive
thought as determining the original Immediate Object of every
Percept. Of course, I must be understood as talking not psychology,
but the logic of mental operations. Subsequent Interpretants
furnish new Semes of Universes resulting from various
adjunctions to the Perceptual Universe. They are, however, all
of them, Interpretants of Percepts.

Finally, and in particular, we get a Seme of that highest of all
Universes which is regarded as the Object of every true
Proposition, and which, if we name it all, we call by the somewhat
misleading title of ‘The Truth.’

That said, let us go back and ask this question: How is
it that the Percept, which is a Seme, has for its direct dynamical
Interpretant the Perceptual Judgment, which is a Pheme?
For that is not the usual way with Semes, certainly. All the
examples that happen to occur to me at this moment of such
action of Semes are instances of Percepts, though doubtless there
are others. Since not all Percepts act with equal energy in this
way, the instances may be none the less instructive for being
Percepts. However, Reader, I beg you will think this matter
out for yourself, and then you can see — I wish I could — whether
your independently formed opinion does not fall with mine.
My opinion is that a pure Perceptual Icon — and many really
great psychologists have evidently thought that Perception is
a passing of images before the mind's eye, much as if one were
walking through a picture gallery, — could not have a Pheme for
its direct Dynamical Interpretant. I desire, for more than one
reason, to tell you why I think so, although that you should
to-day appreciate my reasons seems to be out of the question.
Still I wish you to understand me so far as to know that, mistaken
though I be, I am not so sunk in intellectual night as to be
dealing lightly with philosophic Truth when I aver that weighty
reasons have moved me to the adoption of my opinion; and I am
also anxious that it should be understood that those reasons
have not been psychological at all, but purely logical. My
reason, then, briefly stated and abridged, is that it would be
illogical for a pure Icon to have a Pheme for its Interpretant,
and I hold it to be impossible for thought not subject to self-control,
as a Perceptual Judgment manifestly is not, to be
illogical. I dare say this reason may excite your derision or
286disgust, or both; and if it does I think none the worse of your
intelligence.”

There is an interesting letter dated March 14, 1909, in which
Lady Welby's own Triad of Interpretation is discussed. “I
confess,” he writes, “I had not realized before reading your
Encyclopædia Britannica article, how fundamental your trichotomy
of Sense, Meaning and Significance really is. It is not to
be expected that concepts of such importance should get perfectly
defined for a long time… I now find that my division (of
the three kinds of Interpretant) nearly coincides with yours, as
it ought to do exactly, if both are correct. I am not in the least
conscious of having been influenced by your book in setting my
trichotomy.” He does not believe that there was even an unconscious
reminiscence, and consequently feels “some exultation
in finding that my thought and yours nearly agree.”

He proceeds to inquire how far there is agreement. “The
greatest discrepancy appears to lie in my Dynamical Interpretant
as compared with your ‘Meaning.’ If I understand the latter,
it consists in the effect on the mind of the Interpreter that the
utterer (whether vocally or by writing) of the sign intends to
produce. My Dynamical Interpreter consists in direct effect
actually produced by a Sign upon an Interpreter of it. They
agree in being effects of the Sign upon an individual mind, I
think, or upon a number of actual individual minds by independent
action upon each. My Final Interpretant is, I believe,
exactly the same as your Significance; namely, the effect the
Sign would produce upon any mind upon which circumstances
should permit it to work out its full effect. My Immediate Interpretant
is, I think, very nearly, if not quite, the same as your
‘Sense’; for I understand the former to be the total unanalysed
effect that the Sign is calculated to produce; and I have been
accustomed to identify this with the effect the sign first produces
or may produce upon a mind, without any reflection upon it. I
am not aware that you have ever attempted to define your term
‘Sense,’ but I gather from reading over what you say that it is
the first effect that a sign would have upon a mind well qualified
to comprehend it. Since you say it is Sensal and has no Volitional
element, I suppose it is of the nature of an ‘impression.’ It is
thus, as far as I can see, exactly my Immediate Interpretant.
You have selected words from vernacular speech to express your
varieties, while I have avoided these and have manufactured terms
suitable, as I think, to serve the uses of Science. I might describe
my Immediate Interpretation as so much of the effect of a Sign
287as would enable a person to say whether or not the Sign was
applicable to anything concerning which that person had sufficient
acquaintance.”

Peirce's conception of an ‘Interpretant’ receives further
elucidation in a letter written at the end of 1908, from which we
have already quoted. He there emphasizes that in all questions
of interpretation it is indispensable to start with an accurate and
broad analysis of the nature of a sign. “I define a Sign as anything
which is so determined by something else, called its Object,
and so determines an effect upon a person, which effect I call its
Interpretant, that the latter is thereby mediately determined by
the former. My insertion of ‘upon a person’ is a sop to Cerberus,
because I despair of making my own broader conception understood.
I recognize three Universes which are distinguished by
288three Modalities of being. One of these Universes embraces
whatever has its Being in itself alone, except that whatever is
in this Universe must be present to one consciousness, or be
capable of being so present to its entire Being.” The objects of
this Universe he called Ideas or Possibles, the objects of the second
or actual Universe being Facts, and of the third Necessitants.

The Mode of Being of signs can be ‘possible’ (e.g., a hexagon
circumscribed in or about a conic); or ‘actual’ (as with a barometer);
or ‘necessitant’ (as the word ‘the,’ or any other in the
dictionary). A ‘possible’ sign he calls, as in the Monist article,
a Tone (“though I am considering replacing this by ‘Mark’”);
an ‘actual’ sign, a Token; a ‘necessitant’ sign a Type.

“It is usual and proper to distinguish two Objects of a Sign,
the Mediate without, and the Immediate within the Sign. Its
Interpretant is all that the sign conveys; acquaintance with its
Object must be gained by collateral experience. The Mediate
Object is the Object outside the Sign; I call it the Dynamoid
Object. The Sign must indicate it by a hint; and this hint,
or its substance, is the Immediate Object”

When the Dynamoid object is ‘possible,’ the sign will be
Abstractive (as the word Beauty), when it is actual the sign will
be Concretive (any one barometer or a written narrative of any
series of events); and thirdly, “for a sign whose Dynamoid
Object is a Necessitant, I have at present no better designation
than a ‘Collective,’ which is not quite so bad a name as it sounds
to be until one studies the matter; but for a person like me, who
thinks in quite a different system of symbols to words, it is awkward
and often puzzling to translate one's thought into words!
If the Immediate Object is a ‘Possible’ (that is, if the Dynamoid
Object is indicated, always more or less vaguely, by means of its
Qualities, etc.) I call the Sign a Descriptive; if the Immediate is
an Occurrence, I call the Sign a Designative; and if the Immediate
Object is a Necessitant, I call the Sign a Copulant; for in that
case the Object has to be so identified by the Interpreter that
the Sign may represent a necessitation.”

A Possible can determine nothing but a Possible, and a
Necessitant can be determined by nothing but a Necessitant.
“Hence,” he continues, “it follows from the definition of a
Sign that since the Dynamoid Object determines the Immediate
object,
which determines the Sign itself,
which determines the Destinate Interpretant,
289which determines the Effective Interpretant,
which determines the Explicit Interpretant,
the six trichotomies, instead of determining 729 classes of signs,
as they would if they were independent, only yield 28 classes;
as I strongly opine (not to say almost approve) there are four other
trichotomies of signs of the same order of importance, instead
of making 59,049 classes, these will only come to 66. The additional
4 trichotomies are undoubtedly first Icons (or Simulacra),
Indices, Symbols, and then three referring to the Interpretants.
One of these I am pretty confident is: Suggestives, Imperatives,
Indicatives, where the Imperatives include Interrogatives, Of
the other two I think that one must be into Signs assuring their
Interpretants by Instinct, Experience, and Form. The other I
suppose to be what (in the Monist (1906) article) I called Semes,
Phemes, and Delomes.” 1280290

Appendix E
On negative facts

We may approach the discussion of Facts from many angles,
but perhaps it is best to begin by considering the controversy
about Negative Facts in which the issues come clearly to a
head. In 1917 Mr Raphæl Demos published in Mind the
results of an interrogatory to which he had subjected his more
intelligent non-philosophical acquaintances — as to whether
they had ever personally encountered a negative fact. All
concurred in the opinion that “every case of knowledge expressed
through a negative proposition was in reality of a positive nature,
in a fashion which they were unable to comprehend.”

In his desire not to oppose this verdict of experience without
good reason, the writer ventured to question the orthodox
conclusion that negative facts are an essential constituent of the
universe, and substituted a theory of contrariety between propositions
whereby “John is not in England” is to be construed
as a description of some positive proposition (“John is in Paris” )
incompatible with the positive proposition originally denied
(“John is in England” ). So intrigued was the author of Principia
Mathematica by this logical escapade that, in spite of the almost
unquenchable desire to escape the admission of negative facts
which he had noted as implanted in every human breast, he was
constrained to examine the argument minutely and to traverse
it by pointing out that, ‘incompatible’ being identical with
‘not compatible,’ a negative fact had been illicitly admitted by
the interpretation itself. Should the interpretation be reapplied
to eject this, this application again admits an intruder
and so on.

It is to be noted, however, that in point of time Mr W. E.
Johnson intervened in the pages of Mind with the following
dictum: “We can only say that ‘incompatible’ means ‘incompatible
with compatible’ — or to put it otherwise incompatible
is just as ultimate a positive relation as compatible.” Further
291moves in the game were to be expected; some of them, indeed,
are to be found in Professor Eaton's Symbolism and Truth

The Doctrine of Symbolism allows us, however, quietly to
settle the dispute by turning the attention to what it is about.
We can then apply the Theory of Signs upon which the Doctrine
depends and point out to what the dispute has been due.

It is about the referents of certain complex symbols; those
containing the term ‘not’ or an equivalent. It is about whether
the symbol for one of these is ‘negative fact’ or ‘not a fact,’ and
about the supposed consequences of this momentous decision.
We may best explain by returning at this point to the term Fact,
disregarding for the moment the problem of the negative.

The proposition, or complex symbol, “Charles I. died on the
scaffold,” is used to refer to a certain complex referent. Whenever
a form of words has no referent it fails to be a symbol and is
nonsense. In this case the referent is admitted by historians to
belong to the order of referents which they call ‘historical
events.’

Similarly, the complex sign, “Alexander VI. became a rat-catcher,”
has a referent which historians exclude from the historical
order. They will do this on the ground that all the places
into which this referent might fit are filled by other referents.
They say then (if symbolists) that this referent belongs to some
other order; 1281 either the order of Rabelais' infernal events, or
some other order of imaginary events, or events of some
imagination — all ‘historical’ in the wider sense of events which
have happened.

When the referent of a given symbol belongs to the order
within which we are looking for it, we commonly say “the
symbol (‘Charles I. died on the scaffold’) expresses a fact,” or
“It is a fact that (the symbol)”: more often we say “(The symbol
— viz., Charles I., etc.) is true.” These locutions have the same
referent, the referent more adequately referred to by the complex
symbol: — “The referent belongs to the order to which it is allocated
(by context or openly) by a reference.” 292

A Fact, therefore, is a referent which belongs to the order to
which it is allocated. This definition of ‘a fact’ solves the ‘problem
of negative facts’ with which we began. No other wilt
solve it. The referent in part of the complex symbol (1) “Charles
I. did not die on the scaffold” is also the referent in part of the
complex symbol (2) “Charles I. died on the scaffold,” but with a
different allocation. More clearly stated the expanded form of
(1) is “The referent of the symbol ‘Charles I. died on the scaffold’
belongs to another order than that of historical events.” The
expanded form of (2) is “The referent of the symbol ‘Charles died
on the scaffold’ belongs to the historical order.” Since historians
find the referent of “Charles I. died on the scaffold” in the
historical order we can say that (1) is false and (2) is true; but
in so doing we are merely using alternative locutions.

The converse case of the symbols (1) “Charles I. did not die
293in his bed” and (2) “Charles I. died in his bed” is treated in the
same fashion. (1) expands to “The referent of ‘Charles I. died
in his bed’ belongs to another order than that of historical events.”
(2) expands to “The referent of ‘Charles I. died in his bed’ belongs
to the historical order.” Historians find the ‘place’ in the historical
order which would be filled by this referent filled by some
other referent. We may therefore say that (1) is true and (2) is
false, or that (1) refers to a fact and (2) does not so refer, or
refers to what is not a fact or to a negative fact; but in so saying
we shall merely be using rival shorthands, developed for the
sake of linguistic convenience.

A piece of string will tie up the same parcel whether it has a
knot in it or not. There is no further peculiarity about those
parcels which happen to be tied by string containing knots.
They are neither ‘parcels containing knots’ nor ‘knotty parcels,’
but just honest parcels. Similarly it should now be obvious
that though propositions containing negative elements differ,
qua propositions, from those devoid of nots, the distinction does
not imply parallel differences in the objects referred to, or a
special class of negative objects. And this is of course equally
true when a negative element is used merely as an indication of a
relation between Symbols, as in Peano's Fourth Postulate “o is
not the successor of any number,” and in the case of objects to
which we happen not to be able to refer by other linguistic
means. When we dispute as to whether a fact is positive or
negative, or whether there are ‘negative facts,’ we are engaged
merely in the criticism of rival prose styles.

Now, after much time, there appeared Reason, which said,
“Wherefore hast thou done this thing?”

And Homo said “Speech bewrayed me.”

To whom Reason “Go to now and seek the Doctrine of Symbolism
which showeth that the bee buzzeth not in the Head but
in the Bonnet.”

But Homo hearkened not, and his sin was the greater in that
he was proud and obstinate withal. For as Philosopher and
Economist he said — “We will tend to give the matter our careful
consideration.” And as Returning Warrior, he asked: “What,
grannie, didst thou say in the Great Wars?” And as Plain Man
he continued to splash solemnly about in the Vocabulary of
Ambiguity — and all the while the Noose was tightening and Homo
began to grow inarticulate.

Then had Reason compassion on him, and gave him the Linguistic
Conscience, and spake again softly: “Go to now, be a
Man, Homo! Cast away the Noose of Words which thou hast
woven, that it strangle thee not. Behold! the Doctrine of Symbolism,
which illumineth all things. What are the Laws of
Science? Are they not thine own Conceptual Shorthand?”

And Man blushed.

And Reason asked again, “What is Number? Is it not a
class of classes: and are not classes themselves thine own convenient
Fictions? Consider the Mountain Top — it Hums not
neither does it Spin. Cease then to listen for the noise of the
humming. Weary not thyself in unravelling the web that hath
never been spun.”

And Man replied “Quite.”

Then sang Reason and Man the Hymn 1923, “Glory to Man
in the Highest for Man is the Master of Words” — nineteen
hundred and twenty-three.

And the sound of the Hymn ringeth yet in our ear.

Thus the Realization of Amoeba ended in the Realization of an
Error.

“God laughed when he made the Sahara,” says an old African
proverb — but Man may yet discover the uses of Dust.295