Vous êtes sur la page 1sur 144

V.

THE MARY FLEXNER LECTURES


ON T H E HUMANITIES
III

T h e s e lectures were delivered at B R Y N M A W R C o L L E G E ,


FEBRUARY and M A R C H 1936 on a fund established by
BERNARD F L E X N E R i n of hlS SlSter
THE PHILOSOPHY
OF RHETORIC

I . A. R I C H A R D S

r . 'f

I '
i
1

A Galaxy Book

NEW YORK

OXFORD UNIVERSITY PRESS

!95
Copyright 1936 by Oxford University Press, Inc.
Copyright renewed 1964 by I.A. Richards
First Published as a Galaxy Book 1965
Printed in the United States of America
PREFACE

" p R E F A C E S , " wrote Bacon, "and passages, and excusa-


tions and other speeches of reference to the person,
are great wastes of t i m e ; and though they seem to
proceed of modesty, they are bravery." T h e invita
t i o n to give the M a r y Flexner Lectures is a greater
honour than these outcomes w i l l justify, and the
pleasures of my visit to B r y n M a w r and of theasso-
ciation w i t h one of the great names of modern
America w h i c h the T i t l e of the Lectureship carries
are personal matters.
B u t I may say a w o r d about the f o r m i n which
these remarks are now offered to the reader's eye
after delivery to an audience's ear. T h e t w o modes
of utterance rarely agree. None the less I have here
kept the w r i t t e n w o r d very close to the spoken,
believing that the occasional air is best suited to the
tentative provisional spirit i n w h i c h this subject
should at present be treated. May anything that
seems extreme i n these lectures be thought acci
dental or be taken as a speaker's device.
I . A . R.
Honolulu, April 7th, 1936
V

CONTENTS
LECTURE PACE

I. INTRODUCTORY 3

II. T H E AlMS O F D l S C O U R S E AND T Y P E S O F

CONTEXT 23

III. T H E lNTERINANIMATION OF W o R D S . 47

IV. SOME CRITERIA OF WORDS 69

V. METAPHOR 89

VI. T H E COMMAND OF M E T A P H O R . . . li5


LECTURE I

INTRODUCTORY
Yet beware of being too material, when there is any
impediment or obstruction in men's wills; for preoccupa
tion of mind ever requireth preface of speech; like a
fomentation to make the unguent enter. Francis Bacon,
Of Dispatch.
fc
LECTURE I

INTRODUCTORY

T HESE lectures are an attempt to revive an o l d


subject. I need spend no time, I t h i n k , i n
describing the present state of Rhetoric. Today
i t is the dreariest and least profitable part of the
waste that the unfortunate travel through i n Fresh
man E n g l i s h ! So low has Rhetoric sunk that we
w o u l d do better just to dismiss i t t o L i m b o than to
trouble ourselves w i t h i t unless we can find reason
for believing that i t can become, a study that w i l l
minister successfully to important needs.
As to the needs, there is l i t t l e r o o m for doubt
about them. Rhetoric, I shall urge, should be a
study of misunderstanding and its remedies. We
struggle a l l our days w i t h misunderstandings, and
no apology is required for any study w h i c h
can prevent or remove them. O f course, inevitably
at present, we have n o measure w i t h w h i c h to calcu
late the extent and degree of our h o u r l y losses i n
communication. One of the aims of these lectures
w i l l be to speculate about some of the measures we
should require i n attempting such estimates.
" H o w m u c h and i n how many ways may good com
munication differ f r o m bad?" T h a t i s t o o b i g and
too complex a question to be answered as i t stands,
b u t we can at least t r y to w o r k towards answering
3
4 T H E PHU-OSOPHY OF R H E T O R I C

some parts of i t ; and these explanations w o u l d be


the revived subject of Rhetoric.
T h o u g h we cannot measure o u r losses i n com
munication we can guess at them. W e even have
professional guessers: teachers and examiners,
whose business is t o guess at and diagnose the mis
takes other people have made i n understanding what
they have heard and read and to avoid illustrating
these mistakes, i f they can, themselves. Another
man who is i n a good position f r o m w h i c h to esti
mate the current losses i n communication is an
author l o o k i n g t h r o u g h a batch of reviews, especially
an author who has been w r i t i n g about some such
subject as economics, social or political theory, or
criticism. I t is not very often that such an author
must honestly admit that his reviewers even when
they profess to agree w i t h h i m have seen his point.
T h a t holds, you may say, only of bad writers who
have w r i t t e n clumsily or t>bscurely. B u t bad
writers are commoner than good and play a
larger part i n bandying notions about i n the
world.
T h e m o r a l f r o m this comes home rather heavily
on a Lecturer addressing an audience o n such a
tangled subject as Rhetoric. I t is l i t t l e use appeal
i n g to the hearer as Berkeley d i d : " I do . . . once
for a l l desire whoever shall t h i n k i t w o r t h his while
to understand . . . that he w o u l d not stick i n this
or that phrase, or manner of expression, b u t
candidly collect m y meaning f r o m the whole sum
and tenor of m y discourse, and laying aside the
INTRODUCTORY 5

words as m u c h as possible, consider the bare notions


themselves. . ."
T h e trouble is that we can only "collect the whole
sum and tenor of the discourse" f r o m the words,
we cannot "lay aside the words"; and as to consider
i n g "the b a r e n o t i o n s themselves," well, I s h a l l be
considering i n a later lecture various notions of a no
tion and comparing their merits for a study of com
munication. Berkeley was f o n d of t a l k i n g about
these "bare notions," these "naked undisguised
ideas," and about "separating f r o m them a l l that
dress and encumbrance of words." B u t an idea or a
notion, when unencumbered and undisguised, is no
easier to get h o l d of than one of those oiled and naked
thieves who infest the railway carriages of India.
Indeed an idea, or a notion, like the physicist's u l t i
mate particles and rays, is only k n o w n by what i t does.
A p a r t frojn its dress or other signs i t is not identi
fiable. Berkeley himself, of course, has his doubts:
"laying aside the words as m u c h as possible, con
sider . . ." T h a t "as m u c h as possible" is not very
m u c h ; and is not nearly enough for the purposes
for w h i c h Berkeley hoped to trust i t .
W e haveinstead t o consider m u c h more closely
how words w o r k i n discourse. B u t before p l u n g i n g
i n t o some of the less w h e l m i n g divisions of this
world-swallowing i n q u i r y , let me glance back for a
few minutes at the traditional treatment of the sub
ject ; m u c h m i g h t be learnt f r o m i t that w o u l d help
us. I t begins, of course, w i t h Aristotle, and may
perhaps be said to end w i t h Archbishop Whately,
6 T H E PHH-OSOPHY OF R H E T O R I C

who wrote a treatise o n Rhetoric for the Encyclo-


pcedia Metropolitana that Coleridge planned. I
may remark, i n passing, that Coleridge's o w n Essay
on Method, the preface to that Encyclopaedia, has
itself more bearing o n a possible future for Rhetoric
than anything I k n o w of i n the official literature.
Whately was a prolific writer, b u t he is most
often remembered now perhaps for an epigram.
" W o m a n , " he said, "is an i r r a t i o n a l animal w h i c h
pokes the fire f r o m the t o p . " I am not q u o t i n g tJiis,
here at B r y n M a w r , to prejudice y o u against the
A r c h b i s h o p : any man, when provoked, m i g h t ven
ture such an unwarrantable and imperceptive gener
alization. B u t I do hope to prejudice you further
against his modes of treating a subject i n w h i c h he is,
according to no less an authority than Jebb, the
best modern w r i t e r . Whately has another epigram
which touches the very heart of our problem, and
may be f o u n d either comforting or f u l l of wicked
possibilities as y o u please: here i t is. "Preachers
nobly a i m at n o t h i n g at a l l and h i t i t ! " W e may
w e l l wonder just what the Archbishop meant by
that.
W h a t we have to surmise is how Whately, fol
l o w i n g and summing u p the whole history of the
subject, can proceed as he d i d 1 H e says quite t r u l y
that "Rhetoric is not one of those branches of study
i n w h i c h we can t r a c e w i t h interest a progressive
improvement f r o m age t o age" ; he goes o n to discuss
"whether Rhetoric be w o r t h any diligent cultiva
t i o n " and to decide, rather half-heartedly, that i t
INTRODUCTORY 7

is provided i t be taken not as an A r t of discourse


b u t as the A r t that is to say, as a philosophic
discipline a i m i n g at a mastery of the fundamental
laws of the use of language, not just a set of dodges
that w i l l be f o u n d to w o r k sometimes. T h a t claim
that Rhetoric must go deep, must take a broad
philosophical view of the principles of the A r t is
the climax of his I n t r o d u c t i o n ; and yet i n the
treatise that follows n o t h i n g of the sort is attempted,
nor is i t i n any other treatise that I k n o w of. W h a t
we are given by Whately instead is a very ably ar
ranged and discussed collection of prudential Rules
about the best sorts of things to say i n various argu
mentative situations, the order i n w h i c h to b r i n g
out your propositions and proofs and examples, at
what p o i n t i t w i l l be most effective to disparage your
opponent, how t o recommend oneself to the audi
ence, and like matters. As to a l l of which, i t is fair
to remark, n o one ever learned about them f r o m a
treatise w h o d i d not know about them already; at
the best, the treatise may be an occasion for realizing
that there is skill to be developed i n discourse, b u t
i t does not and cannot teach the skill. W e can t u r n
o n the whole endeavour the words i n w h i c h the
Archbishop derides his arch-enemy Jeremy Ben-
tham: "the proposed plan for the ready exposure of
each argument resembles that by w h i c h children are
deluded, of catching a b i r d by laying salt o n its t a i l ;
the existing doubts and difficulties of debate being
no greater than, on the proposed system, w o u l d be
found i n determining what Arguments were or were
8 T H E PHILOSOPHY OF R H E T O R I C

not to be classified" i n w h i c h places i n t h e system.


W h y has this happened? I t has happened a l l
through the history of the subject, and I choose
Whately because he represents an inherent tendency
i n its study. W h e n he proceeds f r o m these large-
scale questions of the Ordonnance of arguments to
the m i n u t e particulars of discourse under the
r u b r i c of Style the same t h i n g happens. Instead
of a philosophic i n q u i r y i n t o how words w o r k i n
discourse, we get the usual postcard's-worth of crude
common s e n s e : - b e clear,yet don't be d r y ; be v i
vacious, use metaphors when they w i l l be under
stood not otherwise; respect usage; don't be long-
winded, the other hand don't be gaspy; avoid
a m b i g u i t y ; prefer the energetic t o the elegant; pre
serve u n i t y and coherence. . . I need n o t g o over
to the other side of the postcard. W e a l l know w e l l
enough the maxims that can be extracted by patient
readers out of these agglomerations and h o w h e l p f u l
we have a l l f o u n d them 1

What is w r o n g w i t h these too familiar attempts to


discuss the w o r k i n g of words ? H o w words w o r k is
a matter about w h i c h every u s e r o f language is, of
necessity, avidly curious u n t i l these trivialities choke
the flow of interest. Remembering Whately's re
commendation of metaphor, I can p u t the mistake
best perhaps by saying that a l l they do is to poke the
fire f r o m the top. Instead of tackling, i n earnest,
the problem of how language works at a l l , they
assume that n o t h i n g relevant is to be learnt about
INTRODUCTORY 9

i t ; and that the problem is merely one of disposing


the given and unquestioned powers of words to the
best advantage. Instead of ventilating by i n q u i r y
the sources of the whole action of words, they merely
play w i t h generalizations about their effects, gener
alizations that are uninstructive and u n i m p r o v i n g
unless we go more deeply and by another route i n t o
these grounds. T h e i r conception of the study o f
language, i n b r i e f , i s frustratingly distant or macro
scopic and yields no r e t u r n i n understanding
either practical or theoretical unless i t is supple
mented by an intimate or microscopic i n q u i r y which
endeavours t o look into the structure of the mean
ings w i t h w h i c h discourse is composed, not merely
i n t o the effects of various large-scale disposals of
these meanings. I n thisRhetoricians may r e m i n d
us of the Alchemists' efforts to transmute common
substances into precious metals, vain efforts because
they were not able to take account of the internal
structures of the so-called elements.
T h e comparison that I am using here is one which
a modern w r i t e r on language can hardly avoid. T o
account for understanding and misunderstanding,
to study the efficiency of language and its conditions,
we have to renounce, for a while, the view that
words just have their meanings and that what a dis
course does is to be explained as a composition of
these meanings as a w a l l can be represented as a
composition of its bricks. W e have to shift the
focus of our analysis and attempt a deeper and
more m i n u t e grasp and try to take account o f t h e
io T H E PHH-OSOPHY OF R H E T O R I C

structures of the smallest discussable units of mean


i n g and the ways i n which these vary as they are p u t
w i t h other units. Bricks, for a l l practical purposes,
hardly m i n d what other things they are p u t w i t h .
Meanings m i n d intenselymore indeed than any
other sorts of things. I t is the peculiarity of mean
ings that they do so m i n d their company; that is i n
part what we mean by calling them meanings 1 I n
themselves they are n o t h i n g figments, abstractions,
unreal things that we invent, i f you like b u t we
invent them for a purpose. T h e y help us to avoid
taking account of the peculiar way i n which any
part of a discourse, i n the last resort, does what i t
does only because the other parts of the surround
ing, uttered or unuttered discourse and its condi
tions are what they are. " I n the last r e s o r t " - t h e
last resort here is mercifully a l o n g way off and very
deep down. Short of i t we are aware of certain
stabilities w h i c h hide f r o m us this universal rela
t i v i t y or, better, interdependence of meanings.
Some words and sentences still more, do seem to
mean what they mean absolutely and uncondition
ally. T h i s is because the conditions governing their
meanings are so constant that we can disregard them.
So the weight of a cubic centimeter of water seems a
fixed and absolute t h i n g because of the constancy of
its governing conditions. I n weighing out a pound
of tea we can forget about the mass of the earth.
A n d w i t h words w h i c h have constant conditions the
common sense view that they have fixed proper
meanings, w h i c h should be learned and observed, is
INTRODUCTORY

justified. B u t these words are fewer than we sup


pose. Most words, as they pass f r o m context to
context, change their meanings; and i n many dif
ferent ways. I t is their d u t y and their service to us
to do so. O r d i n a r y discourse w o u l d suffer anchy
losis i f they d i d not, and so far we have no ground
for complaint. W e are extraordinarily skilful i n
some fields w i t h these shifts of sense especially
when they are of the k i n d we recognize officially as
metaphor. B u t o u r skill fails; i t is patchy and
fluctuant; and, when i t fails, misunderstanding of
others and of ourselves comes i n .
A chief cause of misunderstanding, I shall argue
later, is the Proper Meaning Superstition. T h a t is,
the common beliefencouraged officially by what
lingers o n i n the school manuals as Rhetoric that
a w o r d has a meaning of its o w n (ideally, only one)
independent of and c o n t r o l l i n g its use and the
purpose for w h i c h i t should be uttered. T h i s
superstition is a recognition of a certain k i n d of
stability i n the meanings of certain words. I t is
o n l y a superstition when i t forgets (as i t commonly
does) that the stability of the meaning of a w o r d
comes f r o m the constancy of the contexts that give
i t its meaning. Stability i n a word's meaning is not
something t o be assumed, b u t always something to
be explained. A n d as we t r y out explanations, we
discover, of course, that as there are many sorts
of constant contexts there are many sorts of stabili
ties. T h e stability of the meaning of a w o r d like
knife, say, is different f r o m the stability of a w o r d like
1* T H E PHILOSOPHY OF R H E T O R I C

mass i n its technicaluse, and then again b o t h differ


f r o m the stabilities of such words, say, as event, ingres-
sion, endurance, recurrence, or object, i n the para
graphs of a very distinguished predecessor i n this
Lectureship. I t w i l l have been noticed perhaps that
the way I propose to treat meanings has its analogues
w i t h M r . Whitehead's treatment of things. B u t i n
deed no one to w h o m Berkeley has mattered w i l l be
very confident as to which is which.
I have been suggestingwith m y talk of macro
scopic and microscopic inquiries that the theory
of language may have something t o learn, not m u c h
b u t a l i t t l e , f r o m the ways i n w h i c h the physicist
envisages stabilities. B u t m u c h closer analogies are
possible w i t h some of thepatterns of Biology. The
theory of interpretation is obviously a branch of
biology a branch that has not grown very far
or very healthily yet. T o remember this may help
us t o avoid some t r a d i t i o n a l mistakes among them
the use of bad analogies w h i c h tie us u p i f we take
them too seriously. Some of these are n o t o r i o u s ;
for example, the opposition between f o r m and con
tent, and the almost equivalent opposition between
matter and f o r m . These are wretchedly incon
venient metaphors. So is that other w h i c h makes
language a dress w h i c h thought puts on. W e shall
do better to t h i n k of a meaning as though i t were a
plant that has grown not a can that has been filled
or a l u m p of clay that has been moulded. These
are obvious inadequacies ; but, as the history of c r i t i
cism shows, they have not been avoided, and the
INTRODUCTORY 13

perennial efforts of the reflective to amend or sur


pass them Croce is the extreme modern example
hardly help.
M o r e insidious and more devastating are the over-
simple mechanical analogies w h i c h have been
brought i n under the heading of Associationism i n
the hope of explaining how language works. A n d
thought as well. T h e two problems are close to
gether and similar and neither can be discussed prof
itably apart f r o m the other. But, unless we drasti
cally remake their definitions, and thereby dodge the
m a i n problems, Language and T h o u g h t are not
need I sayP one and the same. I suppose I must,
since the Behaviorists have so l o u d l y averred that
T h o u g h t is sub-vocal talking. T h a t however is a
doctrine I prefer, i n these lectures, to attack by i m p l i
cation. T o discuss i t explicitly w o u l d take time that
can, I t h i n k , be spent more f r u i t f u l l y . I w i l l only
say that I h o l d that any doctrine identifying T h o u g h t
w i t h muscular movement is a self-refutation of the
observationalism that p r o m p t s i t heroic and fatal.
A n d that an identification of T h o u g h t w i t h an activ
i t y of the nervous system is to me an acceptable hy
pothesis, b u t too large to have interesting applica
tions. I t may be left u n t i l more is k n o w n about
b o t h ; when possibly i t may be developed to a point
at w h i c h i t m i g h t become useful. A t present i t is
still T h o u g h t which is most accessible to study and
accessible largely through Language. W e can a l l de
tect a difference i n our own minds between t h i n k i n g
of a dog and t h i n k i n g of a cat. B u t no neurologist
14 T H E PHH.OSOPHY O F R H E T O R I C

can. Even when no cats or dogs are about and


we are d o i n g n o t h i n g about them except t h i n k i n g
of them, the difference is p l a i n l y perceptible. W e
can also say 'dog' and t h i n k 'cat.'
I must, though, discuss the doctrine of associa
tions briefly, because when we ask ourselves about
how words mean, some theory about trains of asso
ciated ideas or accompanying images is certain to
occur to us as an answer. A n d u n t i l we see how
l i t t l e distance these theories take us they are frus
trating. W e a l l k n o w the outline of these theories:
we learn what the w o r d 'cat' means by seeing a cat
at the same t i m e that we hear the w o r d 'cat' and
thus a l i n k is formed between the sight and the
sound. N e x t t i m e we hear the w o r d 'cat' an image
of a cat (a visual image, let us say) arises i n the m i n d ,
and that is how the w o r d 'cat' means a cat. T h e
obvious objections that come f r o m the differences
between cats; f r o m the fact that images of a grey
persian asleep and of a tabby stalking are very dif
ferent, and f r o m some people saying they never have
any imagery, must then be taken account of, and
the theory grows very complex. Usually, images
get relegated t o a background and become mere
supports to something hard to be precise about
an idea of a c a t w h i c h is supposed then to be
associated w i t h the w o r d 'cat' m u c h as the image
originally was supposed to be associated w i t h i t .
T h i s classical theory of meaning has been under
heavy firefrom many sides for more than a century
f r o m positions as different as those of Coleridge,
INTRODUCTORY 15

of Bradley, of Pavlov and of the gestalt psychologists.


I n response i t has elaborated itself, calling i n the a i d
of the conditioned-reflex and s u b m i t t i n g to the i n
fluence of Freud. I do not say that i t is incapable,
when amended, of supplying us w i t h a workable
theory of m e a n i n g i n fact, i n the next lecture I
shall sketch an o u t l i n e theory of how words mean
w h i c h has associationism among its obvious an
cestors. A n d here, i n saying that simple associa
tionism does not go far enough and is an impediment
unless we see this, I am merely r e m i n d i n g y o u that a
clustering of associated images and ideas about a
w o r d i n the m i n d does not answer o u r question:
" H o w does a w o r d mean?" I t only hands i t o n to
them, and the question becomes: " H o w does an idea
(or an image) mean what i t does?" T o answer that
we have to go outside the m i n d and i n q u i r e i n t o its
connections w i t h what are not mental occurrences.
O r ( i f you prefer, instead, t o extend the sense of the
w o r d ' m i n d ' ) we have to i n q u i r e i n t o connections
between events w h i c h were left out by the traditional
associationism. A n d i n leaving them out they left
out the p r o b l e m .
For o u r purposes here the important points are
two. First, that ordinary, current, undeveloped
associationism is r u i n e d by the crude inapposite
physical metaphor of impressions stamped o n the
m i n d (the image of the cat stamped by the cat), i m
pressions then l i n k e d and combined i n clusters like
atoms i n molecules. T h a t metaphor gives us no
useful account either of perception o r o f reflection,
i6 T H E PHILOSOPHY OF R H E T O R I C

and we shall not be able to t h i n k i n t o or t h i n k out


any of the interesting problems of Rhetoric unless
we improve i t .
Secondly the appeal t o imagery as constituting the
meaning of an utterance has, i n fact, frustrated a
large part of the great efforts that have been made by
very able people ever since the 17th Century to p u t
Rhetoric back into the i m p o r t a n t place i t deserves
among our studies. Let me give you an example.
Here is L o r d Kames who, as a Judge of the Court
of Session i n Scotland, was not w i t h o u t a reputation
for shrewdness being, I believe, really remarkably
silly.
I n Henry V (Act I V , scene I ) W i l l i a m s i n a fume
says this of what "a poor and private displeasure can
do against a m o n a r c h " : " Y o u may as w e l l go about
to t u r n the sun to ice w i t h fanning i n his face w i t h
a peacock's feather." L o r d Kames comments, " T h e
peacock's feather, not to m e n t i o n the beauty of the
object, completes the image: an accurate image can
not be formed of that fanciful operation w i t h o u t
conceiving a particular feather; and one is at a loss
when this is neglected i n the description." (Ele-
ments of Criticism, p. 372.)
T h a t shows, I t h i n k , what the imagery obsession
can do t o a reader. W h o i n the w o r l d , apart f r o m
a theory, w o u l d be "at a loss" unless the sort of
feather we are to fan the sun's face w i t h is specified?
I f we cared to be sillier than o u r author, we could
pursue h i m o n his theory, by asking whether i t is to
be a long or a short feather or whether the sun is at
INTRODUCTORY 17

its height or setting? T h e whole theory that the


p o i n t of Shakespeare's specification is to "complete
the image," i n Kames' sense, is utterly mistaken and
misleading. W h a t peacock does, i n the context
there, is obviously to b r i n g i n considerations that
heighten the idleness, the vanity, i n W i l l i a m s ' eyes,
of "poor and private displeasures against a monarch."
A peacock's feather is something one m i g h t flatter
oneself w i t h . H e n r y has said that i f the K i n g lets
himself be ransomed he w i l l never trust his w o r d
after. A n d W i l l i a m s is saying, " Y o u ' l l never trust
his w o r d after! What's t h a t ! Plume yourself
u p o n i t as m u c h as you like, b u t what w i l l that do
to the k i n g ! "
L o r d Kames i n 1761, blandly enjoying the beauty
and completeness of the lively and distinct and accu
rate image of the feather he has produced for h i m
self, and thereby missing, i t seems, the whole tenor
of the passage, is a spectacle w o r t h some attention.
I shall be r e t u r n i n g to L o r d Kames, i n a later lec
ture, when I discuss metaphor. His theories about
trains of ideas and images are typical 18th Century
Associationism the Associationism of which David
Hartley is the great prophet a n d t h e applications
of these theories i n the detail of Rhetoric are their
o w n refutation. W e have to go beyond these theo
ries, b u t however mistaken they may be, or however
absurd their outcome may sometimes seem, we must
not forget that they are beginnings, first steps i n a
great and novel venture, the attempt to explain i n
detail how language works and w i t h i t t o improve
i8 T H E PHILOSOPHY OF R H E T O R I C

communication. As such, these attempts m e r i t the


most discerning and the most sympathetic eye that
we can t u r n u p o n them. Indeed, i t is impossible
to read Hartley, for example, w i t h o u t deep sympathy
if we realize what a task i t was that he was attempt
ing. N o t only when he writes, i n his conclusion, i n
words w h i c h speak the thoughts of every candid i n
quirer : " T h i s is by no means a f u l l or satisfactory
Account of the Ideas w h i c h adhere t o words by
Association. For the A u t h o r perceives himself to
be still a mere novice i n these speculations; and i t is
difficult to explain Words to the B o t t o m by Words ;
perhaps impossible." ( O n M a n , 277.) B u t s t i l l
more when he says : " A l l that has been delivered by
the Ancients and Moderns, concerning the power of
H a b i t , Custom, Example, Education, A u t h o r i t y ,
Party-prejudice, the Manner of learning the manual
and liberal Arts, Etc., goes u p o n this Doctrine as its
foundation, a n d m a y be considered as the detail of
i t , i n various circumstances. I hope here t o begin
w i t h the simplest case, and shall proceed to more and
more complex ones continually, t i l l I have exhausted
what has occurred to me o n this Subject." (On
M a n , p. 67.)
T h e man who wrote that was not ' p o k i n g the fire
f r o m the top.' H i s way of v e n t i l a t i n g the subject
may not have been perfectly advised, b u t he saw what
needed d o i n g and i t is no wonder that Coleridge for
a while admired Hartley beyond a l l other men. For
u p o n the f o r m a t i o n and transformations of mean
ings w h i c h we must study w i t h and t h r o u g h words
INTRODUCTORY 19

a l l that Hartley mentions, and m u c h more, goes as


its foundation. For i t is no exaggeration to say that
the fabrics of all our various worlds are the fabrics
of our meanings. I began, you recall, w i t h Berkeley,
w i t h to use M r . Yeats' noble lines

G o d appointed Berkeley who proved a l l things a dream,


T h a t this preposterous pragmatical pig of a world, its
farrow that so solid seem,
M u s t vanish o n the instant d i d the m i n d but change its
theme.

Whatever we may be studying we do so only


through the growth of our meanings. T o realize
this turns some parts of this attempted direct study
of the modes of growth and interaction between
meanings, which might otherwise seem a niggling
philosophic juggle w i t h distinctions, i n t o a business
of great practical importance. For this study is
theoretical only that i t may become practical. Here
is the paragraph i n which Hobbes condenses what he
had learnt f r o m his master, Bacon:

" T h e e n d or scope of philosophy is, that we may


make use to our benefit of effects formerly seen, or
that, by the application of bodies to one another, we
may produce the l i k e effects of those we conceive i n our
m i n d , as far forth as matter, strength a n d industry,
w i l l permit, for the commodity of h u m a n life. F o r the
i n w a r d glory a n d t r i u m p h of m i n d that a m a n may
have for the mastery of some difficult a n d doubtful
matter, or for the discovery of some hidden truth, is
not worth so m u c h pains as the study of Philosophy
r e q u i r e s ; nor need any m a n care m u c h to teach an-
other what he knows himself, if h e think that w i l l be
the only benefit of his labour. T h e end of knowledge
80 T H E PHILOSOPHY OF RHETORIC
is p o w e r ; a n d the use of theorems (which, among
geometricians, serve for the finding out of properties)
is for the construction of p r o b l e m s ; and, lastly, the
scope of a l l speculation is the performance of some
action, or thing to be done."

I shall go o n then, i n the next Lecture, by the use


of theorems to the construction of problems, w i t h o u t
further insisting that these problems are those u p o n
which, w i t t i n g l y and u n w i t t i n g l y , we spend our
whole waking life.
LECTURE I I

T H E A I M S O F D I S C O U R S E A N D TYPES
OF C O N T E X T
I repeat, however, that there is a prime part of educa
tion, an element of the basis itself, in regard to which I
shall probably remain within the bounds of safety in de
claring that no explicit, no separate, no adequate plea
will be likely to have ranged itself under any one of your
customary heads of commemoration. Henry James, A
Commemoration Address at Bryn Mawr on The Question
of our Speech.
LECTURE II

T H E A I M S O F D I S C O U R S E A N D TYPES
OF C O N T E X T

I N M Y introductory lecture I urged that there is


r o o m for a persistent, systematic, detailed i n
q u i r y i n t o how words w o r k that w i l l take the place
of the discredited subject w h i c h goes by the name of
Rhetoric. I went on to argue that this i n q u i r y must
be philosophic, or i f you hesitate w i t h that word,
I do myselfthat i t must take charge of the criticism
of its o w n assumptions and not accept them, more
than i t can help, ready-made f r o m other studies.
H o w words mean, is not a question to which we can
safely accept an answer either as an inheritance f r o m
common sense, that curious growth, or as some
t h i n g vouched for by another science, by psychology,
say since other sciences use words themselves and
not least delusively when they address themselves
to these questions. T h e result is that a revived
Rhetoric, or study of verbal understanding and mis
understanding, must itself undertake its own i n q u i r y
i n t o the modes of m e a n i n g n o t only, as w i t h the
o l d Rhetoric, o n a macroscopic scale, discussing the
effects of different disposals of large parts of a dis
coursebut also on a microscopic scale by using
theorems about the structure of the fundamental
conjectural units of meaning and the conditions
23
24 T H E PHILOSOPHY OF R H E T O R I C

through w h i c h they, and their interconnections,


arise.
I n the o l d Rhetoric, of course, there is m u c h that
a new Rhetoric finds useful and m u c h besides
which may be advantageous u n t i l man changes his
nature, debates and disputes, incites, tricks, bullies
and cajoles his fellows less. Aristotle's notes o n the
forensic treatment of evidence elicited under torture
are unhappily not w i t h o u t their u t i l i t y still i n some
very up-to-date parts of the w o r l d .
A m o n g the general themes of the o l d Rhetoric
there is one w h i c h is especially pertinent to our i n
q u i r y . T h e o l d Rhetoric was an offspring of dis
pute ; i t developed as the rationale of pleadings and
persuadings; i t was the theory of the battle of words
and has always been itself dominated by the com
bative impulse. Perhaps what i t has most t o teach
us is the narrowing and b l i n d i n g influence of that
preoccupation, that debaters' interest.
Persuasion is only one among the aims of dis
course. I t poaches on the others especially on that
of exposition, w h i c h is concerned to state a view,
not to persuade people to agree or to do any
t h i n g more than examine i t . T h e review and cor
respondence columns of the learned and scientific
journals are the places i n w h i c h to watch this poach
ing at its liveliest. I t is no bad preparation for any
attempt at exposition above a l l of such debatable
and contentious matters as those to which I am soon
to t u r n to realize how easily the combative impulse
can put us i n mental blinkers and make us take an-
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 25

other man's words i n the ways i n which we can down


h i m w i t h least trouble.
I can p o i n t this moral call i t defensive i f you w i l l
w i t h a small specimen f r o m one of the many l i t t l e
books which i n the Nineteenth Century attempted
a reform of Rhetoric. I t is f r o m Benjamin H u m
phrey Smart's Practical Logic, a l i t t l e book w r i t t e n
for and used for a few decades i n the best young
ladies' seminaries through the middle of the Nine
teenth Century and now as dead as any book well
can be. Smart is discussing the conduct of exposi
t i o n . H e has listed a number of faults commonly
committed and comes to the
T E N T H F A U L T T o B E A v o i D E D , namely: Forget
ting the Proposition.
" O f this error," he writes, "the following instance
may suffice :

'Anger has been called a short madness; and people


of the weakest understanding are the most subject to it.
I t is remarkable that w h e n a disputant is i n the wrong,
he tries to m a k e u p i n violence what he wants i n
argument. T h i s arises from his pride. H e w i l l not
o w n his error, a n d because he is determined not to be
convicted of it, he falls into a passion.'

Here, (Smart comments) instead of going on to show


why Anger has been called a short madness, the
w r i t e r wanders i n t o reflections which have no neces
sary connection w i t h the particular proposition.
He should have reasoned t h u s :

'Anger has been called a short madness. T o be con-


vinced that the appellation is just, let us look to the
26 T H E PHILOSOPHY OF R H E T O R I C
effects of anger. I t disturbs a man's judgment, so that
he inflicts a n injury on his dearest friend, who, the
next moment, he loads w i t h caresses. I t makes h i m
r u n headlong into d a n g e r s , w h i c h , if his m i n d were
clear, h e w o u l d be the first to see a n d avoid. I t is true
that anger does not always disturb the m i n d to this
degree, but that it always disturbs the m i n d i n a degree
proportional to its violence, is c e r t a i n ; a n d therefore
it may be justly characterised as a m a d n e s s . ' "

W h a t necessary connection w i t h the proposition,


may we ask, has this sketch of some scenes f r o m an
early V i c t o r i a n Novel? A n d whence comes this
certainty that anger always disturbs the m i n d i n a
degree proportional to its violence ? However, i t is
better perhaps to take its lesson to heart and remem
ber that anger is not the only w a r p i n g passion.
R i s i b i l i t y and tedium, too, I t h i n k Smart w o u l d have
said, can disturb the judgment.
W a r n e d now of the dangers b o t h of forgetting the
proposition and of the 'short madness' that the com
bative and other passions induce, let me sketch, to
use Hobbes' words, a theorem about meanings w h i c h
may be useful i n constructing the most general prob
lems of a new Rhetoric.
I had better p u t i n another warning, though, here.
W h a t follows is unavoidably abstract and general i n
the extreme. I t may therefore rather illustrate the
difficulties o f c o m m u n i c a t i n g w i t h such highly ab
stract language than achieve as m u c h communica
t i o n as we w o u l d wish. I f so the fault w i l l not lie, I
hope and believe, either i n my stupidity or i n our
j o i n t stupidity. I t w i l l lie i n the abstractness of the
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 27

language. I t has to be abstract here. W h a t i t is


t r y i n g to say cannot, I t h i n k , be p u t safely i n more
concrete terms, for i t is not t a l k i n g about this or that
mode of meaning b u t about a l l meanings. A n d I
cannot here start w i t h illustrations, because all things
equally illustrate what I am saying; and how they
are to be taken is just the problem. But, after this
bout of abstractions, the applications I shall be mak
ing i n the later Lectures w i l l , I believe, clear u p this
dark patch. I n brief, how we use this theorem best
shows us what the theorem is.
If, then, you seem i n the next half hour at times
merely to be hearing words as sounds that come and
go, I must beg your indulgence, or buy i t w i t h the
promise that we shall come out again to practical
problems i n the everyday conduct of words. Mean
while this very difficulty is an illustration of a chief
practical problem.
W h a t I am now going to t r y to say is something
which, i f i t is r i g h t , w e a l l i n a sense know extremely
well already. " I t is not sufficiently considered," said
D r . Johnson, "that men more frequently require
to be reminded than i n f o r m e d . " I shall be t r y i n g to
r e m i n d you of something so simple that i t is hard
to t h i n k of. Something as simple as possible, and, to
quote Hobbes again, "clear and perspicuous to all
men save only to those who studying the hard
writings of the metaphysicians, which they believe
to be some egregious learning, t h i n k they under
stand not when they do." A n d i t may be comfort
i n g to recall that Lotze began a course of lectures
s8 T H E PHILOSOPHY OF R H E T O R I C

on an allied subject by saying that " T h e simplest


of the conceptions here employed, that of a t h i n g
and that of its being, however l u c i d they appear
at first, on closer consideration grow always more
and more obscure." For 'always' I w o u l d say
'for a time.' W e r e t u r n to lucidity. B u t now to
work.
I have two sets of problems i n v i e w : one set I have
just been t a l k i n g about the division of the various
aims of discourse, the purposes for w h i c h we speak
or w r i t e ; i n brief, the functions of language. The
other set of problems goes deeper, and, i f we can set
i t rightly, the problems about the language func
tions are best approached f r o m i t . I can indicate
these deeper problems i n many ways: W h a t is the
connection between the m i n d and the w o r l d by
which events i n the m i n d mean other events i n the
w o r l d ? O r " H o w does a thought come to be ' o f
whatever i t is that i t is a thought of ?" or " W h a t is
the relation between a t h i n g and its name ?" The
last indication may not seem t o carry as far as the
others; b u t they are a l l the same problem and I p u t
the 'name'-formulation i n because an over-simple
view of naming, or rather a treatment of words i n
general as though they were names (usually of ideas)
has been a m a i n defect i n the traditional study.
These are, you w i l l see, really deep problems. As
such we shall not expect any answers w h i c h w i l l be
satisfactory. W e must be content i f the answers we
get are to some degree useful useful among other
things i n i m p r o v i n g themselves.
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 9

I can start the theorem safely by remarking that


we are things peculiarly responsive to other things.
T o develop this we have to consider the peculiari
ties of our responsiveness. W e are responsive i n a l l
sorts of ways. Some of these ways are relatively
simple, i f cut short enough; as when we j u m p at a
l o u d noise or respond to changes of temperature.
B u t even here, i f we compare ourselves to ther
mometers, we see that our responses are of a different
order of complexity. A thermometer responds, the
length of its thread of mercury varies w i t h the tem
perature, b u t only w i t h the present temperature
unless the thermometer is a bad one. W h a t has
happened to i t i n the past, what temperatures i t
formerly recorded, and the order i n which i t re
corded them, a l l that has no bearing upon and does
not interfere w i t h its present response to changes of
temperature. W e can imagine, though, a ther
mometer that, whenever the temperature went u p
and down like this, M , d i d something that could
only be explained by b r i n g i n g i n other things that
happened to i t i n the past when the temperature
went u p and down so, M A n d correspondingly
d i d something else whenever the temperature went
down and up, W Such an imaginary thermom
eter w o u l d be o n the way to showing character
istics of the behavior of l i v i n g systems, of the systems
which, we say, have a m i n d .
N o w consider our o w n minds' simplest operations.
D o we ever respond to a stimulus i n a way which is
not influenced by the other things that happened to
jo T H E PHU,OSOPHY OF R H E T O R I C

us when more or less similar s t i m u l i struck us i n the


past? Probably never. A new k i n d of stimulus
might perhaps give rise to a new k i n d of sensation,
a new k i n d of pain, say. B u t even so we should
probably recognize i t as a pain of some sort. Effects
from more or less similar happenings i n the past
w o u l d come i n to give our response its character and
this as far as i t went w o u l d be meaning. Meaning
of a lowly k i n d , no doubt, the k i n d of meaning that
the least developed animals live by. I t is important
and that is why I have started so. far back w i t h
these elementaries to realize how far back into the
past a l l our meanings go, how they grow o u t of one
another m u c h as an organism grows, and how i n
separable they are f r o m one another.
I can make the same p o i n t by denying that we
have any sensations. T h a t sounds drastic b u t is
almost certainly true i f r i g h t l y understood. A sen
sation w o u l d be something that just was so, o n its
own, a d a t u m ; as such we have none. Instead we
have perceptions, responses whose character comes
to them f r o m the past as w e l l as the present occasion.
A perception is never just of an it; perception
takes whatever i t perceives as a t h i n g of a certain
sort. A l l t h i n k i n g f r o m the lowest to the highest
whatever else i t may be is sorting.
T h a t is an i m p o r t a n t part of the theorem because
it removes, i f i t is accepted, one of the worst troubles
which have distorted traditional accounts of the
meanings of words the troubles that gave rise to
the Nominalist, Realist, Conceptual controversies
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 31

best k n o w n to us t h r o u g h the great B r i t i s h philo


sophical battle of the Eighteenth Century about
whether we have and how we come by abstract ideas
and what they are. T h i s theorem alleges that
meanings, f r o m the very beginning, have a primor
dial generality and abstractness; and i t follows W i l
l i a m James i n saying that the lowliest organism a
polyp or an amoeba i f i t learns at a l l f r o m its past,
i f i t exclaims i n its acts, " H a l l o ! Thingembob
a g a i n ! " thereby shows itself to be a conceptual
thinker. I t is behaving or t h i n k i n g w i t h a concept
not, of course, of one. Its act is abstractive and
general; disregards i n some respects the former sit
uations and so is abstractive, and applies i n some
respects not to one single t h i n g b u t to any of a sort
and so is general.
T h e theorem settles the Eighteenth Century prob
lem by standing i t o n its head. T h a t p r o b l e m was,
H o w do wemanage, f r o m this particular concrete
t h i n g and that particular concrete t h i n g and the
other particular concrete thing, t o arrive at the gen
eral abstract anything ? T h e theorem holds that we
begin w i t h the general abstract anything, split i t , as
the w o r l d makes us, into sorts and then arrive at
concrete particulars by the overlapping or common
membership of these sorts. T h i s b i t of paper here
now i n m y hand is a concrete particular to us sofar
as we t h i n k of i t as paperish, hereish, nowish and
i n m y h a n d ; i t is the more concrete as we take i t as
of more sorts, and the more specific as the sorts are
narrower and more exclusive.
32 T H E PHILOSOPHY OF R H E T O R I C

T h e next step i n the theorem takes us o n to words


and their meanings. I f we sum u p thus far by say
i n g that meaning is delegated efficacy, that descrip
t i o n applies above a l l to the meaning of words, whose
virtue is to be substitutes exerting the powers of
what is not there. T h e y do this as other signs do
i t , though i n more complex fashions, t h r o u g h t h e i r
contexts.
I must explain now the rather special and tech
nical sense I am giving to this w o r d 'context.'
T h i s is the pivotal p o i n t of the whole theorem.
T h e w o r d has a familiar sense i n 'a literary context/
as the other words before and after a given w o r d
w h i c h determine how i t is to be interpreted. T h i s
is easily extended to cover the rest of the book. I
recall the painful shock I suffered when I first came
across, i n a book by D r . Bosanquet, what he called
the Golden Rule of Scholarship, "Never to quote or
comment on anything i n a book w h i c h you have not
read f r o m cover to cover." As w i t h other Golden
Rules a strange peace w o u l d fall u p o n the w o r l d i f
that were observed. I cannot honestly say I either
practice the Rule or recommend i t . T h e r e is a
middle way wiser for the C h i l d r e n of this W o r l d .
However, as I neither am nor hope to be a scholar, I
have no occasion to practise i t .
T h e familiar sense of 'context' can be extended
further to include the circumstances under w h i c h
anything was w r i t t e n or said; wider s t i l l to include,
for a w o r d i n Shakespeare, say, the other k n o w n uses
of the w o r d about that time, wider still finally t o
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 33

include anything whatever about the period, or


about anything else w h i c h is relevant to o u r inter
pretation of i t . T h e technical use I am going to
make of this term 'context' is none of these
though i t has something i n common w i t h them as
having to do w i t h the governing conditions of an i n
terpretation. W e can get to i t best, perhaps, by
considering those recurrences i n nature w h i c h state
ments of causal laws are about.
Put very simply, a causal law may be taken as say
i n g that, under certain conditions, of two events i f
one happens the other does. W e usually call the
first the cause and the second the effect, b u t the two
may happen together, as when I clap my hands and
b o t h palms tingle. I f we are talking about final
causes we reverse them, and the lecture you are going
to hear was the cause of your coming hither. T h e r e
is a good deal of arbitrariness at several points here
w h i c h comes f r o m the different purposes for w h i c h
we need causal laws. W e decide, to suit these pur
poses, how we shall divide u p events; we make the
existence of the earth one event and the tick of a
clock another, and so on. A n d we distribute the
titles of 'cause' and 'effect' as we please. T h u s
we do not please to say that night causes day or day
night. W e prefer to say that given the conditions
the rotation of the earth is the cause of their suc
cession. W e are especially arbitrary i n p i c k i n g out
the cause f r o m among the whole group, or context, of
conditions of p r i o r and subsequent events which
hang together. T h u s the coroner decides that the
34 T H E PHILOSOPHY OF R H E T O R I C

cause of a man's death was the act of a murderer and


not the man's meeting w i t h the murderer, or the
stopping of his heart, or the fact that he was not
wearing a bullet-proof waistcoat. T h a t is because
the coroner is interested i n certain kinds of causal
laws b u t not i n others. So here, i n sketching this
causal theorem of meaning, I am interested only i n
certain kinds of law and am not necessarilysaying
anything about others.
N o w for the sense of 'context.' Most generally
i t is a name for a whole cluster of events that recur
together i n c l u d i n g the required conditions as w e l l
as whatever we may pick out as cause or effect. B u t
the modes of causal recurrence on w h i c h meaning
depends are peculiar t h r o u g h that delegated efficacy
I have been t a l k i n g about. I n these contexts one
item typically a w o r d takes over the duties of
parts w h i c h can then be o m i t t e d f r o m the recur
rence. T h e r e is thus an abridgement of the con
text only shown i n the behavior of l i v i n g things, and
most extensively and drastically shown by man.
W h e n this abridgement happens, what the sign or
w o r d the item w i t h these delegated powers
means is the missing parts of the context.
I f we ask how this abridgement happens, how a
sign comes t o stand for an absent cause and condi
tions, we come u p against the limits of knowledge
at once. N o one knows. Physiological speculation
has made very l i t t l e progress towards explaining that,
though enormous strides have been made this cen
t u r y i n analysing the complexities of the conditioned
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 35

reflex. T h e shift, the handing over, is left still as


inexplicable. Probably this 'learning problem' goes
d o w n as deep as the nature of life itself. W e can
suppose, i f we like, that some sorts of residual effects
are left behind f r o m former occurrences w h i c h later
co-operate w i t h the sign i n determining the response.
T o do so is to use a metaphor drawn f r o m the gross
behavior, taken macroscopically, of systems that are
not l i v i n g p r i n t e d things, gramaphone records and
such. W e can be fairly ingenious w i t h these meta
phors, invent neural archives storing u p impressions,
or neural telephone exchanges w i t h fantastic prop
erties. B u t how the archives get consulted or how i n
the telephone system A gets on to the B i t needs, i n
stead of to the whole alphabet at once i n a j u m b l e ,
remain utterly mysterious matters.
Fortunately linguistics and the theory of meaning
need not wait u n t i l this is remedied. T h e y can
probably go m u c h further than we have yet imagined
w i t h o u t any answer to this question. I t is enogh
for our purposes to say that w h a t a w o r d means is
the missing parts of the contexts f r o m which i t draws
its delegated efficacy.
A t this p o i n t I must r e m i n d you of what I said
a f e w m i n u t e s ago about the p r i m o r d i a l generality
and abstractness of meaning and about how, when
we mean the simplest-seeming concrete object, its
concreteness comes to i t f r o m the way i n which we
are b r i n g i n g i t simultaneously i n t o a number of
sorts. T h e sorts grow together i n i t to form that
meaning. T h e o r y here, as so often, can merely
36 T H E PHILOSOPHY OF R H E T O R I C

exploit the etymological h i n t given i n the w o r d


'concrete.'
I f we forget this and suppose that we start w i t h
discrete impressions of particulars ('fixities and defi-
nites' as Coleridge called them) and then b u i l d these
u p i n t o congeries, the theorem I am recommending
collapses at once i n t o contradictions and absurdities.
T h a t was the fault of the o l d H a r t l e i a n Association
ism I complained of last time. I t d i d not go back
far enough, i t took particular impressions as its
i n i t i a l terms. B u t the i n i t i a l terms for this theorem
are not impressions; they are sortings, recognitions,
laws of response, recurrences of like behaviors.
A particular impression is already a product of
concrescence. Behind, or i n i t , there has been a
coming together of sortings. W h e n we take a n u m
ber of particular impressions of a n u m b e r of dif
ferent white things, say and abstract f r o m them an
idea of whiteness, we are explicitly reversing a
process w h i c h has already been i m p l i c i t l y at w o r k i n
our perception of them as a l l white. O u r risk is to
confuse the abstractness we thus arrive at intellectu
ally w i t h the p r i m o r d i a l abstractness out of w h i c h
these impressions have already grown before ever
any conscious explicit reflection took place.
Things, i n brief, are instances of laws. As Bradley
said, association marries only universals, and out
of these laws, these recurrent likenessess of behavior,
i n o u r minds and i n the w o r l d not out of revived
duplicates of i n d i v i d u a l past impressions the fabric
of our meanings, w h i c h is the w o r l d , is composed.
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 37

So m u c h for the theorem. W h a t are the prob


lems we must use i t to construct ?
Since the whole business of Rhetoric comes down
to comparisons between the meanings of words, the
first problem, I think, should be this. H o w , i f the
meaning of a w o r d is, i n this sense, the missing parts
of its contexts, how then should we compare the
meanings of two words? T h e r e is o p p o r t u n i t y for
a grand misunderstanding here. I t is not proposed
that we should t r y to make these comparisons by a
process of discovering, detailing, and then compar
i n g these missing parts. W e could not do i t and, i f
we could, i t w o u l d be waste of time. T h e theorem
does not pretend to give us quite new ways of dis
tinguishing between meanings. I t only bars out
certain practices and assumptions which are common
and misleading.
T h e office of the theorem is m u c h more negative
than positive; b u t is not the less useful for that. I t
w i l l not perhaps tell us how to do much that we
cannot do w i t h o u t i t already; b u t i t w i l l prevent us
f r o m doing stupid things which we are fond of doing.
So a theory of evolution at least makes i t more diffi
cult to believe that T h e Dog Fritz i n the German
account really d i d the children's sums for them, or
reminded them to salute their 'dear German flag.'
So even an elementary physics puts i n its place
among superstitions M r . Gladstone's firm belief that
snow has "a peculiar power of penetrating leather,"
a power not possessed by water! For lack of that
knowledge of physics i n M r . Gladstone, L o r d
38 T H E PHILOSOPHY OF R H E T O R I C

Rayleigh found i t quite impossible to persuade h i m


i t was not so.
T h e context theorem of meaning w o u l d prevent
our making hundreds of baseless and disabling as
sumptions that we commonly make about meanings,
over-simplifications that create false problems inter
fering w i t h closer comparisons and that is its m a i n
service. I n this, i t belongs w i t h a n u m b e r of other
theorems w h i c h may be called policeman doctrines
because they are designed on the model of an ideal
police-force, not to make any of us do a n y t h i n g b u t
to prevent other people f r o m i n t e r f e r i n g u n d u l y
w i t h our l a w f u l activities. T h e organization of i m
pulses doctrine of values for literary criticism is i n
the same position. These policeman doctrines keep
assumptions that are out of place f r o m frustrating
and misleading sagacity. I shall be i l l u s t r a t i n g the
restraint of these b u l l y i n g assumptions i n most parts
of Rhetoric later. W e had one simple instance w i t h
L o r d Kames' peacock's feather, last time, where what
was discouraged was a nai've view of imagery as the
stuffof meaning.
W e shall have others i n discussing the claims of
usage next week. Preeminently what the theorem
w o u l d discourage, is our habit of behaving as though,
i f a passage means one t h i n g i t cannot at the same
time mean another and an incompatible t h i n g .
Freud taught us that a dream may mean a dozen
different t h i n g s ; he has persuaded us that some
symbols are, as he says, 'over-determined' and mean
many different selections f r o m among their causes.
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 39

T h i s theorem goes further, and regards a l l discourse


outside the technicalities of science as over-de
termined, as having m u l t i p l i c i t y of meaning. I t
can illustrate this view f r o m almost any of the great
controversies. A n d i t offers us by restraining the
One and O n l y One T r u e Meaning Superstition a
better hope, I believe, of profiting f r o m the contro
versies. A controversy is normally an exploitation
of a systematic set of misunderstandings for war-like
purposes. T h i s theorem suggests that the swords of
dispute m i g h t be t u r n e d into plough shares; and a
way f o u n d by w h i c h we may (to revert to Hobbes)
"make use to o u r benefit of effects formerly seen
for the commodity of h u m a n l i f e . "
T h e next problem concerns what happens when
we p u t words together i n sentences. A t least that
is a common way of stating i t . T h e theorem recom
mends us rather to t u r n the p r o b l e m r o u n d and ask
what happens when, out of the integral utterance
w h i c h is the sentence, we t r y to isolate the discrete
meanings of the words of w h i c h i t is composed.
T h a t problem, the analysis of sentences and the
interaction between words i n the sentence, is m y
subject for next week. I t is there that the most
deep-rooted, systematic and persistent misunder
standings arise.
A t h i r d set of problems concerns rivalries between
different types of context w h i c h supply the meaning
for a single utterance. These start w i t h the plain
equivoque as when the w o r d 'reason* may mean
either a cause or an argument. I am simplifying
40 T H E PHILOSOPHY OF R H E T O R I C

this here to make i t a type of a really simple am


biguity. Actually i n most occurrences i t w o u l d be
m u c h more complex and not so easily cleared u p ,
as the shifting meanings of 'cause' and 'argument*
themselves show. T h e context theorem of mean
i n g w i l l make us expect ambiguity to the widest ex
tent and of the subtlest kinds nearly everywhere,
and of course we find i t . B u t where the o l d Rhetoric
treated ambiguity as a fault i n language, and hoped
to confine or eliminate i t , the new Rhetoric sees i t as
an inevitable consequence of the powers of language
and as the indispensable means of most of our most
important utterances especially i n Poetry and Re
l i g i o n . A n d that too I shall be illustrating later.
O f course ambiguities are a nuisance i n exposition
as, i n spite of my efforts, you have certainly been
feeling. B u t neutral exposition is a very special
l i m i t e d use of language, comparatively a late de
velopment to w h i c h we have not (outside some
parts of the sciences) yet adapted i t . T h i s brings
me to those large-scale rivalries between contexts
which shift the very aims of discourse. W h e n the
passions the combative passion and others inter
vene, either i n the formation of an utterance or i n
its interpretation, we have examples of context
action just as m u c h as when the w o r d 'paper/ say,
takes its meaning f r o m its contexts. T h e extra
meaning that comes i n when a sentence, i n addi
t i o n to m a k i n g a statement, is meant to be insulting,
or flattering, or is interpreted so we may call i t
emotive meaningis not so different f r o m plain
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 41

statement as we are apt to suppose. As the w o r d


means the missing part of its contexts and is a sub
stitute for them, so the insulting i n t e n t i o n may be
the substitute for a k i c k , - t h e missing part of its
context. T h e same general theorem covers a l l the
modes of meaning.
I began tonight by speaking of the poaching of the
other language functions o n the preserve of pure
exposition. Pure exposition has its guardian pas
sions no doubt though I do not know their names.
B u t they are not often as strong as the poachers
and are easily beguiled by them. I t has been so
necessary to us, especially since the physical basis of
civilization became technical, to care at least some
times for the t r u t h o n l y a n d keep the poachers some
times out, that we have exaggerated enormously the
extent of pure exposition. I t is a relatively rare
occurrence outside the routine of train services
and the tamer, more settled parts of the sciences.
W e have exaggerated o u r success for strategic rea
sonssome of them good, because encouraging, i f
we do not too m u c h hoodwink ourselves. I have
aimed at points tonight to be merely expository i n
m y remarks, b u t I know better than t o suppose I
have succeeded. W e shall find, preeminently i n the
subject of rhetoric, that interpretations and opinions
about interpretations that are not p r i m a r i l y steps of
partisan policy are excessively hard to arrive at.
A n d thereby we re-discover that the w o r l d so far
f r o m being a solid matter of fact is rather a fabric
of conventions, w h i c h for obscure reasons i t has
48 T H E PHILOSOPHY OF R H E T O R I C

suited us i n tbe past to manufacture and support.


A n d that sometimes is a dismaying re-discovery
which seems to unsettle o u r foundations.
Anyone who publishes a book w i t h the w o r d
'Meaning' i n its title becomes the recipient of a fan-
m a i l of peculiar character. I n comes a d r i b b l e of
letters ever after f r o m people who are quite un
mistakably lunatics. Indeed, i t seems that the sub
ject is a dangerous one. Intense preoccupation w i t h
the sources of our meanings is disturbing, increasing
our sense that our beliefs are a veil and an artificial
veil between ourselves and something that other
wise than t h r o u g h a veil we cannot know. Some
t h i n g of the same sort can happen i n travel. Anyone
who has visited a sufficiently strange country and
come i n t o close contact w i t h its life knows how un
settling and disorientating is the recognition of the
place of conventions i n our mental w o r l d . A n d the
effect is deeper as the contact is closer. Few men
have come i n t o closer and more active contact w i t h
an alien w o r l d than Colonel Lawrence and when, at
the end of the I n t r o d u c t i o n to The Seven Pillars of
Wisdom, he writes of the selves which converse i n the
void, he says, " T h e n madness was very near, as I be
lieve i t w o u l d be near the man who could see things
through the veils at once of two customs, two educa
tions, two environments." H e is w r i t i n g of fatigue,
and the page reeks of the extremities of war and of
the desert the desert which pushes man down to
the limits of his endurance. T h e meditation of a
single code of meanings is not so devastating, and I
AIMS OF DISCOURSE AND TYPES OF C O N T E X T 43

have seen already enough of B r y n M a w r to realize


that i t bears no least resemblance to a desert. We
may then continue undeterred by the implications
of my fan-mail.
T h e subject of the next lecture w i l l be the Doc
trine of Usage and the Interinanimation of Words
and, as the rest of the course w i l l be literary rather
than philosophical and w i l l attempt rather to prac
tise than to theorize, I may close here w i t h some
lines f r o m George Chapman about the theoretic
principles of Rhetoric, the conduct of interpreta
t i o n and " i m p a r t i a l contention" and their proper
relation to action. I t comes i n a poem entitled

To Young Imaginaries in Knowledge.


T h i s rather were the way, if thou wouldst be
A true proficient i n philosophy
Dissemble what thou studiest u n t i l
B y thy i m p a r t i a l contention
T h o u provest thee fit to do as to profess
A n d if thou still profess it not, what less
Is thy philosophy if i n thy deeds
R a t h e r than signs a n d shadows, it proceeds.

I must apologize i f i n this Lecture I have departed


f r o m the spirit o f h i s recommendation.
LECTURE I I I

THE INTERINANIMATION OF WORDS


Since children learn the use of words most evidently
without having any data or fixed points to go upon,
philosophers and candid persons may learn at last to
understand one another with facility and certainty.
David Hartley, On Man.
*
LECTURE I I I

THE I N T E R I N A N I M A T I O N OF WORDS

I T U R N now to that other sense of 'context'the


literary context which I distinguished last time
f r o m the technical sense of 'context/ as a recurrent
group of events, that is convenient for the theorem
of meaning. Let us consider some of the effects on
words of their combination i n sentences, and how
their meaning depends u p o n the other words before
and after them i n the sentence. W h a t happens
when we t r y w i t h a sentence to decide what single
words i n i t mean ?
T h e sentence, of course, as Aristotle taught, is
the u n i t of discourse. W e can hardly give too m u c h
importance here to the influence of our modern way
of separating words i n w r i t i n g . I n conversation we
do not ordinarily separate them sounless we are
asking questions about words. W i t h languages
w h i c h have not been used i n w r i t i n g and thus sub
jected to a special k i n d of grammatical analysis i t
is w o r t h recalling that grammar takes its name f r o m
w r i t i n g t h e r e is often very great uncertainty as to
where one w o r d ends and another begins. The
w r i t t e n f o r m gives words far more independence
than they possess as units of sound i n speech and we
derive thence a habit of supposing that they have far
more independence as regards their meanings than
47
48 T H E PHILOSOPHY OF R H E T O R I C

they usually have i n either w r i t t e n or spoken dis


course.
T h e m u t u a l dependence of words varies evidently
w i t h the type of discourse. A t one end of the scale,
i n the strict exposition of some highly criticized
and settled science t h r o u g h technicalized and r i g i d
speech, a large p r o p o r t i o n of them are independ
ent. T h e y mean the same whatever other words
they are p u t w i t h ; or i f a w o r d fluctuates, i t moves
only into a small number of stable positions, which
can be recorded and are anchored to defini
tions. T h a t is the ideal l i m i t towards which we
aim i n exposition. Unfortunately we tend in
creasingly since the 17th Century to take r i g i d
discourse as the n o r m , and impose its standards u p o n
the rest of speech. T h i s is m u c h as i f we thought
that water, for a l l its virtues, i n canals, baths and
turbines, were really a weak f o r m of ice. The
other end of the scale is i n poetry i n some forms
of poetry rather. W e k n o w very much less about
the behavior of words i n these cases when their
virtue is to have no fixed and settled meaning sepa
rable f r o m those of the other words they occur w i t h .
T h e r e are many more possibilities here than the
theory of language has yet tried to t h i n k out. Often
the whole utterance i n which the co-operating mean
ings of the component words hang on one another
is not itself stable i n meaning. I t utters not one
meaning b u t a movement among meanings. Of
course, even i n the strictest prose we always have
one t h i n g that may be described as a movement of
T H E INTERINANIMATION OF WORDS 49

meaning. W e have change as the sentence develops.


I n " T h e cat is on the m a t " we begin w i t h the cat
and end w i t h the mat. T h e r e is a progression of
some sort i n every explicit sentence. B u t i n the
strictest prose the meanings of the separate words
theoretically stay p u t and thought passes f r o m one
to another of them. A t the other end of the scale
the whole meaning of the sentence shifts, and w i t h
i t any meanings we may try to ascribe to the i n d i
v i d u a l words. I n the extreme case i t w i l l go on
m o v i n g as long as we b r i n g fresh wits to study i t .
W h e n Octavius Caesar is gazing down at Cleopatra
dead, he says,

She looks like sleep,


A s she w o u l d catch another Antony
I n h e r strong toil of grace.

" H e r strong t o i l of grace." Where, i n terms of what


entries i n what possible dictionary, do the meanings
here of toil and grace come to rest ?
B u t m y subject is Rhetoric rather than Poetics
and I want to keep to prosewhich is not too far f r o m
the strict scientific or ' r i g i d ' end of this scale of
dependent variabilities. I n the k i n d of prose I am
talking now, you have usually to wait t i l l I have
gone o n a b i t before you can decide how you w i l l
understand the opening parts of the sentences. If,^
instead, I were reading you the first few theorems
of E u c l i d , that w o u l d not be so. Y o u w o u l d under
stand, as soon as I said 'a triangle,' what the w o r d
meant, and though what I went on to say m i g h t
5 o T H E PHILOSOPHY OF R H E T O R I C

qualify the meaning ('having two sides equal'), i t


would not destroy or completely change the mean
i n g that you had so far given to the w o r d . B u t i n
most prose, and more than we ordinarily suppose,
the opening words have to wait for those that fol
low to settle what they shall mean i f indeed that
ever gets settled.
A l l this holds good not only as to the sense of the
w a i t i n g words b u t as regards a l l the other functions
of language w h i c h we can distinguish and set over
against the mere sense. I t holds for the feeling i f any
towards what I am t a l k i n g about, for the relation
towards my audience I want to establish or main
tain w i t h the remark, and for the confidence I have
i n the soundness of the remark t o m e n t i o n three
m a i n sorts of these other language functions. I n
speech, of course, I have the aid of i n t o n a t i o n for
these purposes. B u t , as w i t h the meanings of words,
so w i t h the i n t o n a t i o n structure. T h e i n t o n a t i o n
of the opening words is l i k e l y to be ambiguous;
i t waits t i l l the utterance is completed for its f u l l
interpretation.
I n w r i t i n g we have to replace i n t o n a t i o n as far
as we can. Most of the more recondite virtues of
prose style come f r o m the skill w i t h w h i c h the r i v a l
claims of these various language functions are recon
ciled and combined. A n d many of the rather mys
terious terms that are usually employed i n discussing
these matters, harmony, rhythm, grace, texture,
smoothness, suppleness, impressiveness, and so o n
are best taken u p for analysis f r o m this p o i n t of
T H E INTERINANIMATION OF WORDS 51

view. O r rather the passages w h i c h seem to ex


emplify these qualities (or fail to) are best exam
ined w i t h the m u l t i p l i c i t y of the language functions
i n m i n d . For we can obviously do n o t h i n g w i t h such
words as these by themselves, i n the blue. T h e y may
mean a l l sorts of different things i n different literary
contexts.
I have been leading u p or down, i f you like
to an extremely simple and obvious b u t fundamen
tal r e m a r k : that no w o r d can be judged as to
whether i t is good or bad, correct or incorrect,
beautiful or ugly, or anything else that matters to
a writer, i n isolation. T h a t seems so evident that I
am almost ashamed to say i t , and yet i t flies straight
i n the face of the only doctrine that for two h u n d r e d
years has been officially inculcated when any doc
trine is inculcated i n these matters. I mean the
doctrine of Usage. T h e doctrine that there is a
r i g h t or a good use for every w o r d and that literary
virtue consists i n making that good use of i t .
There are several bones that can be picked w i t h
that doctrine as i t has been expounded i n many
epochs and, i n particular for us, f r o m the middle
of the 18th Century onwards. I t is the worst legacy
we have f r o m that, i n other ways, happy Century.
A t its best i t can be found i n George Campbell's
Philosophy of Rhetoric otherwise an excellent
book i n many respects. A t its worst, or nearly its
worst, the doctrine can be f o u n d i n most of the
Manuals of Rhetoric and Composition which have
afflicted the schools American schools especially.
5* T H E PHILOSOPHY OF R H E T O R I C

I t asserts that "Good use is the general, present<tay


practice of the best writers." One bone we could
pick w o u l d be w i t h that 'best.' H o w are they the
best writers except by using the words i n the best
ways? W e settle that they are the best writers be
cause we find them using their words successfully.
W e do not settle that theirs is the right, the 'good
usage' of the words because they use them so.
Never was there a crazier case of p u t t i n g the cart
before the horse. I t is as though we were to main
tain that apples are healthy because* wise people
eat them, instead of recognizing that i t is the other
way about that i t is what the food w i l l do for us
w h i c h makes us eat i t , not the fact that we eat i t
w h i c h makes i t good food.
B u t that is not the m a i n bone I have to pick w i t h
the doctrine, w h i c h is that i t blanks o u t and hides
the interinanimation between words. I had better
cite you a sentence or two i n evidence, or you may
t h i n k I am i n v e n t i n g a ghost to exorcize. I w i l l
take them f r o m a Manual of Rhetoric w h i c h carries
the names of three authors : Messrs. Gardiner, K i t -
tredge and A r n o l d . A n d I choose this book because
the regard which I have for M r . Kittredge's name
makes a doctrine w h i c h has that sanction seem the
better w o r t h refuting. T h e authors write : "Usage
governs language. T h e r e is no other standard.
By usage, however, is meant the practice of the best
writers and speakers." ( I have already asked what

* 'Because' is offering to play one of its most troublesome tricks


here, of course, in the shift from 'cause' to *reason.'
T H E INTERINANIMATION OF WORDS 53

standard is supposed to settle w h i c h are the best.)


T h e y go on to consider " f o u r great general p r i n
ciples of choice : correctness, precision, appropriate-
ness and expressiveness," which, they say, " w i t h i n
the limits of good usage and i n every case controlled
by i t . . . should guide us i n the choice of words."
A n d this is what they say of correctness: "Correct
ness is the most elementary of all requirements.
T h e meanings of words are settled by usage. I f we
use a word incorrectly that is i n a sense which
does not customarily belong to i t our readers w i l l
miss our thought, or, at best, they must arrive at i t
by inference and guesswork."
Inference and guesswork! W h a t else is inter
pretation? H o w , apart f r o m inference and skilled
guesswork, can we be supposed ever to understand
a w r i t e r or speaker's thought? T h i s is, I think, a
fine case of p o k i n g the fire f r o m the top. B u t I
have still my m a i n b i t of evidence to give you. M y
authors say: " I n studying the four great principles
of choice, we observe that only the first (correct
ness) involves the question of r i g h t and wrong.
T h e others deal w i t h questions of discrimination
between better and worse that is w i t h the closer
adaptation of words to the thoughts and feelings
which we undertake to express. Further, i t is only
i n dealing w i t h the first principle (correctness) that
we can keep o u r attention entirely on the single
word."
T h e r e ! that is the view I wished to illustrate. Let
us not boggle about the oddities of its expression:
54 T H E PHILOSOPHY OF R H E T O R I C

'right and w r o n g / 'better and worse'; or worry as


to how by keeping " o u r attention entirely on a single
w o r d " we could settle anything at all about i t
except perhaps about its spelling! T h e i m p o r t a n t
point is that words are here supposed just sheerly
to possess their sense, as men have their names i n the
reverse case, and to carry this meaning w i t h them
into sentences regardless of the neighbour words.
T h a t is the assumption I am attacking, because, i f
we follow u p its practical consequences i n w r i t i n g
and reading and trace its effects u p o n interpreta
tion, we shall find among them no small p r o p o r t i o n
of the total of o u r verbal misunderstandings.
I am anxious not to seem to be illustrating this
sort of misunderstanding myself here, u n w i t t i n g l y ,
i n my interpretation of this passage. I know w e l l
enough that the authors probably had i n m i n d such
incorrectness as occurs when people say 'ingenious'
when they mean 'ingenuous'; and I know that the
Usage Doctrine can be interpreted i n several ways
which make i t true and innocuous.
I t can say and t r u l y , for example, that we learn
how to use words f r o m responding to them and
n o t i n g how other people use them. Just how we do
so learn is a deep b u t explorable question. I t can
say equally t r u l y , that a general conformity between
users is a condition of communication. That no
one w o u l d dream of disputing. B u t i f we consider
conformity we see that there are two kinds of con
formity. Conformity i n the general process of i n
terpretation, and conformity i n the specific products.
T H E INTERINANIMATION OF WORDS 55

W e a l l k n o w how the duller critics of the 18th Cen


t u r y (the century that gave us the current Doctrine
of Usage) the people Wordsworth was t h i n k i n g of
when he wrote his Preface, confused the poetic
product w i t h the poetic process and thought a poem
good because i t used poetic diction the words that
former good poets had used and used them i n the
same ways. T h e Usage Doctrine, i n the noxious i n
terpretation of i t , is just that blunder i n a more per
vasive and more dangerous incidence. T h e nox
ious interpretation is the common one. Its evil is
that i t takes the senses of an author's words to be
things we know before we read h i m , fixed factors
w i t h which he has to b u i l d u p the meaning of his
sentences as a mosaic is p u t together of discrete i n
dependent tesserae. Instead, they are resultants
which we arrive at only through the interplay of
the interpretative possibilities of the whole utter
ance. I n brief, we have to guess them and we guess
much better when we realize we are guessing, and
watch out for indications, than when we t h i n k we
know.*
T h e r e are as many morals for the w r i t e r as for
the reader i n a l l this, b u t I w i l l keep to interpreta
tion. A w o r d or phrase when isolated momentarily
f r o m its controlling neighbours is free to develop
irrelevant senses which may then beguile half the
other words to follow i t . A n d this is at least equally
true w i t h the language functions other than sense,
w i t h feeling, say. I w i l l give you one example of
* See the Note at the end of this Lecture.
56 T H E PHILQ$OPHY OF RHETORIC

an erratic interpretation of feeling, and i f I take i t


from the same Manual of Rhetoric that is because
i t illustrates one of the things to which the mosaic
view or habit of interpretation, as opposed to the
organic, often leads.
T h e Authors give the following f r o m Bacon's
Advancement of Learning. A n d i n re-reading i t I
w i l l ask you to note how cunningly Bacon, i n describ
i n g some misuses of learning, takes back w i t h one
hand what he seems to be offering w i t h the other,
indicating both why men do prefer misuses and why
they should not do so.
B u t the greatest error of a l l the rest is the mistaking
or misplacing of the last or furthest end of knowledge.
F o r men have entered into a desire of learning a n d
knowledge, sometimes upon a n a t u r a l curiosity a n d i n -
quisitive appetite; sometimes to entertain their minds
w i t h variety a n d delight; sometimes for ornament a n d
r e p u t a t i o n ; a n d sometimes to enable them to victory
of wit and contradiction; a n d most times for lucre a n d
profession; and seldom sincerely to give a true account
of their gift of reason, to the benefit a n d use of m e n :
as if there were sought i n knowledge a couch, where-
upon to rest a searching or restless s p i r i t ; or a terrace,
for a wandering a n d variable m i n d to walk u p a n d
down with a fair prospect; or a tower of state, for a
proud m i n d to raise itself u p o n ; or a fort or com-
m a n d i n g ground, for strife a n d contention; or a shop,
for profit or s a l e ; a n d not a r i c h storehouse, for the
glory of the Creator a n d the relief of man's estate.

There is much to take to heart here especially as


to the couch aspect of the Usage Doctrine, and, I
must admit, the tower and the fort b u t what the
authors say about i t is t h i s :
T H E INTERINANIMATION OF WORDS 57
H e r e the splendor of the imagery is n o mere em-
bellishment. W i t h o u t it, B a c o n could not have given
adequate expression to h i s enthusiastic appreciation of
learning a n d his fine scorn for the unworthy uses to
w h i c h it is sometimes put. A t the same time, the fig-
ures elevate the passage from the ordinary levels of
prose to a noble eloquence, (p. 372)

W h a t splendor is there i n the imagery? These


images have no splendor as Bacon uses them, b u t are
severely efficient, a compact means for saying what
he has to say. H i s 'enthusiastic appreciation* (a
poor phrase, I suggest, to smudge over h i m ! ) of the
use of knowledge and his 'fine scorn' of unworthy
uses are given only i f we refuse to be beguiled by
the possibilities of splendor i n the isolated images.
Loose them even a l i t t l e f r o m their service, let their
'splendor' act independently, and they begin at once
to fight against his intention. For the terrace, the
tower and the fort, i f they were allowed to 'elevate,'
w o u l d make the misplacings of the last and furthest
end of knowledge seem much grander than "a true
account of their gift of reason, to the benefit and use
of m e n " - a s a terrace or tower of state or a fort w i l l
seem grander than a mere r i c h storehouse.

Let me go on to some further types of the m u t u a l


control and interinanimation between words. So
far I have considered only the influence of words ac
tually present i n the passage, b u t we have to include
words w h i c h are not actually being uttered and are
only i n the background. Take the case of what are
variously called expressive, symbolic, or simulative
58 T H E PHILQSOPHY OF R H E T O R I C

words words which 'somehow illustrate the mean


i n g more immediately than do ordinary speech
forms,' to quote Leonard Bloomfield. Examples are
flip, flap, flop, flitter, flimmer, flicker, flutter, flash,
flush, flare, glare, glitter, glow, gloat, glimmer, bang,
bump, lump, thump, thwack, whack,sniff, sniffle,
snuff. . . W h y should these seem so peculiarly ap
propriate, or fitting, to the meanings we use them
for? T h e popular view is that these words just
simply imitate, are copies of, what they mean. But
that is a short-cut theory which often does not work,
and we can, I t h i n k , go further and do better. As
Bloomfield, i n his excellent book, Language, says,
"the explanation is a matter of grammatical struc
ture, to the speaker i t seems as i f the sounds were
especially suited to the meaning." T h e speaker
usually thinks moreover that the w o r d seems suited
because i n some way i t resembles the meaning, or,
if this seems unplausible, that there must be some
direct connection between them. I f i t is not the
sound of the w o r d which resembles the meaning
then perhaps the tongue and l i p movements instead
imitate something to do w i t h the meaning and so
on. Sir Richard Paget's theories of imitative ges
tures are likely to be appealed to nowadays.
T h e most that the modern linguist who com
pares the very different words w h i c h are used i n
different languages for their meanings is prepared
to allow towards this resemblance of sound and sense
is that "we can distinguish, w i t h various degrees of
clearness and w i t h d o u b t f u l cases on the border line,
T H E INTERINANIMATION OF WORDS 59

a system of i n i t i a l and final root-forming morphemes


of vague signification." Note how guarded Bloom
field is over such a point. #

I must explain what a morpheme is. T w o o r


more words are said to share a morpheme when they
have, at the same time, something i n common i n
t h e i r meaning and something i n common i n their
sound. T h e j o i n t semantic-phonetic u n i t w h i c h
distinguishes them is what is called a morpheme. I t
is the togetherness of a peculiar sound and a peculiar
meaning for a n u m b e r of words.
T h u s flash, flare, flame, flicker, flimmerhave i n
common the sound (fl-) and a suggestion of a 'mov
i n g l i g h t ' a n d this j o i n t possessionis the mor
pheme. Similarly blare, flare, glare, stare have the
sound (-ea) i n common and also the meaning 'big
l i g h t o r noise' shall we say, and this couple sound
and meaning is the morpheme. So w i t h 'smoothly
wet' and (sl-) i n slime, slip, slush, slobber, slide,
slither. B u t pare, pear, pair, though they have a
sound i n common, have no meaning i n common, so
have no common morpheme.
O f course, the existence of a group of words w i t h
a common morpheme has an influence on the forma
t i o n of other words, and on the pronunciation of
other words a s s i m i l a t i n g t h e m to the group.
T h u s , given skid and skate, that is a strong addi
t i o n a l reason, against an English convention, for say
i n g skee rather than shee.
T h i s pedantic l o o k i n g term, morpheme, is useful
because w i t h its help we manage to avoid saying that
6o T H E PHIkpSOPHY OF R H E T O R I C

the sound (sl-) somehow itself means something


like 'smoothly wet or slippery* and gain a way of
saying no more than that a group of words w h i c h
share that sound also share a peculiar meaning.
A n d that is a l l we are entitled to say. T o go further
andsay that the words share the meaning because
they contain this sound and because this sound has
that meaning is to b r i n g i n more than we know
an explanation or theory to account for what we do
know. A n d actually i t is a bad explanation. For
this sound, by itself, means nothing. I t is not the
shared sound b u t each of the words which has the
meaning. T h e sound by itself either means noth
i n g at all as w i t h (fl) i n flame, flare, flash, flicker
or as w i t h (-ea) i n blare, flare, glare, stare i t has by
itself only an irrelevant meaning, namely, that of
air, 'what we breathe.'
T h e theoretical position here is w o r t h close study
because i t is typical of a very large group of posi
tions i n w h i c h we tend, too boldly and too inno
cently, to go beyond our evidence and to assume, as
the obvious explanation, as almost a datum, what
is really the conclusion of a vague and quick and
unchecked inductive argument, often a bad and un
warrantable argument. W h y should a group of
words w i t h a sound i n common have similar mean
ings unless there was a correspondence of some k i n d
between the sound and the meaning? T h a t seems
plausible. B u t state the argument more explicitly,
look over the evidence carefully, and i t becomes un
plausible, for then we have to notice the other words
T H E INTERINANIMATION OF WORDS 61

which share the sound b u t do not share the mean


i n g and the other words which share the meaning
w i t h o u t the sound. T h e n we see that we have been
applying to words the sort of argument which w o u l d
represent a fashion as a spontaneous expression of
original taste on the part of a l l who follow i t . W e
find i n fact that we have been looking at the prob
l e m upside down. T h a t so far f r o m a perceived
correspondence between sound and meaning being
the explanation of the sharing, the existence of a
group of words w i t h a common sound and meaning
is the explanation of our belief i n a correspondence.
T h i s situation, I said a moment ago, is typical.
We can hardly, I think, exaggerate i n an estimate of
the number of literary and rhetorical problems
which, as usually formulated, are upside down i n
this fashion. For example, our common assump
t i o n that when a w o r d such as beautiful or art or
religion or good, is used i n a great variety of ways,
there w i l l be f o u n d something i n common to a l l the
uses, something which is the fundamental or essen
tial meaning of the w o r d and the explanation of its
use. So we spend our wits t r y i n g to discover this
common essential meaning, w i t h o u t considering that
we are looking for i t , most often, only as a result
of a weak and hasty inductive argument. T h i s as
sumption that the same w o r d ought to have or must
have the same meaning, i n an important respect, is
one of those b u l l y i n g assumptions that the context
theorem of meanings w o u l d defend us from i n the
way I discussed i n my lecture last week,
6* T H E PHILOSOPHY OF R H E T O R I C

But to come back to this parallel assumption that


some words, apart f r o m other words, and i n their
own r i g h t i n v i r t u e of their sound must mean cer
tain things. I t was Aristotle who said that there
can be no natural connection between the sound of
any language and the things signified, and, i f we set
the problem r i g h t side u p and remember the other
words before examining i t , we shall have to agree
w i t h h i m . Indeed, i f we ask the question fairly i t
becoihes when we get i t clearnearly senseless.
W h a t resemblance or natural connection can there
be between the semantic and phonetic elements i n
the morpheme ? One is a sound, the other a refer
ence. 'Is (fl-) really like 'moving l i g h t ' i n any way
i n which (sl-) or (gl-) is not ?' Is that not like ask
i n g whether the taste of turkey is like growing i n
some way that the taste of m i n t is not ?
I conclude then that these expressive or symbolic
words get their feeling of being peculiarly fitting
f r o m the other words sharing the morpheme w h i c h
support them i n the background of the m i n d . I f
that is so, a l l sorts of consequences are at once evi
dent. I n translation, for example, the expressive
w o r d i n another language w i l l not necessarily sound
at a l l like the original w o r d . I t w i l l be a w o r d that
is backed u p by other words i n a somewhat analo
gous fashion. Evidently again, a proper apprecia
t i o n of the expressiveness of a w o r d i n a foreign
language w i l l be no matter of merely k n o w i n g its
meaning and relishing its sound. I t is a matter of
having, i n the background of the m i n d , the other
T H E INTERINANIMATION OF WORDS 63

words i n the language which share morphemes w i t h


i t . T h u s no one can appreciate these expressive
features of foreign words justly w i t h o u t a really wide
familiarity w i t h the language. W i t h o u t that our
estimates are merely whimsical.
W e can, and I t h i n k should, extend this n o t i o n
of a w o r d as being backed u p by other words that
are not uttered or thought of. A first extension is
to words that sound alike b u t do not share a mor
pheme, do not have a common meaning b u t only
some relevant meaning. T h u s blare, scare and dare
do not share a morpheme, b u t on occasion the pe
culiar force of blare may w e l l come to i t i n part f r o m
the others. T h i s , of course, is only recognizing on
a larger, wider scale the principle that Lewis Car
r o l l was using i n Jabberwocky. Its relevance to the
theory of rhymes and assonances is obvious.
A n o t h e r and a wider extension w o u l d include not
only influences f r o m words which i n part sound
alike, b u t f r o m other words which i n part overlap
i n meaning. Words, for example, which we m i g h t
have used instead, and, together w i t h these, the
reasons why we d i d not use them. Another such
extension looks to the other uses, i n other contexts,
of what we, too simply, call 'the same w o r d . ' T h e
meaning of a w o r d o n some occasions is quite as
m u c h i n what i t keeps out, or at a distance, as i n
what i t brings i n . A n d , on other occasions, the
meaning comes f r o m other partly parallel uses whose
relevance we can feel, w i t h o u t necessarily being able
to state i t explicitly. B u t w i t h these last leaps I may
64 T H E PHILOSOPHY OF R H E T O R I C

seem i n danger of m a k i n g the force of a w o r d , the


feeling that no other w o r d could possibly do so
w e l l or take its place, a matter whose explanation
w i l l drag i n the whole of the rest of the language.
I am not sure, though, that we need be shy of some
t h i n g very like this as a conclusion. A really mas
terly use of a language i n free or fluid, not techni
cal discourse Shakespeare's use of English for
example, goes a long way towards using the lan
guage as a whole.
Cleopatra, taking u p the asp, says to i t :
Come, thou mortal wretch,
With thy sharp teeth this knot intrinsicate
Of life at once untie ; poor venomous fool,
Be angry, and despatch!
Consider how many senses of mortal, besides 'death-
dealing' come i n ; compare: ' I have i m m o r t a l long
ings i n me/ Consider knot: ' T h i s k n o t intrinsicate
of life': 'Something to be undone,' 'Something that
troubles us u n t i l i t is undone,' 'Something by w h i c h
all holding-together hangs,' ' T h e nexus of all mean
ing.' Whether the homophone not enters i n here
may be thought a d o u b t f u l matter. I feel i t does.
But consider intrinsicate along w i t h knot. Edward
Dowden, f o l l o w i n g the fashion of his time i n m a k i n g
Shakespeare as simple as possible, gives 'intricate'
as the meaning here of intrinsicate. A n d the Ox
f o r d Dictionary, sad to say, does likewise. B u t
Shakespeare is b r i n g i n g together half a dozen mean
ings f r o m intrinsic and intrinse: 'Familiar,' ' i n t i
mate,' 'secret,' 'private,' 'innermost,' 'essential,' 'that
T H E INTERINANIMATION OF WORDS 65

which constitutes the very nature and being of a


t h i n g a l l the medical and philosophic meanings of
his time as w e l l as 'intricate' and 'involved.' W h a t
the w o r d does is exhausted by no one of these mean
ings and its force comes f r o m all of them and more.
As the movement of my hand uses nearly the whole
skeletal system of the muscles and is supported by
them, so a phrase may take its powers f r o m an i m
mense system of supporting uses of other words i n
other contexts.

NOTE

The word usage itself well illustrates some of the more


troublesome shifts of meaning. An improved Rhetoric has
among its aims an improved control over these. Here per-
haps a list of some of the senses of usage may help us in
avoiding misunderstanding.
(1) The most inclusive sense is "the entire range of the
powers which the word can exert as an instrument
of communication in all situations and in co-opera-
tion with any other words."
(In this sense 'Usage, and usage alone, undoubtedly
controls language.')
(2) "Some specific power which, in a limited range of
situations and with a limited type of verbal context
the word normally exerts."
(This is often called a use or sense and is what the
Dictionary attempts to record in its definitions, by
giving other words, phrases and sentences with the
same specific power.)
(3) An instance of 2, at a certain place in Shakespeare,
say, which may be appealed to to show that the word
can have that power.
(4) A supposed fixed 'proper' meaning that the word
must be kept to (has in its own right, etc.) This
T H E PHILOj>OPHY OF R H E T O R I C
notion is derived from i, 2 and 3 by over-simplifica-
tion and a misconception of the working of language
which, typically, takes the meaning of a sentence to
be something built up from separate meanings of its
words instead of recognizing that it is the other
way about and that the meanings of words are de-
rived from the meanings of sentences in which they
occur. This misconception assimilates the process
by which words have their meanings determined
with that by which they have their spelling deter-
minedand is the origin of a large part of misinter-
pretation.
LECTURE IV

SOME C R I T E R I A OF WORDS
There is no warrant for the placing on these inevitably
rather light heads and hearts, on any company of you,
assaulted, in our vast vague order, by many pressing
wonderments, the whole of the burden of a care for
tone.Henry James, The Question of our Speech.
LECTURE IV

SOME CRITERIA OF WORDS

A S T week I was concerned w i t h the interde-


L A pendences of words i n discourse, and the
interinanimation between them. I began by ar
raigning the conventional Doctrine of Usage. I
accused i t of forgetting that a w o r d is always a co
operative member of an organism, the utterance,
and therefore cannot properly i n ordinary free,
fluid, non-technical discourse be thought to have
a meaning of its own, a fixed correctusage, or even
a small l i m i t e d number of correct usages unless by
'usage' we mean the whole how of its successful co
operations w i t h other words, the entire range of the
varied powers which, w i t h their aid, i t can exert.
T h e traditional Usage Doctrine, I said, treated lan
guage on the bad analogy of a mosaic, and conceived
composition and interpretation as though they were
a p u t t i n g together or taking apart of pieces w i t h a
fixed shape and color, whereas, i n fact, the interin
animation of the meanings of words is at least as
great as i n any other mode of mental performance.
A note i n a musical phrase takes its character from,
and makes its c o n t r i b u t i o n only w i t h , the other
notes about i t ; a seen color is only what i t is w i t h
respect to the other colors co-present w i t h i t i n the
visual field; the seen size or distance of an object
69
7o T H E PHILOSOPHY OF R H E T O R I C

is interpreted only w i t h regard to the other things


seen w i t h i t . Everywhere i n perception we see this
interinanimation (or interpenetration as Bergson
used to call i t ) . S o w i t h w o r d s , too, b u t m u c h
m o r e ; the meaning we find for a w o r d comes to i t
only w i t h respect to the meanings of the other words
we take w i t h i t . A n d towards the end of the lec
ture I extended this view to include n o t only the
other words uttered w i t h i t , b u t also unuttered
words i n various relations to i t w h i c h may be back
i n g i t u p though we never t h i n k of them. So, i n
perceiving the size or shape or distance of a t h i n g ,
all sorts of actions we m i g h t take i n w a l k i n g towards
i t , or grasping i t , come i n though we may never
t h i n k of them to guide o u r interpretation. Again,
the etymological h i n t of inter sums u p the whole
story.
I want now to touch u p o n t w o or three considera
tions w h i c h support and illustrate this view before
proceeding to discuss s o m e o f the criteria, or head
ings, under w h i c h we commonly profess to judge
the merits or demerits of words. Y o u w i l l see, I
hope, that these criteria^, vividness, ex
pressiveness, clarity, beauty, are representative i n
stances of t h e m a r e misleading and unprofitable to
study unless we use t h e m w i t h a due recognition of
this interdependence among the words we use them
to describe and an alert distrust of o u r h a b i t of tak
i n g words and their meanings for examination i n
isolation. T h e isolation is never complete, of
course ; a completely isolated w o r d w o u l d be mean-
SOME C R I T E R I A OF WORDS 71

ingless. T h e detachment we attempt is by means


of a supposed standard setting, an imaginary, sche
matic context w h i c h is assumed to be representative.
A n d i t is the habit of trusting such supplied b u t u n
examined contexts ('as generally used,' ' i n ordinary
discourse,' ' i n common parlance' and so on) that I
am attacking. T h e strength of this habit is too
great for anyone once for all, or for l o n g at a time,
to r i d himsetf of i t . T h e view that meanings be
l o n g to words i n their own r i g h t and the more so
phisticated views w h i c h have the same effectare a
branch of sorcery, a relic of the magical theory of
names. A n d m y experience is that the most deter
m i n e d efforts do no more than free us f r o m i t n o w
and then for a few precious moments. T h u s i n ex
h o r t i n g you to discard i t , I feel i n the position of
that Basuto Chief, reported by Casalis i n 1861, who
called his tribe together to warn them against an
other mode of sorcery. "Sorcery," said the Basuto
C h i e f w h o was evidently as much persuading h i m -
setf as t h e m - " S o r c e r y only exists i n the mouths of
those who speak i t . I t is no more i n the power of
a mart to k i l l his fellow by mere effort of his w i l l ,
than i t w o u l d be to raise h i m f r o m the dead. T h a t
is m y o p i n i o n . Nevertheless, y o u sorcerers who
hear me speak, use m o d e r a t i o n ! "
So here, t h o u g h I may be intellectually persuaded,
and persuade you, that a w o r d by itself w i t h o u t a
setting of other words, uttered or supplied, can no
more have a meaning than a patch of color can
have a size or distance w i t h o u t its setting, yet I do
72 T H E PHILOSOPHY OF R H E T O R I C

not expect that our behavior w i l l be m u c h changed.


T h e habitual p u l l of the contrary received assump
t i o n is too strong. O u r best hope is that we may
learn to use moderation i n o u r reliance on that as
sumption.
T h e evil influence of the assumption is most glar
ingly shown w i t h the abstract words u p o n w h i c h a l l
discussions of general theoretical topics t u r n . Out
side of the sciences i n our talk about politics, so
ciety or conduct, or about science itself, i n all
branches of philosophy i n c l u d i n g psychology, i n a l l
discussions of art, literature, language, t r u t h , beauty
and the good, o u r p r i n c i p a l terms incessantly change
their meanings w i t h the sentences they go into and
the contexts they derive from. W e are a l l ready
enough to suspect this, i f not i n our o w n talk at
least i n that of our fellows, and ready to see i n i t a
chief cause for the lamented fact that these subjects
show once we have allowed for current fashions
strangely l i t t l e progress. B u t b o t h the extent and
the plan of these deluding shifts are hidden f r o m us
by the assumption I am attacking. I t leads us to
t h i n k that a shift of meaning is a flaw i n discourse,
a regrettable accident, instead of a virtue. A n d
therefore we neglect to study the plan of the shifts.
T h e assumption is that words have, or should
have, proper meanings w h i c h people should recog
nize, agree about and stick to. A pretty program, i f
i t were possible. But, outside the technical lan
guages of the sciences, i t is not possible. For i n the
topics w i t h w h i c h a l l generally interesting discussion
SOME C R I T E R I A OF WORDS 73

is concerned, words must shift their meanings thus.


W i t h o u t these shifts such m u t u a l understanding as
we achieve w o u l d fail even w i t h i n the narrowed re
sultant scope. Language, losing its subtlety w i t h its
suppleness, w o u l d lose also its power to serve us.
T h e remedy is not to resist these shifts b u t to
learn to follow them. T h e y recur i n the same
forms w i t h different w o r d s ; they have similar plans
and common patterns, w h i c h experience enables us
to observe and obey i n practice sometimes w i t h a
s k i l f u l ease w h i c h seems amazing when we examine
i t . W e may reasonably hope that systematic study
w i l l i n time p e r m i t us to compare, describe and ex
p l a i n these systematic ambiguity or transference
patterns o n a scale as m u c h surpassing o u r best
present-day Dictionary T e c h n i q u e as, say, o u r pres
ent competence i n chemistry surpasses that of Bacon
who foresaw i t . Even now, i f we could take sys-
tematic cognizance of even a small part of the shifts
we fleetingly observe, the effect w o u l d be like that
of i n t r o d u c i n g the m u l t i p l i c a t i o n table to calculators
who just happened to k n o w the w o r k i n g of a few
sums and no others. A n d w i t h such a clarification,
such a translation of o u r skills i n t o comprehension,
a new era of h u m a n understanding and co-operation
i n t h i n k i n g w o u l d be at hand. I t w o u l d not be
difficult to do not a l i t t l e towards this at once.
W h a t stands i n the way is chiefly the Proper Mean
i n g Superstition and the effort i t sustains towards
increased r i g i d i t y i n fields where r i g i d i t y is inap
propriate.
74 T H E PHILOSOPHY OF R H E T O R I C

These shifts deceive us most when they affect ab


stract words, because then they are hardest to follow.
B u t they occur as m u c h and as variously w i t h seem
ingly simple concrete words. W i t h these we often
follow them so easily that we may not suspect that
any shifts are taking place. T h e w o r d book, for
example, troubles no one. A n d yet compare the
use of book, i n w h i c h we distinguish a book f r o m a
magazine or j o u r n a l , w i t h that i n w h i c h a majority
of speakers i n England (call them uneducated i f y o u
like) describe a weekly as a book. O r compare the
senses of book i n " I t ' s a formidable volume, b u t it's
not a book." " H e hashis m i n d f u l l of his book."
" W r i t i n g a book." " B i n d i n g a book." " P r i n t i n g
a book." "Rearranging the books i n the cata
logue." I n each of these we have shifted the sense
of book, sometimes to positions incompatible w i t h
one another. N o one, for example, w i l l ever b i n d
the book which I am now m a k i n g of these lectures.
W h a t w i l l be p r i n t e d and what w i l l be b o u n d are
different things altogether f r o m what I am now
w o r k i n g o n (a set of ideas), though of course they
are connected i n obvious ways w i t h i t .
W e f o l l o w these shifts w i t h o u t trouble because we
are f a m i l i a r w i t h them. W e are n o t yet so familiar
w i t h the shifts of the more heavily worked abstract
words of reflection. I t is the hope and the great
o p p o r t u n i t y for intellectual improvement that we
may i n time become equally f a m i l i a r w i t h them.
T h a t , I w o u l d say, is fundamentally the aim and the
justification of advanced verbal education, a t h i n g
SOME C R I T E R I A OF WORDS 75

otherwise often hard to j u s t i f y ; and the best an


swer to the troublesome question, " W h y should we
w o r r y ourselves w i t h i t ? " is that thereby we may
better find out what we and others are t h i n k i n g .

Towards the end of the last lecture I was suggest


i n g that our words commonly take meaning t h r o u g h
the influence of other words w h i c h we may never
t h i n k of b u t w h i c h i n the back of the m i n d co-oper
ate i n controlling them. A n d I concluded w i t h the
remark that a great w r i t e r often gains his a i m by
making a single phrase p u l l w i t h or against large
ranges of the language. T h i s , i f t r u e , is, of course,
an additional argument against the Proper Usage Su
perstition. W e can say of the single w o r d what
Donne said of single sentences, "Sentences i n Scrip
ture, like hairs i n horsetails, concur i n one root of
beauty and strength; b u t being plucked out one by
one, serve only for springes and snares." W e need
especially to beware of p l u c k i n g words out one by
one when we are tempted to judge new additions
that are being made to the language, for these are
more easily detached and carry less latent, or as
sumed context w i t h them. N o t h i n g better tests
our ideas about the choice of words than the rea
sons we find ourselves giving for l i k i n g or disliking
a new w o r d and n o t h i n g better exposes the doctrine
of usage. O u r language is growing faster and i n
more variedways than at any time since Elizabeth's.
I t is estimated that even for that conservative sec
t i o n of the English speaking peoples w h i c h lives i n
76 T H E PHILOSOPHY OF R H E T O R I C

England, as many a s t h i r t y new words a year have


recently been coming into general use. T h i s quite
apart f r o m the technicalities of the trades and the
sciences. A large number of them crossed the A t
lantic and i t is fair to say, I t h i n k , that among these
w i l l be f o u n d many that have been most heartily
welcomed i n England.
But new words are not always or even usually
welcomed, certainly not by the vocal few who pro
fess to have explicit reasons for their opinions about
them. A new w o r d commonly stirs u p complaints
and some of these complaints are w o r t h study for
the interesting l i g h t they t h r o w u p o n current as
sumptions about language.
Let me take first the complaints most often heard
against the new words that are coined i n the sciences
against those of them that pass f r o m scientific use
i n t o general currency. These complaints are, com
monly, that they are awkward or difficult to pro
nounce or too long, and that they are not labels
but compacted descriptions or explanations. The
prejudice is sometimes so strong that even the lexi
cographers succumb to i t . Y o u w i l l not find, i n the
two-volume Shorter Oxford Dictionary, for example,
either the w o r d extraversion (extravert) or intro
version (introvert) i n their Jungian senses indis
pensable though these words have been to an enor
mous amount of conversation.
N o w what do these complaints amount to?
W h a t sort of case can be made out for them ? First
as to the awkwardness the uncertainty that may be
SOME C R I T E R I A OF WORDS 77

felt as to how to pronounce them, as to whether, for


example, epistemology should be e'pistemology, epis'-
temology, episte'mology or epistemoVogy. Sir James
M u r r a y i n h i s Preface to the Oxford Dictionary,
notes that, on applying "directly to the introducer
of a word, to k n o w how he pronounces i t , or means
i t to be pronounced," on several occasions, the an
swer received was "that he has never thought of its
pronunciation, does not presume to say how i t
ought to be pronounced, and leaves i t to people to
pronounce as they like, or to the D i C T i O N A R Y to say
what is the right pronunciation."
T h i s , Sir James complained, inverts the estab
lished order by which speech comes first. B u t
surely the surprised introducers were justified. Be
i n g themselves the very fountain-heads of usage they
knew better than to trust i t ! T h e y knew that the
good and bad of words have other standards. H o w a
w o r d should be pronounced is, at least i n part, a
matter r e q u i r i n g reference to how the other words
i n the language are pronounced. A n d that evi
dently is the lexicographer's not the philosopher's or
the psychologist's business. B u t unfortunately the
lexicographer is sometimes himself i n these mat
ters of pronunciation as i n matters of interpretation
too much under the spell of crude usage theo
ries. Daunted by his responsibilities he falls back
u p o n the task of making a phonetic record of pro
nunciations that have been established, or he takes
refuge i n a variety of the usage doctrine which I can
ticket here as obedience to the C l u b Spirit.
78 T H E PHILOSOPHY OF R H E T O R I C

T h i s is an important mode of the usage doctrine.


Essentially i t makes the conduct of language sub
servient to manners to the manners of a special
set of speakers. I f you belong to a certain sort of
C l u b you thereby enter u p o n an engagement to be
have, while there, i n certain ways or rather an en
gagement not to behave i n certain other ways. As
usual i t is m u c h easier to say what you w i l l not do
there, than what you w i l l . Similarly, i n using a
language, you j o i n a more or less select company
of correct users of the language. Deviation f r o m
their customs is incorrectness and is visited w i t h a
social penalty as such. A n d no account is taken, i n
this, of whether what you do is better or worse than
the customs of the company. I t is enough that i t is
different to b r i n g you i n t o condemnation.
T h i s specialized f o r m of control by usage, this
social or snob control over all language, is obviously
very wide and rigorous. One of the tasks of an
improved Rhetoric is to question i t , whether i t con
cerns pronunciation or matters of meaning and i n
terpretation. For the moment I am concerned w i t h
pronunciation, b u t I w o u l d like to insist that what
I am saying about pronunciation has its parallels
throughout the whole field and can be applied too
to variations and specializations of meaning. Thus
i t was the C l u b Spirit i n England which made us say
' T m s o r r y ! " when you over here say " I beg your
pardon I " A p a r t f r o m the C l u b Spirit, your phrase
is perhaps the better of the two, w i t h less risk of
confusing the important w i t h the t r i v i a l . The
SOME C R I T E R I A OF WORDS 79

effects of questioning the u t i l i t y of the Rule of the


C l u b Spirit w o u l d I believe be extremely drastic.
T h e reasons for questioning i t are, since i t is itself a
social rule, themselves social. I n the past, I am w i l l
i n g to believe, such snob control was often useful
to the whole community, not rnerely, as now, to the
members of the Club, who find i n i t an advantage
over their fellow citizens. T h i s use of verbal differ
ences as weapons i n the class-war dates for us, i n
some i m p o r t a n t respects, f r o m the 17 t h Century.
I n Shakespeare's age i t seems probable that a less de
rogatory and a more humorous notice was taken of
differences i n speech. T h e r e was less need to be
scornful. I t was because a new stratification of so
ciety had arisen that the early 18th Century began
to observe that niceties of pronunciation and expres
sion constituted the most certain differentiation be
tween a gentleman and his valet, between a lady and
a mantua-maker. T h e new effort towards u n i f o r m
spelling is another aspect of the same change. A n d
i t was then that a preoccupation w i t h correctness ( i n
this C l u b Spirit sense) became the obsession of the
grammar-book merchants of those (Steele is an
example) who purveyed instruction to the new
gentry about how they were to make i t clear that
they were reallygentry.
T h a t is the h u m i l i a t i n g side of the R u l e of the
C l u b Spirit. B u t there is a w o r t h i e r side. I n the
i 8 t h C e n t u r y - w h e n comparatively few were edu
cated, and when the education given was relatively
all of a piece this sort of correctness d i d give some
8o T H E PHILOSOPHY OF R H E T O R I C

reliable indication of culture i n a deeper sense.


Nowadays we educate a far larger p r o p o r t i o n of a
population ten times more numerous, and, what is
more important, education i n the humanities is no
longer u n i f o r m . I f asked for a definition of the
humanities now, we w o u l d have to reply, " A n y t h i n g
that has anything to do w i t h anything i n the Metro
politan or the British Museums." A n d , w i t h that,
the Rule of the C l u b Spirit ceases to guarantee any
t h i n g i m p o r t a n t about the depth of the culture of
those who speak correctly or incorrectly, according or
not according to its terms.
A n d y e t h o w strong i t still is. I suppose a large
part of the justification of lectures i n many colleges,
is that they ensure that those who attend them w i l l
know how to pronounce the names of Italian paint
ers and Greek heroines. Even extensive reading i n
prose translations of the classics w i l l not protect us
f r o m the awful dangers of saying Saloam, or Penny-
lope or Hermy-one. I w e l l remember a worthy
young auto-didact f r o m Manchester bursting i n
upon me to announce w i t h deep enthusiasm that he
had become, i n the vacation, "Desperately keen o n
D a n t a n d Goath." I do not believe that he w o u l d
have read them w i t h a w h i t more genuine profit had
he k n o w n that he was really reading Dante and
Goethe.
Enough about these breaches of the C l u b Spirit.
Let us go back to the complaints against new scien
tific terms. 'Conciseness and p r o n u n c i a b i l i t y / said
Jeremy Bentham, are merits i n a new<xnned w o r d .
SOME C R I T E R I A OF WORDS 81

A queer doctrine, we may t h i n k , to come f r o m a


man who, himself, for example, coined as key-terms
i n the analysis of metaphor which he d i d so m u c h
to advance the words archetypation and phraseo-
plerosis: for the discovery of the ground of the shift
of meaning and the filling i n of the phrase to rep
resent that ground. H o w about conciseness as a
merit? Must we not qualify i t w i t h that useful
phrase 'other things being equal'? A n d shall we
not then agree that, with words, other things are
never equal and that very often length too
( w i t h i n limits) may be a m e r i t i n a word? Most
especially when, as often w i t h these scientific words,
the meanings they have to carry are complex. I t
is an advantage w i t h many scientific words that they
should look scientific and should r e m i n d us that
they belong to a system and depend u p o n assump
tions which we must take account of. A n d there
comes i n the answer to that other complaint that
so many of these words (introversion and extraver
sion, for example) are explanations, not labels.
For familiar things, i t is often said, we need a label,
not a description. Yes! i f the things are really fa
m i l i a r . B u t the dangers of mere labels when the
things are not really familiar of labels w h i c h give
no h i n t of what they are attached to should hardly
need p o i n t i n g out.
T o pass to another type of complaint that these
words are cumbrous and ugly i n themselves. I have
seen i t urged and by no slight authority that
whereas good o l d words like mind and thought are
8* T H E PHILOSOPHY OF R H E T O R I C

neat, concise and beautiful, a w o r d like psychology


is cumbrous and disagreeable. H o w sound a com
plaint is that? A n d is i t really a complaint against
the f o r m of the w o r d or against some of its uses?
Let us grant that some of the derivative uses of
psychology are objectionable because inconven
iently and unnecessarily ambiguous as when some
one persists i n t a l k i n g and w r i t i n g of Shakespeare's
psychology w i t h o u t l e t t i n g us see whether he means
(1) Shakespeare's theories, i f any, about the m i n d ,
(2) the assumptions Shakespeare unconsciously
made about mental processes, (3) the inference as
to mental processes we m i g h t arrive at f r o m Shake
speare's work, or (4) (to go n o further i n t o these
possibilities) just the way Shakespeare's o w n m i n d
worked. These vagaries of the w o r d are typical and
unfortunate. T h e y endanger discourse, and m u c h
use of that sort of language r i g h t l y discredits a
speaker or writer. B u t s u c h uses otpsychology are
no ground for complaint against the w o r d i n its use
for the theoretical study of how the m i n d works,
or i n derivative uses where the context takes care
of them. T h e complaint against Shakespeare's psy-
chology is really against the inadequate contextual
control. A n d for controlled uses of the w o r d see
i n g what a cumbrous subject psychology is a cum
brous w o r d may have therein its recommendation.
Such a w o r d may, however, be tainted, for many,
even i n its prime use, by its associations w i t h u n
happy uses. T h i s is a common case especially
w i t h new words and, I t h i n k , instructive. I t can
SOME C R I T E R I A OF WORDS 83

be illustrated w i t h colorful. A violent distaste


for this w o r d , as for ta$te/w/,isprevalent i n many
circles i n England. C o l o r f u l came i n seemingly
about 1890. O n the principle that any stick w i l l
do to beat a dog w i t h , a l l sorts of reasons are
brought out to justify dislike for i t . T h a t i t is
a h y b r i d , that i t is vulgar, that we do not say
soundful or lightful, that i f we use i t we shall
soon be using lifeful and laughterful, that we al-
readyhave ' f u l l of color' which does its work, and
so on. None of these objections w i l l bear exami
nation. O u r language has too many lusty and valu
able hybrids i n i t already for that to matter (com
pare beautiful and joyful which w o u l d be equally
hybrids). Most of the other objections are drawn
f r o m analogy and may fairly be met by pursu
i n g the analogies further. Y o u w i l l see that i n
doing s o w e w o u l d be considering the same sorts of
situations, i n w h i c h a w o r d is backed u p by other
words, that we considered w i t h the morphemes last
week and that the analogies which w o u l d matter
w o u l d not be those that were only evident to the
philologist b u t those actually operative, on occasion,
i n backing u p or deflecting (i.e., steering) our use of
the words.
T h e interesting objection, against such words
as colorful, concerns vulgarity. I t is an objec
t i o n that many new words are open to because
they are often taken u p most readily by people
the objectors like to describe so. Moreover, that
a w o r d should be popular is often the chief con-
84 T H E PHILOSOPHY OF R H E T O R I C

d i t i o n for its admission i n t o the language and


to some people popular means what vulgar means.
B u t a w o r d like colorful can evidently be used i n
many different ways w i t h different meanings, and to
compare these is the way to judge i t . W e w o u l d
not judge a quadruped w i t h o u t deciding whether
i t is a horse or a dog. So, we ought not to judge
a w o r d w i t h o u t considering what sorts of uses i t is
peculiarly suited to and w i t h o u t remembering that
a good ferret makes a bad rabbit. W h a t then are
the peculiar utilities of the w o r d colorful?
First, I w o u l d draw attention to the ironical i m
plications i t is capable of conveying. L i k e the
phrases, hard-working, painstaking, does his best,
industrious and well-intentioned,-and others w h i c h
t u r n u p i n schoolmasters' reports, i t can suggest
that i f this is the best that can be said about some
thing, well, we know where we are! T o call a
prose-style, or a dramatic production, colorful and
leave i t at that, can be a very polite and therefore
very effective way of damning i t w i t h faint praise.
I t is the more effective because i t suggests that those
who w i l l be content w i t h i t as straight praise are
not b r i n g i n g a sufficient critical apparatus to bear
u p o n the matter. T h e r e are, of course, uses of
colorful which have no such implications where,
for example, that a t h i n g should be f u l l of color
is all we can ask, where no ironical reserves and no
disparagement can be intended. A n d indeed i t is
the fact that there are these other straight uses, and
that confusion between the straight and the derog-
SOME C R I T E R I A OF WORDS 85

atory uses is easy and frequent, that gives t h e w o r d


its peculiar subtlety on occasion.
A n d this confusion too is, I think, the source of
the distaste for the word. I f i t is used straight,
when the ironical implication w o u l d be i n place, i t
suggests a lack of discrimination i n the user, which
unless we analyse we allow to infect the word
itself. I f beautiful were not such an old and w e l l
understood w o r d we m i g h t let the same t h i n g hap
pen w i t h i t . Gross uses of beautiful m i g h t make
the w o r d itself a t h i n g suited only to gross uses.
W h e n people talk of 'beautiful food' some are apt
to shudder. W h e n M r . E l i o t i n The WasteLand
makes one of his characters say,

Well, that Sunday Albert was home, they had a hot


gammon,
And they asked me in to get the beauty of it hot

he is using that shudder, and a l l the pathetic rever


berations f r o m its occasion and its contrasts. T h a t
is the f u l l use of language which dramatic w r i t i n g
more than any other, of course, requires. I t takes
its w o r d , not as the repository of a single constant
power b u t as a means by which the different powers
i t may exert i n different situations are brought to
gether and again w i t h an interinanimating appo
sition.
I have taken colorful as a type word. Its own
peculiar problems are local, perhaps temporary and
u n i m p o r t a n t b u t i f we pursue them we find that
they lead us to most of the problems of the choice
86 T H E PHILOSOPHY OF R H E T O R I C

of words, and f u r t h e r s t i l l that they b r i n g i n t o view


most of the problems of aesthetics. T o realize that
i t is idle to ask of a w o r d , "Is i t beautifulP"-unless
we are ready to ask thoroughly, " W h a t w i l l i t do
i n its variedincidencesP"-is a first step and a long
step i n the aesthetics of language. A parallel step
must be made for every branch of aesthetics. A dis
cussion of the reasons for the choice of words w h i c h
too often seems a t r i v i a l exchange of whimsies can
become an i n t r o d u c t i o n to the theory of a l l choices.
T h e art of so transforming i t f r o m a tea-table topic
i n t o the central discipline of education waits to
be rediscovered, b u t the better we understand what
place words h o l d i n our lives the readier we shall
be to admit that to t h i n k about their choice is the
most convenient mode of t h i n k i n g about the p r i n
ciple of a l l o u r choices.
LECTURE V

METAPHOR
Again, the more the mind knows, the better it under
stands its forces and the order of nature; the more it
understands its forces or strength, the better it will be
able to direct itself and lay down rules for itself; and the
more it understands the order of nature, the more easily
it will be able to liberate itself from useless things:
of this, as we have said, consists the whole method.
Spinoza, De intellectus emendatione.
LECTURE V

METAPHOR

I T W A S Aristotle, no lesser man, who said, i n


The Poetics, " T h e greatest t h i n g by far is to
have a command of metaphor." B u t he went on
to say, " T h i s alone cannot be imparted to another:
i t is the mark of genius, for to make good metaphors
implies an eye for resemblances." I do not know
how m u c h influence this remark has h a d : or whether
i t is at a l l responsible for our feeling that what
i t says is common-sense. B u t question i t for a mo
ment and we can discover i n i t , i f we w i l l to be
malicious, here at the very beginning of the sub
ject, the evil presence of three of the assumptions
which have ever since prevented the study of this
'greatest t h i n g by far' f r o m taking the place i t de
serves among o u r studies and f r o m advancing, as
theory and practice, i n the ways open to i t .
One assumption is that 'an eye for resemblances'
is a gift that some m e n h a v e b u t others have not.
B u t we a l l live, and speak, only through our eye
for resemblances. W i t h o u t i t we should perish
early. T h o u g h some may have better eyes than
others, the differences between them are i n degree
only and may be remedied, certainly i n some meas
ure, as other differences are, by the r i g h t kinds of
90 T H E PHILOSOPHY OF R H E T O R I C

teaching and study. T h e second assumption denies


this and holds that, though everything else may be
taught, " T h i s alone cannot be imparted to another."
I cannot guess how seriously Aristotle meant this
or what other subjects of teaching he had i n m i n d
as he spoke. But, i f we consider how we a l l of us
attain what l i m i t e d measure of a command of meta
phor we possess, we shall see that no such contrast
is valid. As individuals we gain o u r command of
metaphor just as we learn whatever else makes us
distinctively human. I t is a l l imparted to us f r o m
others, w i t h and t h r o u g h the language we learn,
language w h i c h i s u t t e r l y unable to aid us except
through the command of metaphor w h i c h i t gives.
A n d that brings u p the t h i r d and worst assumption
that metaphor is something special and excep
tional i n the use of language, a deviation f r o m its
normal mode of w o r k i n g , instead of the omnipres
ent principle of a l l its free action,
j T h r o u g h o u t the history of Rhetoric, metaphor
j has been treated as a sort ofjhappy extra trick w i t h
words, an o p p o r t u n i t y to exploit the accidents of
their versatility, something i n place occasionally b u t
I r e q u i r i n g unusual skill and caution. I n brief, a
grace or ornament or added power of language, not
i its constitutive f o r m . Sometimes, i t is true, a w r i t e r
w i l l venture on speculations that go deeper. I have
just been echoing Shelley's observation that " L a n
guage is v i t a l l y metaphorical; that is, i t marks the
before unapprehended relations of things and per
petuates their apprehension, u n t i l words, w h i c h

METAPHOR 91

represent them, become, through time, signs for


portions or classes of thought instead of pictures of
integral thoughts: and then, i f no new poets should
arise to create afresh the associations which have
been thus disorganised, language w i l l be dead to
all the nobler purposes of h u m a n intercourse."
B u t that is an exceptional utterance and its i m p l i
cations have not yet been taken account of by rhet
oricians. N o r have philosophers, as a body, done
m u c h better, though historians of language have
l o n g ' t a u g h t that we can find no w o r d or descrip
t i o n for any of the intellectual operations which,
if its history is k n o w n , is not seen to have been
taken, by metaphor, f r o m a description of some
physical happening. O n l y Jeremy Bentham, as
successor to Bacon and Hobbes, insisted w i t h his
technique of archetypation and phraseoplerosis
upon one inference that m i g h t be d r a w n ; namely,
that the m i n d and a l l its doings are fictions. H e
left i t to Coleridge, F. H . Bradley and Vaihinger to
point to the f u r t h e r inference; namely, that matter
and its adventures, and a l l the derivative objects of
contemplation, are fictions too, of varied rank be
cause of varied service.
I have glanced for a moment at these deep waters
i n t o which a serious study of metaphor may plunge
us, because possibly fear of them may be one cause
why the study has so often not been enterprising
and why Rhetoric traditionally has l i m i t e d its i n
q u i r y to relatively superficial problems. B u t we
shall not advance i n even these surface problems
9 T H E PHILOSOPHY OF R H E T O R I C

unless we are ready to explore, as best we can, the


depths of verbal interaction w h i c h give rise to
them.
T h a t metaphor is the omnipresent principle of
language can be shown by mere observation. W e
cannot get through three sentences of ordinary fluid
discourse w i t h o u t i t , as you w i l l be noticing through
out this lecture. Even i n the r i g i d language of
the settled sciences we do not eliminate or prevent
> i t w i t h o u t great difficulty. I n the semi-technicalised
subjects, i n aesthetics, politics, sociology, ethics, psy
chology, theory of language and so on, our constant
chief difficulty is to discover how we are using i t
and how our supposedly fixed words are shifting
their senses. I n philosophy, above all, we can take
no step safely w i t h o u t an unrelaxing awareness of
the metaphors we, and our audience, may be em
p l o y i n g ; and though we may pretend to eschew
them, we can attempt to do so only by detecting
them. A n d this is the more true, the more severe
and abstract the philosophy is. As i t grows more
abstract we t h i n k increasingly by means of meta
phors that we profess not to be relying on. T h e
metaphors we are avoiding steer our thought as
much as those we accept. So i t must be w i t h any
utterance for w h i c h i t is less easy to k n o w what
we are saying than what we are not saying. A n d
i n philosophy, of w h i c h this is almost a definition,
I w o u l d h o l d w i t h Bradley that o u r p r e t e n c e to
do w i t h o u t metaphor is never more than a bluff
waiting to be called. B u t i f that is a t r u t h , i t is
METAPHOR - 9S

easier to utter than to a c c e p t w i t h its consequences


or to remember.
T h e view that metaphor is omnipresent i n speech
can be recommended theoretically. I f you r e c a l l '
what I tried to say i n my Second Lecture about the
context theorem of m e a n i n g ; about meaning as the
delegated efficacy of signs by w h i c h they b r i n g to
gether i n t o new unities the abstracts, or aspects,
w h i c h are the missing parts of their various con
texts, y o u w i l l recollect some insistence that a w o r d ;
is normally a substitute for (or means) not o n e i
discrete past impression b u t a combination of gen-!
eral aspects. N o w that is itself a summary accountj
of the principle of metaphor. I n the simplest \
f o r m u l a t i o n , when we use a metaphor we have two
thoughts of different things active together and s u p
ported by a single w o r d , or phrase, whose meaning
is a resultant of their interaction.
"As to metaphorical expression," said D r . J o h n -
son, " t h a t is a great excellence i n style, when i t is
used w i t h propriety, for i t gives you two ideas for
one." H e is keeping, you see, to the l i m i t e d tradi
tional view of metaphor. As to the excellence of a
style that gives you two ideas for one, that depends
on what the two ideas do to one another, or con
j o i n t l y do for us. W e find, of course, when we
look closer that there is an immense variety i n these
modes of interaction between co-present thoughts,
as I w i l l call them, or, i n terms of the context the
orem, between different missing parts or aspects of
t h e d i f f e r e n t contexts of a word's meaning. I n
94 T H E PHELOSOPHY OF R H E T O R I C

practice, we distinguish w i t h marvellous s k i l l be


tween these modes of interaction, though o u r skill
varies. T h e Elizabethans, for example, were far
more widely skilled i n the use of metaphor b o t h
i n utterance and i n interpretation than we are.
A fact which made Shakespeare possible. T h e i 8 t h
Century narrowed its skill down, defensively, t o cer
tain modes only. T h e early 19th Century revolted
against this and specialized i n other modes. The
later 19th Century and my generation have been
recovering f r o m these two specializations. T h a t , I
suggest, is a way of reformulating the Classic-
Romantic antithesis which i t w o u l d be interesting to
try out.
B u t i t could not be tried out w i t h o u t a better
developed theory of metaphor than is yet available.
T h e traditional theory noticed only a few of the
modes of metaphor; and l i m i t e d its application of
the t e r m metaphor to a f e w of t h e m only. A n d
thereby i t made metaphor seem to be a verbal mat-
jter, a shifting and displacement o f w o r d s , w h e r e a s
fundamentally i t is a b o r r o w i n g between and inter
course of thoughts, a transaction between contexts.
Thought is metaphoric, and proceeds by compari
son, and the metaphors of language derive there
from. T o improve the theory of metaphor we must
remember this. A n d the method is to take more
note of the skill i n thought which we possess and
are i n t e r m i t t e n t l y aware of already. W e must
translate more of o u r skill into discussable science.
Reflect better u p o n what we do already so cleverly.
METAPHOR 95

Raise o u r i m p l i c i t recognitions i n t o explicit dis


tinctions.
As we do so we find that a l l the questions that
matter i n literary history and criticism take on a
new interest and a wider relevance to h u m a n needs.
I n asking how language works we ask about how
thought and feeling and all the other modes of the
mind's activity proceed, about how we are to learn
to live and how that "greatest t h i n g of a l l , " a com
mand of metaphor which is great only because i t
is a command of life may best, i n spite of Aristotle,
"be imparted to another." B u t to profit we must
remember, w i t h Hobbes, that "the scope of a l l spec
ulation is the performance of some action or t h i n g
to be done" and, w i t h Kant, t h a t - " W e can by no
means require of the pure practical reason to be
subordinated to the speculative, and thus to reverse
the order, since every interest is at last practical,
and even that of the speculative reason is b u t con
ditional, and is complete only i n its practical use."
O u r theory, as i t has its roots i n practice, must also
have its f r u i t i n improved skill. " I am the c h i l d , "
says the Sufi mystic, "whose father is his son, and
the wine whose vine is its j a r , " summing u p so the
whole process of that meditation which does not
forget what i t is really about.

T h i s m u c h has been an i n t r o d u c t i o n or prepara


t i o n to p u t the theory of metaphor i n a more i m
portant place than i t has enjoyed i n traditional
Rhetoric. I t is time to come down f r o m these h i g h
96 T H E PHILOSOPHY OF R H E T O R I C

speculations to consider some simple steps i n analy


sis which may make the translation of o u r s k i l l w i t h
metaphor i n t o explicit science easier. A first step
is to introduce two technical terms to assist us i n
distinguishing f r o m one another what D r . Johnson
called the two ideas that any metaphor, at its sim
plest, gives us. L e t me call them the tenor and
the vehicle. One of the oddest of the many odd
things about the whole topic is that we have no
j agreed distinguishing/terms for these two halves of
a m e t a p h o r i n spite of the immense convenience,
almost the necessity, of such terms i f we are to make
any analyses w i t h o u t confusion. For the whole task
is to compare the different relations which, i n dif
ferent cases, these two members of a metaphor h o l d
to one another, and we are confused at the start i f
we do not k n o w w h i c h of the two we are t a l k i n g
about. A t present we have only some clumsy de-
j scriptive phrases w i t h w h i c h to separate them. ' T h e
original idea' and 'the borrowed one'; 'what is really
being said or thought o f and 'what i t is compared
to'; 'the u n d e r l y i n g idea' and 'the imagined nature';
'the p r i n c i p a l subject' and 'what i t resembles' or,
still more confusing, simply 'the meaning' and 'the
metaphor' or 'the idea' and 'its image.'
H o w confusing these must be is easily seen, and
experience w i t h the analysis of metaphors f u l l y con
firms the worst expectations. W e need the w o r d
'metaphor' for the whole double u n i t , and to use i t
sometimes for one of the two components i n sepa
ration f r o m the other is as injudicious as that other
METAPHOR 97

trick by which we use 'the meaning* here sometimes


for the w o r k that the whole double u n i t does and
sometimes for the other component the tenor, as
I am calling i t the underlying idea or principal
subject which the vehicle or figure means. I t is
not surprising that the detailed analysis of meta
phors, i f we attempt i t w i t h such slippery terms as
these, sometimes feels like extracting cube-roots i n
the head. Or, to make a more exact comparison,
what w o u l d the most elementary arithmetic feel like,
if we used the w o r d twelve (12) sometimes for
the number one (1), sometimes for the number
two (2) and sometimes for the number twenty-one
(21) as well, and had somehow to remember, or see,
unassisted by our notation, which uses we were
making of i t at different places i n our calculations ?
A l l these words, meaning, expression, metaphor,
comparison, subject, figure, image, behave so, and
when we recognize this we need look no further
for a part, at least, of the explanation of the back
ward state of the study. W h y rhetoricians have not
long ago remedied this defect of language for their
purpose, w o u l d perhaps be a profitable matter for
reflection. I do not know a satisfactory answer.
As the best teacher I ever knew, G. E. Moore, once
remarked, " W h y we should use the same form of
verbal expression to convey such different meanings
is more than I can say. I t seems to me very curi
ous that language should have grown u p as i f i t were
expressly designed to mislead philosophers; and I
do not k n o w why i t should have."
g8 T H E PHILOSOPHY OF R H E T O R I C

T h e words 'figure' and 'image' are especially and


additionally misleading here. T h e y b o t h some
times stand for the whole double u n i t and some
times for one member of i t , the vehicle, as opposed
to the other. B u t i n addition they b r i n g i n a con
fusion w i t h the sense i n which an image is a copy
or revival of a sense-perception of some sort, and
so have made rhetoricians t h i n k that a figure of
speech, an image, or imaginative comparison, must
have something to do w i t h the presence of images,
i n this other sense, i n the mind's eye or the mind's
ear. But, of course, i t need not. N o images of
this sort need come i n at any point. W e had one
instance of the vicious influence of this red-herring
i n my first lecture L o r d Kames' antic w i t h the
mental picture he supposed we must f o r m of Shake
speare's peacock-feather. W h o l e schools of rhetoric
and criticism have gone astray after i t . Lessing's
discussion of the relations of the arts, for example,
.is grievously spoilt by i t . We cannot too firmly
recognize that how a figure of speech works has
nothing necessarily to do w i t h how any images, as
copies or duplicates of sense perceptions, may, for
reader or writer, be backing u p his words. I n spe
cial cases for certain readers they may come i n
then is a long chapter of i n d i v i d u a l psychology
which is relevant here. B u t the words can do al
most anything w i t h o u t them, and we must p u t no
assumption about their necessary presence into o u r
general theory.
I can illustrate b o t h the convenience of such
METAPHOR 99

technical terms as tenor and vehicle and the evil


influence of the imagery assumption, w i t h another
citation f r o m L o r d Kames, f r o m Chapter 20, para
graph 6, of his Elements of Criticism. You w i l l
see f r o m the very difficulty of m a k i n g out just what
he is saying, how m u c h we n e e d r i g i d technicalities
here. His p o i n t is, I think, evidently mistaken;
b u t before we can be satisfied that i t is mistaken, we
have to be certain what i t i s ; and what I want first
to direct your attention u p o n is the clumsy and
distracting language i n which he has to state i t .
H e is preparing to set u p a rule to be observed by
writers i n 'constructing a metaphor.' H e says, " I n
the f o u r t h place, the comparison . . . being i n a
metaphor sunk by imagining the principal subject
to be that very t h i n g which i t only resembles; an
opportunity is furnished to describe i t (i.e., the
principal subject) i n terms taken strictly or literally
w i t h respect to its imagined nature."
T o use m y proposed terms we can describe or
qualify the tenor by describing the vehicle. He
goes on, " T h i s suggests another r u l e : T h a t i n con
structing a m e t a p h o r , t h e w r i t e r ought to make use
of such words only as are applicable literally to the
imagined nature of his subject." T h a t is, he must
not use any further metaphor i n describing the
vehicle. "Figurative words," he says, "ought care
f u l l y to be avoided; for such complicated figures,
instead of setting the principal subject i n a strong
light, involve i t i n a c l o u d ; and i t is w e l l i f the
reader, w i t h o u t rejecting by the l u m p , endeavour
ioo T H E PHILOSOPHY OF R H E T O R I C

patiently to gather the plain meaning, regardless of


the figures."
L e t me invite you to consider what is being done
here very carefully, for i t illustrates, I believe, most
of the things w h i c h have made the t r a d i t i o n a l
studies of metaphor not very profitable. A n d no
tice first how i t shows the i 8 t h Century assumptions
that figures are a mere embellishment or added
beauty and that the plain meaning, the tenor, is
what alone really matters and is something that,
'regardless of the figures/ m i g h t be gathered by the
patient reader.
A modern theory w o u l d object, first, that i n many
of the most i m p o r t a n t uses of metaphor, the co-
presence of the vehicle and tenor results i n a mean
i n g (to be clearly distinguished f r o m the tenor)
which is not attainable w i t h o u t their interaction.
T h a t the vehicle is not normally a mere embel
lishment of a tenor which is otherwise unchanged
by i t b u t that vehicle and tenor i n co-operation give
a meaning of more varied powers than can be
ascribed to either. A n d a modern theory w o u l d go
on to point out that w i t h different metaphors the
relative importance of the contributions of vehicle
and tenor to this resultant meaning varies i m
mensely. A t one extreme the vehicle may become
almost a mere decoration or coloring of the tenor,
at the other extreme, the tenor may become almost
a mere excuse for the i n t r o d u c t i o n of the vehicle,
and so no longer be 'the principal subject/ A n d the
METAPHOR

degree to w h i c h the tenor is imagined " t o be that


very t h i n g w h i c h i t only resembles" also varies i m
mensely.
These are differences I r e t u r n to next week. L e t
us study L o r d Kames a l i t t l e longer first: H o w
about this suggested r u l e that we should carefully
avoid m o u n t i n g metaphor u p o n metaphor? W h a t
w o u l d be the effect of taking i t seriously ? I t w o u l d ,
i f we accepted and observed i t , make havoc of
most w r i t i n g and speech. I t is disregarding under
cover of the convenient excuse that they are dead
the most regular sustaining metaphors of a l l
speech. I t w o u l d make, I think, Shakespeare the
faultiest w r i t e r who ever held a p e n ; and i t turns
an obstinately b l i n d eye u p o n one of the most ob
vious features of current practice i n every m i n u t e
of our speech. Look, for example, a t L o r d Kames'
o w n sentence. "Such complicated figures, instead
of setting the p r i n c i p a l subject i n a strong light,
involve i t i n a cloud/' W h a t about that 'strong'
light? T h e light is a vehicle and is described
w i t h o u t anyone experiencing the least difficulty
by a secondary metaphor, a figurative w o r d . B u t
you may say, " N o ! Strong is no longer a figurative
w o r d as applied to light. I t is as literally descrip
tive of l i g h t as i t is of a man or a horse. I t carries
not two ideas b u t one only. I t has become 'ade-
quated,' or is dead, and is no longer a metaphor."
B u t however stone dead such metaphors seem, w e l
can easily wake them up, and, i f Kames w e r e r i g h t , \ 1
io T H E PHILOSOPHY OF R H E T O R I C

to wake them u p w o u l d be to risk i n v o l v i n g the


tenor i n a cloud, and n o t h i n g of the sort happens.
T h i s favourite o l d distinction between dead and liv
i n g metaphors (itself a two-fold metaphor) is, i n
deed, a device which is very often a hindrance to
the play of sagacity and discernment throughout the
subject. For serious purposes i t needs a drastic
reexamination.
W e are i n fact immeasurably more adroit i n han
d l i n g complicated metaphors than Kames w i l l allow
us to be. H e gives an example of a breach of his
rule which is w o r t h examining i f only to show how
easily a theory can paralyse n o r m a l aptitude i n such
things. H e takes these two lines

A stubborn and unconquerable flame


Creeps in his veins and drinks the streams of life.

" L e t us analyse this expression," he says. " T h a t


a f e v e r m a y be imagined aflame, I a d m i t ; though
more than one step is necessary t o come at the re
semblance." I , for m y part, w o u l d have supposed,
on the contrary, that we could,hardly find a simpler
transference, since b o t h a fever and a flame are i n
stances of a rise i n temperature! B u t he goes o n
to detail these steps. " A fever by heating the body,
resembles fire; and i t is no stretch to imagine a
fever to be a fire. Again, by a figure of speech,
flame may be p u t for fire, because they are com
monly conjoined; and therefore a fever may be
termed a flame. B u t now, a d m i t t i n g a fever to be
a flame, its effects ought to be explained i n words
METAPHOR 105

that agree literally to a flame. T h i s rule is not ob


served here; for a flame drinks figuratively only, not
properly."
W e l l and g o o d ! B u t who, for a l l that, has any
difficulty i n understanding the lines? T h e inter
actions of tenor and vehicle are not i n the least ham
pered by the secondary vehicle.
I have taken this instance of vain pedantry chiefly
to accustom you to m y use of these technical terms,
b u t partly too to support the contention that the
best part of the traditional discussion of metaphor
is hardly more than a set of cautionary hints to over-
enthusiastic schoolboys, hints masquerading as fun
damental theory of language. L o r d Kames is not
exceptionally l i m i t e d i n his treatment or abnormally
obtuse. Y o u w i l l find similar things i n Johnson
when he discusses Cowley and Donne for example,
i n Monboddoe, and Harris and Withers, and Camp
bell, i n a l l the chief 18th Century Rhetoricians.
N o t u n t i l Coleridge do we get any adequate set
t i n g of these chief problems of language. B u t
Coleridge's thought has not even yet come i n t o its
j
o w n . A n d , after Coleridge, i n spite of the possi
bilities w h i c h he opened, there was a regrettable
slackening of interest i n the questions. T h e 18th
Century was mistaken i n the way i t p u t them and
i n the technique i t attempted to use, b u t i t at least
knew that they were i m p o r t a n t questions and that
there is u n l i m i t e d w o r k to be done u p o n them.
A n d so L o r d Kames' Elements of Criticism, though
I may seem to have been making f u n of i t i n places,
io4 T H E PfflLOSOPHY OF R H E T O R I C

and though i t is so f u l l of similar things as to be


most absorbing reading, is still a very valuable and
instructive book offering a model not only of mis
conceptions to be avoided b u t of problems to be
taken up, reframed and carried forward. T u r n i n g
his pages you w i l l again and again find points raised,
which, i f his treatment of them is unsatisfactory,
are none the less points that no serious study of
language should neglect. One such w i l l serve me
as a peg for a pair of warnings or morals of which
any ambitious attempt to analyse metaphors is con
stantly i n need.
Kames quotes f r o m Othello the single l i n e

Steep'd me in poverty to the very lips

and comments, " T h e resemblance is too faint to be


agreeable Poverty must here be conceived to be
a fluid which i t resembles not i n any manner." Let
us look at Othello's whole speech. W e shall find
that i t is not an easy matter to explain o r j u s t i f y
that 'steep'd.' I t comes, y o u w i l l recall, when
Othello first openly charges Desdemona w i t h u n
faithfulness,
Had it pleas'd heaven
To try me with affliction, had he rain'd
All kinds of sores, and shames, on my bare head,
Steep'd me in poverty to the very lips,
Given to captivity me and my utmost hopes,
I should have found in some part of my soul
A drop of patience; but alas! to make me
The fixed figure for the time of scorn
To point his slow and moving finger at;
METAPHOR
Yet could I bear that too; well, very well.
But there, where I have garner'd up my heart,
Where either I must live or bear no life,
The fountain from the which my current runs,
Or else dries up; to be discarded thence I
Or keep it as a cistern for foul toads
To knot and gender in I

W h a t are we to say of that w o r d steep, how answer


Kames? H e is indeed too m i l d , i n saying "the re
semblance is too faint to be agreeable." I t s not a
case of a lack of resemblance b u t of too m u c h d i
versity, too m u c h sheer oppositeness. For Poverty,
the tenor, is a state of deprivation, of desiccation;
b u t the vehicle the sea or vat i n w h i c h Othello is
to be steeped gives an instance of superfluity. I n
poverty a l l is outgoing, w i t h o u t income; were we
"steeped to the very l i p s " i t w o u l d be the incomings
that we w o u l d have to fight against.* Y o u w i l l have
noticed that the whole speech returns again and
again to these l i q u i d images: "had they rained," "a
d r o p of patience," " T h e fountain f r o m the w h i c h
m y current runs, O r else dries u p . " None of these
helps steep out, and one of them "a drop of patience"
makes the confused, disordered effect of steep seem
m u c h worse. I do not myself find any defence of
the w o r d except this, w h i c h seems indeed quite suffi
cientas dramatic necessities commonly are that
O t h e l l o is himself h o r r i b l y disordered, that the ut
terance is part of "the storm of h o r r o u r and outrage"

In the partly parallel 'And steep my senses in forgetfulness' (Henry


IV, P. I I , in, i) Lethe, by complicating the metaphor, removes the
difficulty.
io6 T H E PHILOSOPHY OF R H E T O R I C

w i t h which he is assailing Desdemona and that a mo


mentarily deranged m i n d speaks so and is obsessed
w i t h images regardless of their fittingness. Othello,
we m i g h t say, is d r o w n i n g i n this storm, (Cf. Act I I ,
i , 212-21) and knows i t .
T h e morals I w o u l d point w i t h this instance are:
First, that n o t t o see how a w o r d can work is never
by itself sufficient proof that i t w i l l not work. Sec
ond, conversely, that to see how i t ought to work
w i l l not prove that i t does. A n y detailed examina
t i o n of metaphor brings us into such risk of pedantry
and self-persuasion, that these morals seem w o r t h
stress. Yet a critical examination of metaphor, w i t h
these morals i n m i n d , is just now what literary c r i t i
cism chiefly needs.
T o come back to Kames, his objection that "the
resemblance is too faint to be agreeable" (notice
the amusing assumption that a w r i t e r must of course
always a i m to be agreeable!)-assumed that tenor
and vehicle must be l i n k e d by their resemblance and
that their interaction comes about through their re
semblance one to another. A n d yet Kames himself
elsewhere takes some pride, and justifiably, i n point
ing out a type of figure which does not depend u p o n
resemblance b u t upon other relationsbetween tenor
and vehicle. H e says that i t has been overlooked
by former writers, and that i t must be distinguished
f r o m other figures as depending on a different p r i n
ciple.
"Giddy brink, jovial wine, daring wound are ex
amples of this figure. Here are adjectives that can-
METAPHOR 107

not be made to signify any quality of the substan- !


tives to w h i c h they are j o i n e d : a brink, for example, |
cannot be termed giddy i n a sense, either proper or
figurative, that can signify any of its qualities or
attributes. W h e n we examine attentively the ex
pression, we discover that a brink is termed giddy
f r o m p r o d u c i n g t h a t effect i n those who stand o n
i t . . . H o w , " he asks, "are we to account for this
figure, w h i c h we see lies i n the thought ( I am not
sure what lies means here. I t h i n k he means 'has
its g r o u n d or explanation i n the thought' not 'utters
falsehood.') and to what principle shall we refer i t ?
Have the poets a privilege to alter the nature of
things, and at pleasure to bestow attributes u p o n a
subject to which they do not belong?" Most mod
erns w o u l d say " O f course, they have !" B u t Kames
does not take that way out. H e appeals instead to a
principle of contiguous association. " W e have had
often occasion to inculcate, that the m i n d passeth eas
i l y and sweetly along a train of connected objects, and,
when the objects are intimately connected, that i t is
disposed to carry along the good or bad properties of
one to another, especially when i t is i n any degree i n
flamed w i t h these properties." H e then lists eight va
rieties of these contiguous inflammations w i t h o u t , I
t h i n k , at a l l clearly realizing what an immense exten
sion of the theory of possibilities of metaphoric inter
action he has made w i t h this new principle. Once
we begin 4o examine attentively' interactions which
do not w o r k through resemblances between tenor and
vehicle, b u t depend u p o n other relations between
io8 T H E PHILOSOPHY OF R H E T O R I C

them i n c l u d i n g disparities, some of o u r most prev


alent, over-simple, r u l i n g assumptions about meta
phors as comparisons are soon exposed.
B u t let us take one more glance at this giddy
brink first. Is Kames r i g h t i n saying that a brink
cannot be termed giddy i n a sense that can signify
any of its qualities or attributes? Is he r i g h t i n
t u r n i n g giddy i n t o giddy-making "*, b r i n k is
termed giddy f r o m producing that effect i n those
who stand on i t " ? Is i t not the case that at the mo
ment of giddiness the b r i n k itself is perceived as
swimming? As the man totters i n vertigo, the
w o r l d spins too and the b r i n k becomes not merely
giddy-making b u t actually vertiginous, seems itself
to stagger w i t h a dizziness and to w h i r l w i t h a be
w i l d e r i n g rapidity. T h e eyes nystagmically r o l l i n g
give away their m o t i o n to the w o r l d i n c l u d i n g the
b r i n k . T h u s the b r i n k as perceived, w h i c h is the
b r i n k that the poet is speaking of, actually itself ac
quires a giddiness. I f so, we may doubt for a mo
ment whether there is a metaphor here at all u n t i l
we notice how this w h i r l i n g that infects the w o r l d
as we grow giddy comes to i t by a process w h i c h is
itself radically metaphoric. O u r eyes twitch, b u t
i t is the w o r l d which seems to spin. So i t is w i t h a
large part, perhaps, i n the final account, w i t h all o u r
perceptions. O u r w o r l d is a projected w o r l d , shot
through w i t h characters lent to i t f r o m our o w n life.
" W e receive b u t what we give." T h e processes of
metaphor i n language, the exchanges between the
METAPHOR 109

meanings of words which we study i n explicit verbal


metaphors, are super-imposed u p o n a perceived
w o r l d which is itself a product of earlier or unwit
t i n g metaphor, and we shall not deal w i t h them
justly i f we forget that this is so. T h a t is why, i f
we are to take the theory of metaphor further than
the 18th Century took i t , we must have some gen
eral theorem of meaning. A n d since i t was Cole
ridge who saw most deeply and clearly i n t o this
necessity, and, w i t h his theory of the imagination,
has done most to supply i t , I may fittingly close this
Lecture w i t h a passage f r o m A p p e n d i x C of The
Statesman's Manual, i n which Coleridge is stating
that theory symbolically.
A symbol, for h i m , is a translucent instance, which
" w h i l e i t enunciates the whole,abides itself as a
l i v i n g part of that u n i t y of which i t is the repre
sentative." So here he takes the vegetable kingdom,
or any plant, as an object of meditation through and
i n which to see the universal mode of imagination
of those metaphoric exchanges by which the i n d i
vidual life and its w o r l d grow together. I f we can
follow the meditation we are led, I believe, to Cole
ridge's conception of imaginative growth more easily
and safely than by any other road. For, as the plant
here is a symbol, i n his sense, of all growth, so the
passage too is itself a symbol, a translucent instance,
of imagination.
H e has been speaking of the book of Nature that
"has been the music of gentle and pious minds i n
no T H E PHILOSOPHY OF R H E T O R I C

a l l ages, i t is the poetry of a l l h u m a n nature, to read


i t likewise i n a figurative sense, and to find therein
correspondences and symbols of the spiritual w o r l d .
" I have at this moment before me, i n the flowery
meadow, o n w h i c h m y eye is now reposing, one of
its most soothing chapters, i n w h i c h there is no la
m e n t i n g w o r d , no one character of g u i l t or anguish.
For never can I look and meditate on the vegetable
creation, w i t h o u t a feeling similar to that w i t h
w h i c h we gaze at a beautiful infant that has fed i t
self asleep at its mother's bosom, and smiles i n its
strange dream of obscure yet happy sensations.
T h e same tender and genial pleasure takes posses
sion of me, and this pleasure is checked and drawn
i n w a r d by the like aching melancholy, by the same
whispered remonstrance, and made restless by a
similar impulse of aspiration. I t seems as i f the
soul said to herself: F r o m this state hast thou fallen !
Such shouldst t h o u still become, thy Self a l l per
meable to a holier power I thy self at once h i d d e n
and glorified by its o w n transparency, as the acci
dental and dividuous i n this quiet and harmonious
object is subjected to the life and l i g h t of nature
which shines i n i t , even as the transmitted power,
love and wisdom, of God over a l l , fills and shines
through n a t u r e ! B u t what the plant is, by an act
not its o w n and unconsciously that must t h o u
make thyself to become! must by prayer and by a
watchful and unresisting spirit, j o i n at least w i t h the
preventive and assisting grace to make thyself, i n
METAPHOR iu

that l i g h t of conscience which inflameth not, and


w i t h that knowledge which puffeth n o t u p !
"Butfurther. . . Iseemtomyselftobeholdinthe
quiet objects o n which I am gazing, more than an
arbitrary illustration, more than a mere simile, the
w o r k of m y o w n fancy. I feel an awe, as i f there
were before m y eyes the same power as that of the
reason the same power i n a lower dignity, and
therefore a symbol established i n the t r u t h of things.
I feel i t alike, whether I contemplate a single tree
or flower, or meditate o n vegetation throughout the
w o r l d , as one of the great organs of the life of na
ture. L o ! - w i t h the rising sun i t commences its
outward life and enters i n t o open communion w i t h
all the elements at once assimilating them to itself
and t o each other. A t the same moment i t strikes
its roots and unfolds its leaves, absorbs and respires,
steams f o r t h its cooling vapour and finer fragrance,
and breathes a repairing spirit, at once the food and
tone of the atmosphere, i n t o the atmosphere that
feeds it. L o ! at the touch of l i g h t h o w i t returns
an air akin t o light, and yet w i t h the same pulse
effectuates its o w n secret growth, still contracting to
fix what expanding i t had refined. L o ! how up
h o l d i n g the ceaseless plastic m o t i o n of the parts i n
the profoundest rest of the whole, i t becomes the
visible organismus of the whole silent or elementary
life of nature and therefore, i n incorporating the
one extreme becomes t h e s y m b o l of the other; the
natural symbol of that higher life of reason."
na T H E PHILOSOPHY OF R H E T O R I C

W h a t Coleridge has here said of this "open com


m u n i o n " is true also of the word i n the free meta-
phoric discursive sentence. " A r e not words," he
had asked nineteen years before, " A r e not words
parts and germinations of the plant?"
LECTURE V I

METAPHOR (continued)
All life therefore comes back to the question of our
speech, the medium through which we communicate with
each other; for all life comes back to the question of
our relations with one another. Henry James, The Ques-
tion of our Speech.
LECTURE VI

THE C O M M A N D OF METAPHOR

H E N , i n my last lecture, I spent so much time


V u p o n L o r d Kames' theories of metaphor i t
was because he, better than anyone else I know, illus
trates the limitations of traditional treatment, and
shows why these limitations are unnecessary. T h e
neglect of the study of the modes of metaphor i n
the later 19th Century was due, I t h i n k , to a gen
eral feeling that those methods of i n q u i r y were un
profitable, and the time was not ripe for a new
attack. I am not sure that i t is yet ripe i n spite of
all that Coleridge and Bentham d i d towards ripen
i n g i t . Very likely a new attempt must again lead
into artificialities and arbitrarinesses. I f so, their
detection may again be a step o n the road. I n this
subject i t is better to make a mistake that can be
exposed than to do n o t h i n g , better to have any ac
count of how metaphor works (or thought goes on)
than to have none. Provided always that we do not
suppose that our account really tells us what hap
pensprovided, that is, we do not mistake our the
ories for our skill, or our descriptive apparatus for
what i t describes. T h a t is the recurrent mistake
which the 18th Century doctrines exemplify and
that a l l doctrines are likely to exemplify unless we
are o n our guard against i t . I t is what W i l l i a m
"5
n6 T H E PHILOSOPHY OF R H E T O R I C

James called the Psychologist's Fallacy, the mistak


i n g of a doctrine, w h i c h may be good as far as i t
goes, for the very processes i t is about. As if, to use
Bridges' lines f r o m The Testament of Beauty,
as if the embranglements
of logic wer the prime condition of Being,
the essence of things; and man in the toilsome journey
from conscience of nothing to conscient ignorance
mistook his tottery crutch for the main organ of life.
O u r skill w i t h metaphor, w i t h thought, is one t h i n g
prodigious and inexplicable; our reflective aware
ness of that skill is quite another t h i n g v e r y i n
complete, distorted, fallacious, over-simplifying. Its
business is not to replace practice, or to tell us how
to do what we cannot do already; b u t to protect o u r
natural skill f r o m the interferences of unnecessarily
crude views about i t ; and, above, all, to assist the
i m p a r t i n g of that skill that command of metaphor
from m i n d to m i n d . A n d progress here, i n trans
lating our skill into observation and theory, comes
chiefly f r o m profiting by our mistakes.
Last time I generalized, or stretched, the sense
of the term metaphor almost, you may t h i n k , to
breaking point. I used i t to cover all cases where
a word, i n Johnson's phrase, 'gives us two ideas for
one,' where we compound different uses of the w o r d
into one, and speak of something as though i t were
another. A n d I took i t further still to include, as
metaphoric, those processes i n which we perceive
or t h i n k of or feel about one t h i n g i n terms of
another as when looking at a b u i l d i n g i t seems to
T H E COMMAND OF METAPHOR 117

have a face and to confront us w i t h a peculiar


expression. I want to insist that this sort of
t h i n g is normal i n f u l l perception and that
study of the growth of our perceptions (the ani
mistic w o r l d of the child and so on) shows that i t
mustbeso. - : : , r r -
Let me begin now w i t h the simplest, most familiar
case of verbal metaphor the leg of a table for ex
ample. We call i t dead b u t i t comes to life very
readily. N o w how does i t differ f r o m a plain or
literal use of the word, i n the leg of a horse, say?
T h e obvious difference is that the leg of a table has
only some of the characteristics of the leg of the
horse. A table does not walk w i t h its legs; they
only h o l d i t u p and so on. I n such a case we call
the common characteristics the ground of the meta
phor. Here we can easily find the ground, but very
often we cannot. A metaphor may work admirably
w i t h o u t our being able w i t h any confidence to say
how i t works or what is the ground of the shift.
Consider some of the metaphors of abuse and en
dearment. I f we call some one a p i g or a duck, for
example, i t is little use l o o k i n g f o r some actual re
semblance to a p i g or a duck asthe ground. We do
not call someone a duck to i m p l y that she has a b i l l
and paddles or is good to eat. T h e ground of the
shift is much more recondite. T h e Oxford Dic-
tionary hints at i t by defining a 'duck' i n this use
as 'a charming or delightful object.' A n extremely
simplified account of the ground here w o u l d make
i t s o m e t h i n g like t h i s : that some feeling, of 4ender
n8 T H E PHILOSOPHY OF R H E T O R I C

and amused regard,' say, that i t is possible to have


towards ducks is being felt towards a person.
A very broad division can thus be made between
metaphors which w o r k through some direct resem
blance between the two things, the tenor and vehicle,
and those which w o r k through some common atti
tude which we may (often t h r o u g h accidental and
extraneous reasons) take u p towards them both.
T h e division is not final or irreducible, of course.
That we like them both is, i n one sense, a common
property that two things share, though we may, at
the same time, be w i l l i n g to admit that they are
utterly different. W h e n I like tobacco and logic,
that is no very obvious character that they have
i n common. B u t this division, though i t does
not go very deep, may at a certain level help us
sometimes to avoid one of the worst snares of the
study the assumption that i f we cannot see how a
metaphor works, i t does not work.
Let us go back to leg for a moment. W e notice
that even there the boundary between literal and
metaphoric uses is not quite fixed or constant. T*o
what do we apply i t literally? A horse has legs
literally, so has a spider, b u t how about a chim
panzee ? Has i t two legs or four? A n d how about
a star-fish? Has i t arms or legs or neither? A n d ,
when a man has a wooden leg, is i t a metaphoric or
a literal leg? T h e answer to this last is that i t is
both. I t is literal i n one set of respects, metaphoric
i n another. A w o r d may be simultaneously b o t h
literal and metaphoric, just as i t may simultaneously
T H E COMMAND OF METAPHOR 119

support many different metaphors, may serve to fo


cus i n t o one meaning many different meanings.
T h i s p o i n t is of some importance, since so much
misinterpretation comes f r o m supposing that i f a
word works one way i t cannot simultaneously work
i n another and have simultaneously another
meaning.
Whether, therefore, a w o r d is being used literally
or metaphorically is not always, or indeed as a rule,
an easy matter to settle. W e may provisionally
settle i t by deciding whether, i n the given instance,
the w o r d gives us two ideas or one; whether, i n the
terms I suggested last time, i t presents b o t h a tenor
and a vehicle which co-operate i n an inclusive mean
i n g . * I f we cannot distinguish tenor f r o m vehicle
then we may provisionally take the w o r d to be
l i t e r a l ; i f we can distinguish at least two co-operat
i n g uses, then we have metaphor.
For example, when Hamlet says:
" W h a t should such fellows as I do crawling be
tween earth and heaven?" O r when Swift makes
* '

the Brobdingnagian K i n g say to G u l l i v e r : " T h e


b u l k of your natives appear to me to be the most
pernicious race of little odious v e r m i n that nature
ever suffered to crawl upon the face of the earth,"
are crawling and crawl to be regarded as literal or
metaphoric ?
M y answer is that they are metaphoric. Hamlet
#
This carries assumptions about aspect-selection and thing- or
idea-making, whose examination would be necessary in a further de
velopment.
120 T H E PHILOSOPHY OF R H E T O R I C

or man may crawl literally as babies and big-


game hunters undoubtedly do at times b u t i n b o t h
passages there is an unmistakable reference to other
things that crawl, to the motions of foul insects, to
vermin, and this reference i s t h e vehicle as Hamlet,
or man and his ways, are the tenor. By this test, of
course, most sentences i n free or fluid discourse
t u r n out to be metaphoric. L i t e r a l language is
rare outside the central parts of the sciences. W e
t h i n k i t more frequent than i t is through the influ
ence of that f o r m of the usage doctrine w h i c h
ascribes singlefixedmeanings to words and that is
why I have spent s o m u c h time i n these lectures i n
veighing againstthat doctrine.
Let us consider, now, some of the varying rela
tions between tenor and vehicle. I t is convenient
to begin w i t h the remark, which you w i l l meet w i t h
everywhere, that a metaphor involves a comparison.
What is a comparison ? I t may be several different
things: i t may be just a p u t t i n g together of two
things to let them work together; i t may be a study
of them b o t h to see how they are like and how u n
like one another; or i t may be a process of calling
attention to their likenesses or a method of drawing
attention to certain aspects of the one through the
co-presence of the other. As we mean by compari
son these different things we get different concep
tions of metaphor. I f we mean calling attention
to likenesses,we get a m a i n i 8 t h Century doc-
t r i n e o f metaphor. D r . Johnson, for example,
praises Denham's lines on the Thames because "the
T H E COMMAND OF METAPHOR 121

particulars of resernblance are so perspicaciously col


lected." These are the lines,
O could I flow like thee, and make thy stream
My great exemplar as it is my theme!
Though deep, yet clear; though gentle, yet not dull;
Strong without rage; without o'erflowing, full.
Here the flow of the poet's m i n d , we may say, is
the tenor, and the river the vehicle; and i t is w o r t h
noting, as an exercise i n analysis, that i n the last
two lines there is a repeated alternation of the rela
tive positions of tenor and vehicle and of the direc
t i o n of the shift between them. " T h o u g h deep,
yet clear": the words are literally descriptive of the
vehicle, the r i v e r ; derivatively or metaphorically
descriptive of the m i n d . " T h o u g h gentle yet not
d u l l " : "gentle" certainly is literally descriptive of
the m i n d , the tenor, derivatively of the river, the
other way a b o u t ; b u t " d u l l , " I suppose, goes from
the r i v e r to the m i n d again. " S t r o n g w i t h o u t rage"
goes, for me, unquestionably f r o m m i n d to river,
and " w i t h o u t o'erflowing, f u l l " goes back again from
river, does i t not ? to m i n d . A l l through, of course,
i t is not etymology b u t how we take the words which
settles these questions.
These details of order are not important to no
tice i n themselves though to do so gives practice
i n the peculiar sort of attention which is the method
of the whole study. Still, this alternating movement
i n the shifts may have not a l i t t l e to do w i t h the
rather mysterious power of the couplet, the way i t
exemplifies what i t is describing:
122 T H E PHILOSOPHY OF R H E T O R I C
Though deep yet clear; though gentle, yet not dull;
Strong without rage; without o'erflowing, full.
A n d also i t may have something to do w i t h what
Johnson is r i g h t l y r e m a r k i n g when he says that "the
flow of the last couplet is so smooth and sweet that
the lines have not been overpraised."
" T h e particulars of resemblance (between tenor
a n d v e h i c l e ) are so perspicuously collected," that
( is a typical i 8 t h Century conception of the k i n d of
comparison that metaphor should supply, the proc
ess of p o i n t i n g out likenesses perspicuously collect
ing particulars of resemblance. B u t i t does not
really apply as an account of how these lines work.
T h e more carefully and attentively we go over the
senses and implications of deep, clear, gentle, strong
and full as they apply to a stream and to a m i n d , the
lessshall we find the resemblances between vehicle
and tenor counting and the more w i l l the vehicle,
the river, c o m e t o seem an excuse for saying abcut
the m i n d something which could not be said about
the river. Take deep. Its m a i n implications as re
gards a river are, 'not easily crossed, dangerous, navi
gable, and suitable for swimming, perhaps.' As ap
plied to a m i n d , i t suggests 'mysterious, a l o t going
on, r i c h i n knowledge and power, not easily ac
counted for, acting f r o m serious and important rea
sons.' W h a t the lines say of the m i n d is something
that does not come f r o m the river. B u t the river
is not a mere excuse, or a decoration only, a g i l d i n g
of the m o r a l p i l l . T h e vehicle is still controlling
the mode i n which the tenor forms. T h a t appears
T H E COMMAND OF METAPHOR 1*3

at once i f we try replacing the river w i t h , say, a cup


of tea!

Though deep, yet clear; though gentle, yet not dull;


Strong without rage; without o'erflowing, full.

Comparison, as a stressing of likenesses, is not the


whole mode of this metaphor though i t commonly
is i n 18th Century w r i t i n g w h e r e , too, the tenor
is usually the most important partner i n the meta
phor. T h e opposed conception of comparison as
a mere p u t t i n g together of two things to see what
w i l l happen is a contemporary fashionable aber
ration, w h i c h takes an extreme case as the n o r m .
Here i t is, i n a summary and exaggerated form.
T h i s is A n d r e Breton, the leader of the French
Super-Realists, stating the doctrine very p l a i n l y :
" T o compare two objects, as r e m o t e f r o m one
another i n character as possible, or by any other
method p u t them together i n a sudden and striking
fashion, this remains the highest task to which poetry
can aspire." (Les vases communicants)
" ' T o p u t them together i n a sudden and striking
fashion' ""les mettre en presence d'une maniere
brusque et saisissante" T h a t , as "the highest task
to which poetry can aspire" ! I t is a d o c t r i n e w e l l
w o r t h some examination. L i k e M r . M a x Eastman,
w i t h his insistence ( i n The Literary Mind) that
metaphor works by attempting 'impracticable iden
tifications,' M . Breton sees no need to consider what
should be p u t w i t h what provided they are suffi
ciently remote f r o m one another nor does he
i24 T H E PHILOSOPHY OF R H E T O R I C

distinguish between the very different effects of such


collocations. T h i s is the opposite position f r o m
Johnson's, for whereas Johnson objected to compari
sons being, like Cowley's, 'far fetched,' i t is the dis
tance of the fetching here w h i c h is the m e r i t . M r .
Eastman shares this indifference as to the precise
effect of the encounter of disparates. For h i m the
poet "communicates a k i n d of experience not else
where accessible" and, to do so, M r . Eastman says,
he "must rouse a reaction and yet impede i t , cre
ating a tension i n o u r nervous system sufficient and
r i g h t l y calculated to make us completely aware that
we are l i v i n g something and no matter what."
(The Literary Mind, p. 2 0 5 . ) " N o matter what?"
These last words are heroic certainly. T i e a man
down and approach h i m w i t h a red-hot p o k e r ; you
w i l l arouse a reaction and sufficiently impede i t to
make h i m completely aware, I believe, that he is
l i v i n g something. T h i s same heroism haunts a good
deal of current literary theory and practice not
only i n the Super-Realists' cult of artificial para
noias. I t comes, I t h i n k , f r o m a crude conception
of the mode of action of metaphors, a conception
which is an excessive reaction f r o m the sort of t h i n g
we had last week i n L o r d Kames.
Let us consider more closely what happens i n the
m i n d when we p u t together i n a sudden and strik
i n g fashion two things belonging to very different
orders of experience. T h e most important happen-
j ings i n addition to a general confused reverbera-
1 t i o n and strain are the mind's efforts to connect
THE COMMAND OF METAPHOR 125

them. T h e m i n d i s a connecting organ, i t works 1

only by connecting and i t c a n connect any two


things i n a n indefinitely large n u m b e r o f different
ways. W h i c h of t h e s e i t chooses is settled by refer
ence to some larger whole or aim, and, though we
may not discover its aim, the m i n d is never aimless.
I n all interpretation we are filling i n connections,
and for poetry, of course, o u r freedom to fill i n
the absence of explicitly stated intermediate steps
is a m a i n source of its powers. As M r . Empson w e l l
says ( i n his Seven Types of Ambiguity, p. 3 2 ) ,
"Statements are made as i f t h e y w e r e connected,
and the reader is forced to consider their relations
for himself. T h e reason why these statements
should have been selected is left for h i m to i n v e n t ;
he w i l l invent a variety of reasons and order them
i n his own m i n d . T h i s is the essential fact about
the poetical use of language." T h e reader, I w o u l d
say, w i l l t r y out various connections, and this experi
m e n t a t i o n w i t h the simplest and the most com
plex, the most obvious and the most recondite col
locations alike is the movement w h i c h gives its
meaning to all fluid language.
As the t w o t h i n g s p u t together are more remote,
the tension created is, of course, greater. T h a t ten
sion is the spring of the bow, the source of the
energy of the shot, b u t we ought not to mistake the
strength of the b o w f o r the excellence of the shoot
ing ; or the strain for the aim. A n d bafflement is an
experience of which we soon tire, and rightly. But,
as we know, what seems an impossible connection,
1*6 T H E PHILOSOPHY OF R H E T O R I C

an 'impracticable#identification/ can at once t u r n


i n t o an easy and powerful adjustment i f the r i g h t
h i n t comes f r o m the rest of the discourse. Here
is an instance.
A n incautious recent w r i t e r on the general theory
of language says: " I n England the symbol house
may symbolise a reference to many different kinds
of houses; metaphorically its reference may be so
generalised as to refer to many more other t h i n g s ;
b u t i t can hardly ever have the same reference as,
let us say, bread! 9

T h a t sets us a p r o b l e m ; find an occasion i n which


bread may be metaphorical for house, or house for
bread. I t w o u l d not be hard, I think, to find sev
eral b u t here is a fairly obvious one, f r o m Gerard
Manley Hopkins. F r o m that rather distressing and
unhappy poem, The Drummer Boy's Communion,
when Hopkins is speaking of thewafer as the dwell
i n g of the D i v i n e Presence. T h i s is the l i n e :

Low-latched in leaf-light housel his too huge godhead.

There is no strain, surely, i n speaking of the bread


here as the l i t t l e house, housel.
B u t i t is the rest of the poem that makes the con
nection easy and obvious, which witnesses to a
general t r u t h . T h e m i n d w i l l always t r y to find con
nections and w i l l be guided i n its search by the rest
of the utterance and its occasion.-
I conclude then that these contemporary exploit
ers of the crude 'clash them together no matter
what' view of metaphor are beguiling themselves
T H E COMMAND OF METAPHOR 1*7

w i t h by-products of the process of interpretation and


neglecting the more important cares of critical the
ory. B u t still one point of importance emerges
clearly f r o m examining these exaggerations. W e
must not, w i t h the i 8 t h Century, suppose that the
interactions of tenor and vehicle are to be confined
to their resemblances. There is disparity action
too. W h e n H a m l e t uses the w o r d crawling its force
comes not only f r o m whatever resemblances to ver
m i n i t brings i n b u t at least equally f r o m the dif
ferences that resist and control the influences of
their resemblances. T h e i m p l i c a t i o n there is that
man should not so crawl. Thus, talk about the iden
tification or fusion that a metaphor effects is nearly
always misleading and pernicious. I n general, there
are very few metaphors i n which disparities between
tenor and vehicle are not as much operative as the
similarities. Some similarity w i l l commonly be the
ostensive ground of the shift, b u t the peculiar m o d i - 5

fication of the tenor which the vehicle brings about


is even more the work of their unlikenesses than of
their likenesses.
T h i s has, I believe, very important consequences
for literary practice and theory at innumerable
points. Insufficient analysis here has led not only
to false doctrine and crude reading b u t to attempts
i n w r i t i n g to make words behave i n fashions which
c o n f l i c t w i t h the nature of language as a medium.
T o take the danger of false doctrine first. One
of the most influential of modern critics has been
T . E. H u l m e . H i s death i n the W a r was a very
128 T H E PHILOSOPHY OF R H E T O R I C

heavy loss for many reasons not least, perhaps, be


cause his doctrine of metaphor was left at a half-way
stage f r o m which, I believe, he w o u l d certainly have
developed i t . As i t stands, i n the interpretation i n
which i t has been vigorously infective for the last
nineteen years and especially since his papers on
' M o d e r n A r t ' and on 'Romanticism and Classicism'
were published i n 1924 i n the volume called Specu-
lations i t seems to me most deceiving.
I t says (p. 137) " P l a i n speech is essentially inaccu
rate. I t is only by new metaphors . . . that i t can
be made precise." T h i s you w i l l see is only Shel
ley's p o i n t again, and we can accept i t , w i t h a de
m u r r e r as to some of the implications of 'new' here
a demurrer that H u l m e himself hints on an earlier
page when he says, " 'Works of art aren't eggs,' " and
so need not be fresh or new laid. B u t he added
various points about the precision that he supposed
m e t a p h o r t o aim at, and i t is these that give occasion
for mistakes. " T h e great a i m , " he says, "is accu
rate, precise and definite description." Poetry,
fluid discourse, as opposed to prose, "is not a lan
guage of counters, b u t , " he holds, "a visual concrete
one. I t is a compromise for a language of i n t u i t i o n
which w o u l d hand over sensations bodily. I t always
endeavours to arrest you, and make you continu
ously see a physical thing, to prevent you g l i d i n g
through an abstract process."
I have three quarrels w i t h this account. First
w i t h that always. O n l y remember Shakespeare and
you w i l l not say t h a t t h e language of poetry always
T H E COMMAND OF METAPHOR 129

does anything of this sort. M y second quarrel is


w i t h the words visual and see: "make you continu
ously see a physical t h i n g and prevent you g l i d i n g
through an abstract process." T h a t is patently
false.
If thou didst ever hold me in thy heart
Absent thee from felicity awhile
And in this harsh world draw thy breath in pain
To tell my story.

You need see nothing while reading that, and the


words certainly do not work by m a k i n g y o u see
anything. Besides, you already have the actors to
look at. M y t h i r d quarrel is w i t h this fear of the
abstract. T h e language of the greatest poetry is
frequently abstract i n the extreme and its aim is pre
cisely to send us " g l i d i n g through an abstract
process."
This she? N 0 , this is Diomed's Cressida.
If beauty have a soul, this is not she,
If souls guide vows, if vows be sanctimony,
If sanctimony be the gods' delight,
If there be rule in unity itself,
This is not she.

We are not asked by Shakespeare here to perceive


beauty, b u t to understand i t through a metaphoric
argument as the 'rule i n u n i t y itself and to under
stand its place i n the soul's growth.

W h a t can have happened to make so shrewd and


acute a w r i t e r as H u l m e blunder i n this gross fash
ion? I have two explanations, w h i c h combine.
i30 T H E PHILOSOPHY OF R H E T O R I C

T h e first is that he is t r i c k i n g himself w i t h the w o r d


see i n t o supposing that he means i t literally when
his doctrine w o u l d only be sanctioned i f he were
using i t metaphorically. Obviously if, i n an argu
ment, we say " I see your p o i n t ! " we are using see
metaphorically. So when H u l m e wrote see and
visual here, the words are to be taken metaphori
cally too or the doctrine must be condemned at once.
W h a t discourse 'always endeavours to do is to make
us apprehend, understand, gain a realizing sense of,
take i n , whatever i t is that is being meant w h i c h
is not necessarily any physical t h i n g . B u t i f we say
"a realizing sense," we must remember that this is
not any 'sense' necessarily, such as sense-perception
gives, b u t may be a feeling or a thought. W h a t is
essential is that we should reallytake i n and become
fully aware of whatever i t is.
T h i s blunder w i t h the vtorasee may seem too
crude to be likely. B u t the patient t o i l of scores of
teachers is going every day, i n courses about the
appreciation of poetry, i n t o the effort to make chil
dren (and adults) visualize where visualization is a
mere distraction and of no service. A n d l i t t l e books
appear every few months encouraging just this gross
misconception of language. For words cannot, and
should not attempt to "hand over sensations bod
i l y " ; they have very much more important w o r k
to do. S o f a r f r o m verbal language being a "com
promise for a language of i n t u i t i o n " - a t h i n , b u t
better-than-nothing, substitute for real experience,
language, w e l l used, is a completion and does what
T H E COMMAND OF METAPHOR 131

the intuitions of sensation by themselves cannot do.


Words are the meeting points at which regions of
experience w h i c h can never combine i n sensation
or i n t u i t i o n , come together. T h e y are the occasion
and the means of that growth which is the mind's
endless endeavour to order itself. T h a t is why we
have language. I t is no mere signalling system. I t
is the instrument of a l l our distinctively h u m a n
development, of everything i n which we go beyond
the other animals.
Thus, to present language as w o r k i n g only
through the sensations i t reinstates, is to t u r n the
whole process upside down. I t overlooks what is
important i n Mallarme's dictum that the poet does
not write w i t h thoughts (or w i t h ideas or sensations
or beliefs or desires or feelings, we may add) b u t w i t h
words. " A r e not words," so Coleridge asked, "parts
and germinations of the plant? A n d what is the
law of their growth? I n something of this sort,"
he wrote, " I w o u l d endeavour to destroy the old
antithesis of Words and T h i n g s : elevating, as i t
were, Words into Things and l i v i n g things too."
W e must do so i f we are to study metaphor profit
ably. H u l m e and the school teachers are forgetting
everything that matters most about language i n
treating i t as just a stimulus to visualization. T h e y
t h i n k the image fills i n the meaning of the w o r d ;
i t is rather the other way about and i t is the w o r d
which brings i n the meaning which the image and
its original perception lack.
T h a t is one part, I think, of the explanation of
i$2 T H E P H I L O S O P H Y O F R H E T O R I C

these disorders of thought the mistaking of see and


perceive i n the literal sense instead of a wide and
open metaphoric sense. B u t the other part of the
explanation goes deeper: i t is t h e m i s t a k i n g of what
I have been calling the tenor-vehicle antithesis for
that between the metaphor (the double u n i t i n
cluding tenor and vehicle) and its meaning. These
- two antitheses are easy to confuse, indeed, i t is hard
to keep them steadily distinct especially when
metaphor (and its synonyms), as I illustrated last%

time,sometimes means Vehicle,' sometimes means


Vehicle and tenor together.' N o t h i n g b u t habitua
t i o n makes this shift manageable and keeps i t f r o m
deceiving us. I t h i n k i t deceived H u l m e here and
I k n o w i t deceives others. W h e n he says, " T h e
great a i m is accurate, precise and definite descrip
t i o n " we can agree, i f that is saying no more than
"the words somehow must make us f u l l y and r i g h t l y
aware of whatever i t is, the language must really
utter its meaning." T h a t is, the metaphor (the
whole thing, tenor and vehicle together) should
mean what i t should. B u t H u l m e turns his remark
i n t o something about a supposedly needful accu
racy of correspondence between vehicle and tenor,
and so into something which is false. " P l a i n speech
is essentially inaccurate. I t is only by . . . meta
phors . . . that i t can be made precise. W h e n the
analogy has not enough connection w i t h the t h i n g
described to b e q u i t e parallel w i t h i t , when i t over
lays the t h i n g i t describes and there is a certain
excess" i t is inferior. " B u t where the analogy is
T H E COMMAND OF METAPHOR 133

every b i t of i t necessary for accurate description . . .


I f i t is sincere, i n the accurate sense, when the whole
of the analogy is necessary to get o u t the exact curve
of the feeling or t h i n g you want to express there
you seem to me (he says) to have the highest verse."
I n part of this, H u l m e is t h i n k i n g of the whole
metaphor and its m e a n i n g ; i n other parts he is
t h i n k i n g of the vehicle and tenor. Something
which is obvious and true of the whole metaphor
and its meaning thus lends an illusory plausibility
to a fabe view of the correspondence of vehicle to
tenor. H u l m e seems n o t to be distinguishing these
two couples and i t is as fatal to confuse them as i t
w o u l d be i n chemistry to mistake the order of com
plexity of a molecule and an electron, or i n algebra,
to ignore the brackets. H i s confidence i n a truism
that speech should mean what i t should mean
makes h i m (as I read his pages) certain that vehicle
must correspond to tenor the whole of the analogy
be necessary to get o u t the exact curve and that,
i n the sense i n w h i c h I read h i m , is not a truism,
b u t an easily demonstrable error, a misdescription
of a l l o u r current practice.
For one thing, there is no whole to any analogy,
we use as m u c h of i t as we need; and, i f we tact
lessly take any analogy too far, we break i t down.
T h e r e are n o s u c h limits to the r e l a t i o n s o f tenor
and vehicle as this account puts. T h e result of the
doctrine may be seen i n those anxious, over-careful
attempts to copy perceptions and feelings in words,
to " h a n d over sensations b o d i l y , " of which modern
iM T H E PHILOSOPHY OF R H E T O R I C

prose at its most distinguished too often consists.


Words are not a m e d i u m i n which to copy life.
T h e i r true work is to restore life itself to order.
T h e error of mistaking l h e tenor-vehicle relation
for the relation between tenor plus vehicle together
and what they mean, has consequences which go far
beyond what we are apt to regard (on a l i m i t e d
view) as literary matters. T h e y enter into the ways
we envisage all our most important problems. For
example, i n t o the question of belief. M u s t we be
lieve what an utterance says i f we are to understand
i t fully? Does the D i v i n e Comedy, or the Bible
tell us something which we must accept as true i f
we are to read i t aright ? These are questions that
we cannot possibly answer satisfactorily unless we
are clear as to the ways i n which metaphoric utter
ances may say or tell us something. M r . E l i o t re
marks somewhere of the D i v i n e Comedy that the
whole poem is one vast metaphor. I t is. A n d , i f
so, what is i t that we m i g h t believe i n i t ? Is i t the
tenor or the vehicle or their j o i n t presentation; or
is i t 'that tenor and vehicle are thus and thus re
lated there'? O r is the b e l i e f r e q u i r e d no more
than a readiness to feel and w i l l and live, i n cer
tain respects, i n accordance w i t h the resultant mean
i n g i n so far as we apprehend t h a t m e a n i n g o r
rather i n so far as that meaning apprehends, grasps,
takes control of, us ? W e are accustomed to distin
guish between taking an utterance literally and tak
i n g i t metaphorically or anagogically, but, at the
simplest, there are at least four possible modes of
THE COMMAND OF METAPHOR 135

interpretation to be considered, not two. A n d the


kinds of believing that w i l l be appropriate w i l l as
a r u l e be different. W e can extract the tenor and
believe that as a statement; or extract the vehicle;
or, taking tenor and vehicle together, contemplate
for acceptance or rejection some statement about
their relations, or we can accept or refuse the direc
t i o n which together they w o u l d give to our living.
W e need not go to the Alexandrian schools of early
Christian interpretation, or to the similar exegetical
developments of other religions, to find instances
to show how immense the consequences for belief
of these choices may be. T h e varying possibilities
of understanding of any metaphoric utterance w i l l
show them.
A 'command of metaphor'a command of the
interpretation of metaphors can go deeper still
into the control of the w o r l d that we make f o r o u r -
selves to live i n . T h e psycho-analysts have shown
us w i t h their discussions of 'transferenceanother
name for m e t a p h o r h o w constantly modes of re
garding, of loving, of acting, that have developed
w i t h one set of things or people, are shifted to an
other. T h e y have shown us chiefly the pathology
of these transferences, cases where the vehicle the
borrowed attitude, the parental fixation, say tyran
nizes over the new situation, the tenor, and behavior
is inappropriate. T h e v i c t i m is unable tosee the
new person except i n terms of the old passion and
its accidents. H e reads the situation only i n terms
of the figure, the archetypal image, the vehicle. B u t
136 T H E PHILOSOPHY OF R H E T O R I C

i n h e a l t h y growth, tenor and vehicle t h e n e w


h u m a n relationship and the f a m i l y constellation
co-operate freely; and the resultant behavior de
rives i n due measure f r o m b o t h . T h u s i n happy
l i v i n g the same patterns are exemplified and the
same risks of error are avoided as i n tactful and
discerning reading. T h e general f o r m of the inter
pretative process is the same, w i t h a small-scale i n
stancethe r i g h t understanding of a figure of
- speech or w i t h a large scale instance the conduct
of a friendship.
B u t the literary instance is easier to discuss and
more accessible to investigation. I t is an o l d dream
that i n time psychology m i g h t be able to tell us so
m u c h about o u r minds that we w o u l d at last be
come able to discover w i t h some certainty what we
mean by o u r words and h o w we mean i t . A n oppo
site or complementary dream is that w i t h enough
improvement i n Rhetoric we may i n t i m e learn so
much about words that they w i l l tell us h o w o u r
minds work. I t seems modest and reasonable to
combine these dreams and hope that a patient per
sistence w i t h the problems of Rhetoric may, w h i l e
exposing the causes and modes of the misinterpreta
t i o n of words, also t h r o w l i g h t u p o n and suggest a
remedial discipline for deeper and more grievous
disorders; that, as the small and local errors i n
our everyday misunderstandings w i t h language are
models i n m i n i a t u r e of the greater errors w h i c h
disturb the development of o u r personalities, their
study m a y a l s o s h o w us more about how these large
T H E COMMAND OF METAPHOR 137

scale disasters may be avoided. T h a t at least was


Plato's hope, as i t was Spinoza's belief that there is
b u t one end for the sciences. "Above a l l things, a
method must be thought out of healing the under
standing and p u r i f y i n g i t at the beginning, that i t
may w i t h the greatest success understand things cor
rectly." These Lectures, which began by claiming
that the study of Rhetoric must, i n a certain sense,
be philosophical, may end w i t h a passage, f r o m the
Timaeus, where Plato is speaking of this hope i n
myth.
" T h e circuits of the years, passing before our eyes,
have discovered unto us number and given to us a
notion of t i m e ; and set us seeking to know the na
ture of the A l l ; whence we have gotten us Philos
ophy than w h i c h no greater good hath come, nor
ever shall come, as gift f r o m Gods to m o r t a l k i n d . "
T h a t , i f we wrest the words, may seem a singularly
bitter charge to b r i n g against the Gods. B u t Plato
meant otherwise. H e goes on, "Concerning Sound
and Hearing let the same t h i n g be said that they
also have been bestowed by the Gods to the same
end as sight. For to this end hath Speech been or
dained, and maketh thereto the largest contribu
t i o n ; and moreover a l l that part of Music hath
been given us for the sake of Harmony, and Har
mony, having her courses k i n unto the revolutions
i n our Soul, hath been given by the Muses to be a
helper to the man who, w i t h understanding, shall
use their art, not for the getting of unreasonable
pleasure w h i c h is commonly esteemed the use of
138 T H E PHILOSOPHY OF R H E T O R I C

Music b u t for the ordering of the circuit of our


Soul which hath fallen out of Harmony, and the
b r i n g i n g thereof i n t o concord w i t h itself. . . N o w
unto the Divine Part i n us the motions that are k i n
are the Thoughts and Circuits of the A l l . These
must every man follow, that he may regulate the
Revolutions i n his Head which were disturbedwhen
the Soul was b o r n i n the Flesh and by thoroughly
learning the Harmonies and Circuits of the A l l may
make that which understandeth like unto that which
is understood, even as i t was i n the b e g i n n i n g ; and,
having made i t like, may attain u n t o the perfection
of that Best L i f e which is offered u n t o men by the
Gods, for the present time and for the time here
after."