CONCEPTS OF LITERATURE

STYLE AND
STYLISTICS
Graham Hough
STYLE AND STYLISTICS

Linguistics and literary criticism have

in part the same subject matter, but
they have different aims and are widely
separated in method. The principal
bridge between them is stylistics —
the study of style, sometimes scientific,
sometimes largely intuitive. This book
gives a short survey of stylistics from
the literary point of view, and tries to
answer the question of how much
stylistics has to contribute to the
understanding of literature. It brings
together Continental European work
on stylistics and Anglo-American crit¬
ical writing which has a similar purpose
though usually under a different name.
In calling the attention of the student
of literature to trains of thought with
which he is not generally familiar.
Professor Hough provides important
new critical insights.
Style and Stylistics
CONCEPTS OF LITERATURE

GENERAL EDITOR: WILLIAM RIGHTER

Department of English
University of Warwick
Style and Stylistics

by Graham Hough
Professor of English Literature
University of Cambridge

LONDON

ROUTLEDGE & KEGAN PAUL

NEW YORK : HUMANITIES PRESS
First published in 1969
by Routledge & Kegan Paul Ltd
Broadway House, 68-y4 Carter Lane
London E.C.4
Printed in Great Britain
by Willmer Brothers Limited
© Graham Hough 1969
No part of this book may be reproduced
in any form without permission from
the publisher, except for the quotation
of brief passages in criticism
SBN yioo 6416 o (C)
SBN yioo 6qiy 9 (P)
General Editor’s Introduction

The study of literature has normally centred on the

consideration of work, author, or historical period. But
increasingly there is a demand for a more analytic ap¬
proach, for investigation and explanation of literary con¬
cepts of crucial ideas and issues—topics which are of
general importance to the critical consideration of par¬
ticular works. This series undertakes to provide a clear
description and critical evaluation of such important ideas
as ‘symbolism’, ‘realism’, ‘style’ and other terms used in
literary discussion. It also undertakes to define the rela¬
tionship of literature to other intellectual disciplines:
anthropology, philosophy, psychology, etc., for it is in
connection with such related fields that much important
recent critical work has been done. Concepts of Literature
will both account for the methodology of literary study,
and will define its dimensions by reference to the many
activities that throw light upon it. Individual works will
describe the fundamental outlines of particular problems
and explore the frontiers that they suggest. The series as a
whole will provide a survey of recent literary thought.
It is often claimed that stylistic study using linguistic
methods may provide a new intellectual discipline for
literary criticism. Professor Hough examines the claims
VI STYLE AND STYLISTICS

of stylistics, in a variety of its forms, to determine both

its benefits and its limitations. His is consciously a literary
approach, continually asking, ‘How does this method
advance our understanding of particular literary works, or
in a wider context, of the phenomenon of literature
itself?’ His scepticism about quantitative analysis is not
due to any hostility towards scientific procedures in their
own right, but rather to his rigorous questioning of the
criteria of what is relevant in any given literary case. The
resulting appraisal has important bearings on the scope
and character of the critic’s task.

WILLIAM RIGHTER
Contents

Preface ix

1 The concept of style and the origins of style-study i

Older Concepts of Style i
The Modern Concept of Style g
The Beginnings of Modern Style-study 12

2 Linguistic style-study 2i
Up to Saussure 21
Bally and His Successors 25

3 Literary stylistics: methods and problems 31

S tylistics and Li ter ary Art 31
Special Expressive Devices 33
Individual Style 38
Period Style 48
History of Style 52
Statistical Methods 53

4 Some practitioners 59
Leo Spitzer 59
Erich Auerbach 68
Damaso Alonso 73
I. A. Richards 80
William Empson 90
John Holloway 95
Stephen Ullmann 98
Donald Davie 99

5 Conclusion: limits and possibilities 103

Select Bibliography in
I would maintain that to formulate observation by means
of words is not to cause the artistic beauty to evaporate in
vain intellectualities; rather, it makes for a widening and
deepening of the aesthetic taste. It is only a frivolous love
that cannot survive intellectual definition; great love
prospers with understanding.

LEO SPITZER
Preface

The object of this essay is to give a short account of the

modern study of literary style. It is necessarily selective
and incomplete, but I have tried to indicate the main
directions that such work has taken, and the directions
it might take in the future. Style-study has often grown
from linguistics, sometimes from other starting-points.
But whatever its origin, stylistics is inevitably a study of
language. The only matter for dispute is how literary
language should be studied. Linguistics is by now a for¬
midable and autonomous discipline, and its relation with
literary studies has not been easy. Many of its concerns
are irrelevant to literature, and some of its methods are
disliked by most literary students. Yet in the end it cannot
be irrelevant. The study of language and the study of
literature obviously have a common frontier, and stylis¬
tics is the border area.
This book is written from the literary point of view.
I hope what I have said will not seem positively wrong
to professional linguists; but it is not my aim to satisfy
their demands or their criteria of relevance. I have not
tried to inquire how linguistics could revolutionize the
study of literature, but how much it can contribute to the
X STYLE AND STYLISTICS

study of literature as that is ordinarily understood. A

subsidiary aim has been to bring together Continental
work in stylistics and English work which, though it has
never been called ‘stylistics’, has actually the same object.
There is something to be gained by seeing them in the
same perspective.
I should like to thank Peter Seuren, of Darwin College
and the Cambridge Department of Linguistics, for kindly
reading my typescript and clarifying some linguistic
matters that were very dimly present to my consciousness.

Darwin College G.H.

Cambridge
 1
The concept of style and
the origins of style-study

Older Concepts of Style

It is a paradox that the term ‘style’ has tended to dis¬

appear from the main stream of modern criticism, while
a quasi-independent study of ‘stylistics’ has simultaneously
made its appearance. If we look into the causes of this
we shall go far towards defining the rationale of the
modern study of literary style.
The concept of style is an old one; it goes back to the
very beginnings of literary thought in Europe. It appears
in connection with rhetoric rather than poetic, and there
seems to be no special reason for this, except that style
is regarded as part of the technique of persuasion and
therefore discussed largely under the head of oratory.
Ancient rhetoric distinguished between ceremonial, politi¬
cal and forensic oratory, and each has its own appropri¬
ate occasion and appropriate repertory of devices. If you
wish to produce this particular effect these are the means
to bring it about; the proper vocabulary, type of syntax
and figures of speech can be prescribed for the purpose in
hand (Aristotle, Rhetoric, Bk. Ill; Quintilian, Institutes of
Oratory, Bk. VIII). The tone of this ancient rhetoric is
largely prescriptive—the giving of instructions for appro¬
priate and effective composition. Even a writer like
2 STYLE AND STYLISTICS

Longinus, who is much concerned with the moral and

spiritual sources of ‘the sublime’, still goes on to detail
the rhetorical figures by which it can be achieved. Ancient
rhetoric in its later phases tended to enlarge its discussion
to historians and other prose writers. In the Middle Ages
and the Renaissance this immense body of rhetorical pre¬
cept was largely incorporated into poetic, where it had a
deep influence not only on critical ideas, but, as recent
studies have shown, on the composition of poetry itself.
The tradition carries on a lingering existence even into
the eighteenth century.
But for us all this is a vanished history. Prescriptive
criticism has not been a central literary activity for the
last 300 years. In a post-Romantic age it survives only
in odd corners—schools of journalism, classes in ‘creative
writing’. Modern literary study does not presume to dic¬
tate to poets; it does not offer instructions towards the
forming of a style, it examines styles that are already
formed. It is parallel in this respect to linguistic study,
which no longer lays down rules for correct grammar,
but studies the rules that are actually adhered to by
particular cultural groups. The aim is not to give laws for
human utterance, but to understand the utterances that
actually occur. We can conceive a time when linguistics
and criticism might resume their legislative role; but ex¬
cept in countries where literature is decisively subor¬
dinated to political necessity this is remote from any
currently active way of thinking, and we shall not con¬
sider it further.
However, there are other legacies from the old rhetori¬
cal concept of style that cannot be dismissed so easily.
It was fundamental in traditional rhetoric and criticism
to make a separation between matter and manner, what
is said and the way of saying it. Such things are often
spoken of in metaphor, perhaps necessarily so; and here
 THE CONCEPT OF STYLE 3

the commonest metaphor is to speak of language as the

dress of thought. Thought is imagined as existing in
some pre-verbal form, and it is then clothed in language.
We can illustrate this from a passage in Dry den’s Preface
to Annus Mirabilis:

So then the first happiness of the poet’s imagination is

properly invention, or finding of the thought; the second
is fancy, or the variation, deriving or moulding of that
thought, as the judgment represents it proper to the subject;
the third is elocution, or the art of clothing or adorning
that thought so found and varied in apt, significant and
sounding words.

On this theory it is easy to see what style is. Language

is the dress of thought, and style (often, following
Quintilian, referred to as ‘elocution’) is the particular cut
and fashion of the dress.
This way of thinking has a number of consequences.
The cut or fashion can be looked at from different points
of view. Dryden sees it as mainly dictated by the subject;
the thought must be moulded in a manner ‘proper to
the subject’; the words must be ‘apt’ to the subject. This
is in conformity with the general neo-classic theory of
literary kinds. Each genre has its own appropriate style;
the style of a tragedy is not to be the same as that of a
pastoral because they have different subject-matters; and
this is nothing to do with the private tastes of the author,
but part of the nature of things. At a rather later date,
with the advent of expressive theories of literature, style
is seen as largely dictated by the nature of the author
himself. It is the expression of his personality. Le style
c’est l’homme meme, as we always say in this context,
perverting Buffon to our own purpose. By an extension of
this approach we can go on to talk about the style of
a period or a literary school. But from then on it begins
4 STYLE AND STYLISTICS

to be doubtful whether we can stay within the limits

of the old metaphor—language as the dress of thought.
Is the style of the Romantic poets different from that
of the school of Pope because they were saying the same
things in different ways, or because they were saying
different things? Probably the latter; and the more we
reflect on it the more doubtful it becomes how far we
can talk about different ways of saying; is not each
different way of saying in fact the saying of a different
thing?
The process of reflecting on this subject goes on inter¬
mittently through the eighteenth century. Gray could
write to his friend Mason in 1762, of a passage in one
of his odes: ‘It is flat, it is prose.... If the sentiment
must stand, twirl it a little into an apophthegm, stick a
flower in it, gild it with a costly expression, let it strike
the fancy, the ear or the heart, and I am satisfied.’
This is apt to make a post-Romantic generation shudder.
If Mason’s sentiment was flat, we feel, no amount of
gilt or costly expression is going to improve it. By the
time we reach the criticism of our own day we find
that this whole distinction between matter and manner
has been decisively rejected. To reject it has indeed be¬
come one of the principal dogmas of current critical
thought. The work of literary art is seen as an organic
unity, in which matter and manner, thought and ex¬
pression are indissolubly one; and what began perhaps
as an aesthetic doctrine is equally prevalent in consider¬
ing non-literary utterance. Bloomfield, in an article on
‘Linguistic Aspects of Science’, says: ‘It is a well-tried
hypothesis of linguistics that formally different utter¬
ances always differ in meaning.’ Some recent developments
in generative grammar make this look at least doubtful,
and the status of synonymy is still open to discussion.
But Bloomfield’s point of view is still general. On this
 THE CONCEPT OF STYLE 5
hypothesis we cannot talk about different ways of ex¬
pressing the same thought, but only of different
thoughts. What then has become of style? It seems to
have disappeared. The case has been put very clearly by
Richard M. Ohmann (Style in Prose Fiction, 2):

For if style does not have to do with ways of saying some¬

thing, just as style in tennis has to do with ways of hitting
a ball, is there anything at all which is worth naming
‘style’ ? ... The critic can talk about what the writer says,
but talk about the style he cannot, for his neat identity—
one thought, one form—allows no margin for individual
variation, which is what we ordinarily mean by style. Style,
then, becomes a useless hypothetical construct.

And it is for this reason that the word ‘style’ makes

very little appearance in the main stream of modern
criticism.

The Modern Concept of Style

Whatever has happened to the word, however, the con¬

cept of style cannot in practice be simply evaporated;
for the kind of considerations that used to shelter under
its name are still critically active. In England and America
there is a huge complex of critical and educational prac¬
tice that, without using the name, relies largely on stylis¬
tic analysis. The experiments in ‘practical criticism’ con¬
ducted by I. A. Richards in the twenties; what is, or was,
called the New Criticism in America; the widespread
techniques of ‘explication’ or ‘close reading’ of poetry—
these are all cases in point. The idea that the nature of a
whole work can be deduced from the qualities exhibited
in a short passage is still widely current; and this is a
stylistic dogma. Indeed, all that body of modern criticism
which prides itself on its close contact with the verbal
6 STYLE AND STYLISTICS

texture of literature is a kind of style-study. But the

modern critic does not talk about style any more than
he talks about beauty. Because the word is out of
fashion and the concept now ill-defined, it has been felt
necessary to invent new terminologies, and to be ex¬
cessively cautious about relapsing into old doctrines now
regarded as heresies. If we could continue to talk about
style—without bringing in undesired and outmoded con¬
notations—it would be an obvious simplification. It would
also have the advantage of enlisting at least a little aid
from the linguistic disciplines to a kind of criticism that
has often had no discipline at all.
In general the linguists seem to have been little em¬
barrassed by the difficulties we have been considering. In
spite of the dictum of Bloomfield quoted above, many
students of linguistics who have concerned themselves
with style are quite content to talk about different ways
of saying the same thing. Charles Bally, one of the found¬
ing fathers of modern stylistics, defined it as the study of
the ‘affective’ elements in language—these affective ele¬
ments being conceived as optional additions to an already
determinate meaning. More recently Hockett’s Course in
Modern Linguistics asserts that ‘two utterances in the
same language which convey approximately the same
information, but which are different in their linguistic
structure, can be said to differ in style: He came too soon
and He arrived prematurely’. Stephen Ullmann quotes a
sentence of Proust, then offers a rearranged version of
it and says, ‘Both sentences mean the same thing.’ He
goes on to analyse the difference between them as a
matter of effectiveness—effectiveness in expressing a given
meaning—thus following precisely the definition of style
he has quoted earlier from Stendhal: ‘ajouter d une
pensee donnee toutes les circonstances propres d pro¬
duce tout l’effet que doit produire cette pensee’.
 THE CONCEPT OF STYLE 7
The attitude of professional linguistics to this problem
is perhaps in process of modification. Late developments
of Chomsky’s generative grammar have led to the hypo¬
thesis (it is yet only a hypothesis) that the deep structure
of sentences may be the universal semantic basis of all
languages, that they assume different grammatical forms
in different languages, and that a single semantic complex
in an individual language may indeed assume different
but synonymous grammatical forms. The differences be¬
tween synonymous sentences may then be called stylistic;
in fact, a return to the old view of language as the dress
of thought. Both the state of the case and the limitations
of my knowledge forbid further discussion of the question
here.
So I shall not meddle any further with the philosophi¬
cal or linguistic aspects of the ‘one thought, one form’
doctrine—not from lack of interest, but from lack of
competence. I would claim too that the literary critic
has the right to borrow Occam’s razor, and may legitim¬
ately shear off a good many entities that are not necessary
to his purpose. Among them are all discussions about
forms of propositions, their relation to the syntactical
forms of sentences, and most of what is said about the
meaning of meaning. Such considerations occur at a level
of abstraction where literary criticism has difficulty in
following, where it has no contribution to make and no
competence to decide. In these matters the critic may be
a probabilist, and use for his purposes any doctrine
decently attested by a reputable authority. The value of
his conclusions is to be judged by their success in inter¬
preting literature, not by the nature of the non-literary
tools he has used on the way.
I think the concept of style can be rescued in three
ways, none of them obviously disreputable.
(i) The critic can rest on ordinary language and re-
B
8 STYLE AND STYLISTICS

ceived opinion, ‘the common sense of readers uncorrupted

by literary prejudice’. Such readers obstinately persevere
in distinguishing between matter and manner, the thing
said and the way of saying it. The end of literature is to
come home to men’s business and bosoms, and it can be
argued that it is best discussed in the terms in which
men’s business is normally conducted, to which their
bosoms normally return an echo. There is no need to
prove that there ‘is’ such a thing as style; it is enough
that it is a convenient and natural term to use, and that
in practice everybody knows what is being talked about.
(2) The critic can deny the doctrine that formally
different utterances always differ in meaning, as I. A.
Richards goes near to doing in Interpretation in Teaching.
This may need some ingenious casuistry, but it is possible
to establish the position sufficiently for literary purposes.
(3) The critic may accept this doctrine, and agree that
difference of form is always difference of meaning. But
he can still deny that the concept of style has disappeared.
It has not disappeared; it has become subsumed in mean¬
ing. Style is a part of meaning, but a part which can
properly and reasonably be discussed on its own.
For myself I should prefer the third of these possi¬
bilities—that style is an aspect of meaning; but I should
expect to find literary critics employing all three, if not
indiscriminately at least on different occasions and for
different purposes. I shall not attempt to decide between
these three positions—any one of them gives the critic
all the room he needs to work in. What is now necessary
is to show in practical and literary terms what aspects of
the literary work it is intended to discuss under the head
of style.
Whatever view we may take of its nature, it is clear
that in talking about style we are talking about choice—
choice between the varied lexical and syntactic resources
 THE CONCEPT OF STYLE 9
of a particular language. And this is a secondary choice,
a choice of means. In discussing whatever it is we mean
by style we assume that the primary choice, the choice
of subject-matter in the large sense, has already been
made. The decision to write about the Trojan War, or
whale-fishing, or a country childhood is not a stylistic one.
We can talk about the secondary choice, the stylistic one,
in various ways. We can say that it is the choice of the
best verbal means to express a pre-determined subject-
matter; or if we decline to do this we can talk about
choice between the various meanings or shades of mean¬
ing that cluster round a given subject. (In practice it will
be found to make surprisingly little difference which way
we set about it.) We can regard the choice as conditioned
by the subject-matter and the occasion, or as conditioned,
perhaps unconsciously, by the character and tempera¬
ment of the author. But whatever point of view we adopt,
it will be the verbal ordonnance that we discuss, not the
outlines of the myth, the facts of the case, the ideological
or biographical substructure.
Let us begin by taking a simple case—virtually a non¬
literary one—the case of a scientific paper reporting the
results of an experiment. It has several points of interest
for our purpose. In the first place it could surely be main¬
tained with some plausibility that there is here a pre-
linguistic matter to be clothed in verbal dress. It would be
possible not to write the paper at all, but simply to
summon those with whom it was desired to communicate
to witness a wordless repetition of the experiment. Or
the result could be expressed entirely in mathematical
formulae. If ever it is reasonable to talk about language
as the dress of some pre-verbal complex of thought, surely
it is here. But the strictly stylistic problem is limited;
the range of choice is extremely narrow. Feeling is ex¬
cluded in a scientific paper; tone (attitude towards the
IO STYLE AND STYLISTICS

reader) is neutral; verbal play is superfluous or taboo.

Yet even within these limits the stylistic choice still
exists. It is possible to announce the subject of inquiry,
to describe the experimental means used to investigate
it, to state the results, in a variety of ways—economically
or diffusely, clearly or obscurely. Above all it is possible
to write the paper in such a way that curiosity is aroused,
the resulting tension sustained for the appropriate time,
the curiosity satisfied at the logically and psychologically
appropriate points. And these things are all parts of what
we will call style. It is possible that so much of style
can occur even in a non-verbal medium—as when a
mathematical paper is commended for its ‘elegance’.
In a properly literary context the matter is always more
complex and the possibilities of choice are far greater.
The feeling of the writer towards his subject, his attitude
towards the reader, both become significant. And here
it is surely less appropriate to talk of a predetermined
subject, which the writer can choose to express in one
of a number of ways. The feeling and the tone are parts
of what is to be expressed. They are parts of the meaning,
not supererogatory embellishments or means of ingratia¬
tion. But however we choose to look at them, they are
there to be discussed, and the verbal means by which
such attitudes are established are open to observation.
The choice of this word rather than that, of this kind of
syntactical construction rather than another, is a visible
fact, whose nature and effects can be examined. The
initial response to a work of literature is general; we re¬
spond to the whole complex, without being aware of
what goes to make it up. Most readers stop there, and
feel no impulse to go farther. The minority whose in¬
terests lead them to closer inquiry may take a variety
of courses. They may be led back from the work to its
genesis in the author’s experience; they may examine
 THE CONCEPT OF STYLE 11

the philosophic or religious ideas that it embodies; they

may concern themselves about its moral or social effects.
But if their interest is in the maker’s art itself, rather
than in its causes or its effects, they will inquire into the
particular verbal means by which the total form has been
achieved. In that case the inquiry will be a stylistic one.
If its results bring about a new understanding of the
total literary form, the stylistic inquiry will have had real
critical results.
The organic unity of a work of literature is not some¬
thing ready-made; it is not an entire and perfect chrysolite
found lying about in nature; it is something achieved.
This organic whole may be arrived at in a variety of
ways. The lyric poet sometimes has a rhythm in his
head before he knows what words will ever be found
to fit it; he may conceive a line or a phrase before he
knows what poem it is going to belong to; only on rare
occasions does he experience a whole poem as given,
dictated at once in its final shape. The expository writer
begins with an argument, perhaps vaguely apprehended
as a general form, but refined and knit together only in
the process of working out the appropriate organization,
syntax and vocabulary. The novelist can be aware of
fictional characters and situations, just as he can be
aware of characters and situations in life, without any
thought of their linguistic presentation. Most writing in¬
volves a process of revision, conducted either on paper
or in the mind before anything is written down. There is
some evidence that different writers look on this revising
process in very different lights; some see it as the pro¬
gressively more accurate embodiment of a preconceived
meaning; some see it as a continual change and modifica¬
tion of meaning itself. In either case it is best for the
critic to look at the matter prospectively. The work of
literature is a project; when it is complete the result is a
12 STYLE AND STYLISTICS

unity, a whole. But it is a whole composed of linguistic

elements that we also know in other combinations; we
can therefore by a process of abstraction become aware
of them separately and discuss them as contributions to
the whole. The word that is magical in a particular line of
poetry may be quite inert in a different sentence; the
construction that is merely a bungle in one context may
serve a powerful expressive purpose in another. It is with
these phenomena, whatever our philosophy of meaning,
whatever our theory about the psychology of the cre¬
ative process, that the study of style is concerned.

The Beginnings of Modern Style-study

The effective impulses to modern style-study were

various, but they can probably be reduced to two. One
comes from historical linguistics and one from literary
criticism. The impulse from criticism was largely an
Anglo-American affair, that from linguistics was largely
continental European, arising particularly out of Romance
philology. The two still remain imperfectly united. Their
aims were partly different, and there is no point in trying
to reconcile studies that have different aims. However,
in their most interesting work the two schools are really
concerned with the same things, and it should be possible
to see them both in the same perspective. In both cases
the motive force was a reaction against scholastic dis¬
ciplines that seemed mechanistic and external, hardly
related at all to the living texture of literature. The
European situation is described in a vivid passage of
autobiography by the great Romance philologist Leo
Spitzer:

When I attended the classes of French linguistics of my

great teacher Meyer-Lubke no picture was offered us of
 THE CONCEPT OF STYLE 13

the French people, or of the Frenchness of their language:

in these classes we saw Latin a moving, according to relent¬
less phonetic laws, toward French e (pater, pere); there we
saw a new system of declensions springing up from
nothingness, a system in which the six Latin cases came to
be reduced to two, and later to one.... I was a long while
realizing that Meyer-Liibke was offering only the pre¬
history of French (as he established it by a comparison
with the other Romance languages), not its history. And
we were never allowed to contemplate a phenomenon in
its quiet being, to look into its face: we always looked at
its neighbours or at its predecessors.

When I changed over to the classes of the equally great

literary historian Philipp August Becker, that ideal French¬
man seemed to show some faint signs of life—in the spirited
analyses of the events in the Telerinage de Charlemagne,
or of a Moli^re comedy; but it was as if the treatment of
the contents were only subsidiary to the really scholarly
work, which consisted in fixing the dates and historical
data of these works of art, in assessing the amount of auto¬
biographical elements and written sources which the poets
had supposedly incorporated into their artistic productions
(Linguistics and Literary History, 2).

It was from a dissatisfaction with this state of affairs

that Spitzer came to evolve his own methods of style-
study, designed to achieve a far more intimate and in¬
ward contact with the work of literary art. The passage
above was written in 1948, of a period forty years earlier;
and the discontents it is describing mark a movement in
literary studies, paralleled, of course, in many other fields,
away from the positivist methods of the nineteenth cen¬
tury. Spitzer combined in an unusual degree the equip¬
ment of the professional linguist with the tastes of an
ardent and enthusiastic student of literature. Other
European linguists approached the matter from a less
i4 STYLE AND STYLISTICS

literary point of view. Charles Bally, for example, con¬

ceived of stylistics as the study of the expressive re¬
sources of a given language, and excluded from it the
study of literary language, language organized for an
aesthetic purpose. This self-denying ordinance was not
to be maintained, and Bally’s followers, Cressot,
Marouzeau and Devoto, interested themselves also in the
literary manifestations of language. It remains true, how¬
ever, that Continental style-study was conducted by
students of language rather than by men of letters in the
wider sense; and this has continued to affect the general
orientation of the subject.
In England the technical study of language was not
on the whole so highly developed, but the situation was
not greatly different from that described by Spitzer.
Literary studies at the beginning of this century were
largely dominated by Germanic philology; and much time
and energy was devoted to the linguistic study of Old
and Middle English. But the procedure remained arrested
at the Meyer-Ltibke stage; it was very little extended
towards literary study and not at all towards more
modern literature. Literary studies were largely a matter
of literary history and annotation—facts about literature
rather than literature itself. What Spitzer says of Becker’s
lectures was largely true of most literary study in the
Anglo-American sphere before the First World War; and if
these restrictions were evaded it was mainly by a way¬
ward and unfocused belles-lettres.
It so happened that just about this time, in the years
preceding 1914, a new literary movement made its ap¬
pearance. The poetry and criticism of T. S. Eliot, the
poetry and propaganda of Ezra Pound, the prose of
Joyce and the polemics of Wyndham Lewis were at
first a sort of underground movement in opposition to the
literary establishment; but they soon became something
 THE CONCEPT OF STYLE 15

like a literary establishment themselves. While doing

many other things as well, these writers all tended, in
their several ways, to focus attention on the verbal tex¬
ture of the work of art, on literary craftsmanship, and to
cast off as an obsolete irrelevance most of the external
literary history and conventional literary judgment that
had preceded them. This new critical outlook was at first
quite divorced from the academies and from scholastic
literary studies; but in a surprisingly short time it cap¬
tured them, or was captured by them. No one lectured
any more about the middle period of the novel of pas¬
sion or the effect of the French Revolution on Words¬
worth and Coleridge. The interest was in the texture of
the poets’ language, seen as a symptom of the quality of
their imagination, a way into the heart of the creative
process. In this way a strong impulse towards what can
be called in the broadest sense style-study came from
a totally new direction.
The literary history of this period is not my concern
now, but a few illustrations should be offered. If we
compare the immensely influential criticism of Eliot with
that of, say, Arnold, we are at once struck by Eliot’s much
freer use of quotation. The argument is supported by
close reference to particular passages. The prolonged
elucidation of the nature of poetical wit in the essays on
Marvell, the Metaphysical poets and Dryden is mainly
developed by showing what is going on internally in
particular poems and particular stanzas, and by con¬
trasting these passages with others by poets of a different
period. Wyndham Lewis opens his Men without Art, a
book which develops into a piece of savage literary buc¬
caneering, by contrasting a passage from Henry James
with a greatly inferior passage from Aldous Huxley. The
critics were hardly interested in detached scholarship;
they were writing polemic and propaganda for their own
16 STYLE AND STYLISTICS

literary movement. But as a by-product they were focus¬

ing attention on the intrinsic and interior constituents
of the work of art, and so, in a way not at all to be
expected, reveal some community of intention with a
brilliant, purely academic scholar, such as Leo Spitzer.
The injection of these new literary ideas into academic
study occurred at Cambridge in the 1920’s. The real
distinction of the Cambridge English school in its earlier
and more challenging days was that it was the only
university department in the world that knew anything
about modern literature. It may be said that even Cam¬
bridge did not know much; but that would be an illegiti¬
mate hindsight. It was remarkable at the time we are
discussing to find any awareness of modern literature
and its new critical orientations in an academic setting.
(Experto crede: in my own early education I saw some¬
thing of what was happening on both sides of the fence.)
I. A. Richards in his early teaching and writing called
attention to Eliot’s poetry and criticism; Eliot reviewed
Richards’s Science and Poetry in The Dial. The inter¬
change thus begun was continued by F. R. Leavis. A
brilliant, erratic Cambridge teacher, Mansfield Forbes, who
left hardly any published work, was influential in two
ways. His own lectures and classes offered examples of
minute and sensitive internal analysis of poetry; and his
effect on the development of Richards’s work was very
great. Richards came to literary studies from experi¬
mental psychology, and the application of his thinking
to literature was enormously assisted by Forbes.
Richards, with his interest in semantics and communi¬
cation theory, was deeply dissatisfied with the standards
of reading and comprehension achieved by conventional
literary training. The experiment recorded in his Practical
Criticism was a piece of field-work designed to investigate
this state of affairs. Short poems of varying degrees of
 THE CONCEPT OF STYLE 17

complexity and merit were given to a mixed class of

students and dons, without any indication of authorship
or date, and they were asked to comment on them. The
results revealed that without the aid of external informa¬
tion, literary history and the traditional judgments it
brings in its train, a surprising number of educated readers
were bewildered and helpless in interpreting what was
before them. In the second part of the book Richards
drew some theoretical conclusions from his experiment,
and made a number of extremely valuable suggestions
for improving matters. It is hardly too much to say
that he provided a new apparatus for reading and in¬
terpreting literary texts. It was of course to a large extent
a formalization of what the skilful reader had always
done intuitively; but the fact remains that the method had
never been made explicitly and generally accessible before.
The effects of this work were rapid and far-reaching.
Something like an educational revolution in the teaching
of literature took place in the thirties. The weight was
taken off literary history, facts about literature, and
attention was more and more concentrated on close
reading and interpretation of individual works. Practical
criticism, originally devised as a technique of inquiry, came
to be used as an educational method, as it still is. Genera¬
tions of literary students have been brought up to write,
not historical essays, but analyses and interpretations of
the works themselves, with close attention to verbal
texture and organization. Richard’s most brilliant pupil
William Empson interested himself in the question of
‘ambiguity’, or the plurality of meaning in any highly
organized piece of writing. The whole movement soon
spread to the United States, where it formed the basis
of what was there called the New Criticism, exemplified
in the work of John Crowe Ransom, Cleanth Brooks and
many others. The feed-back from this academic work to
l8 STYLE AND STYLISTICS

general criticism became evident before long. Un¬

supported judgments, mere statements of taste and pre¬
ference, were felt to be inadmissible; they had to be
backed up by precise reference and analysis; and criti¬
cism, which by now meant largely stylistic criticism,
began to make a claim for itself as an independent
discipline.
This did not take place without a good deal of con¬
troversy, and the battle was long ago fought to the
point of exhaustion. It is now obvious enough that
positions tended to harden in a slightly absurd fashion;
and the change in method, for all its indubitable bene¬
fits, began to reveal some attendant disadvantages. A false
dichotomy between ‘scholarly’ and ‘critical’ studies grew
up, and the new kind of stylistic criticism developed a
somewhat cavalier attitude towards both historical and
linguistic considerations. Gerard Manley Hopkins’s phrase,
‘low-latched in leaf-light housel’, was gaily glossed by
saying that it gained its concrete force from the word
‘housel’, which originally meant ‘little house’. It never
did, of course, mean anything of the sort, being derived
from a Germanic root which means ‘sacrifice’ or ‘holy
thing’. The current meaning, in fact, is the same as the
etymological one. Ingenious psychological explanations
of metaphors and images were given when straightforward
historical ones were staring us in the face. Most of these
matters were arguable, and some of them are still argued
about. What I believe is now evident is that this whole
school of Anglo-American stylistic analysis, lively and
invigorating as it often was, was also an extremely un¬
disciplined affair, and strangely innocent of scholarship.
It was not based on any positive body of knowledge. In
effect it disdained the appeal to knowledge of any kind,
and rejoiced in being an intuitive free-for-all. To be im¬
partially disrespectful to both sides, we may say that if
 THE CONCEPT OF STYLE 19

the vice of the Continental stylistics was pedantry, that

of the Anglo-American school was irresponsibility.
In English and American scholarship ‘practical
criticism’, ‘explication’, ‘close reading’, or whatever we
like to call it, has developed into a convention without
acquiring much more in the way of principles than it had
to start with. It is now a convention less deadening than
the external literary history that it replaced, but not much
less. Yet no one can doubt that its gains were real. The
intelligent student today has far better means of coming
to grips with a work of literature in its own essence than
any that were available fifty years ago. It seems likely
that what is needed now is an injection of some real
linguistic knowledge. It is not likely that the stylistic
study of literature will ever become a science, but there
is no need for it to be a riot of subjective fancy. Pro¬
fessional linguists are apt to say that their science has
literary bearings to which students of literature pay in¬
sufficient attention. I think it is obvious that most of what
the science of linguistics now does cannot be usefully
related to literature at all; but there are bridges to be
built, and it is in the area of stylistics that the oppor¬
tunities for doing this are greatest. In the chapters that
follow we shall examine what has been done in this
field, and if possible see how different lines of a study
could be brought together.
 2

Linguistic style-study

Up to Saussure

As soon as we look at it at all closely, the difference be¬

tween the literary and the linguistic study of style be¬
comes evident. Charles Bally points out that until the
nineteenth century language was never studied for its
own sake. It was always a question of what advantage
could be derived from linguistic study—for the logical
formulation of thought, for correctness of style, above
all for the understanding of the classical writers, re¬
garded not only as literary models but as linguistic norms.
This brought a number of consequences in its train:
a reverence for the written language; a corresponding
devaluation of the spoken language, which was regarded
as vulgar; the superstition of an unchangeable classic
language, proper as a model for all time, which should
be guarded against all innovation by a jealous purism.
The innovations, of course, occurred; the forms of purism
changed from age to age, and the situation was complic¬
ated by the fact that the ultimately unchallengeable stan¬
dards were situated in the dead languages, not in the
living vernaculars. The idea of a stylistic standard fixed
in the past or in another country is deeply rooted in our
culture. Horace in the Ars Poetica refers the Latin writers
22 STYLE AND STYLISTICS

to Greek models. The prestige of Ciceronian Latin was so

high in the Renaissance that it became the obvious duty
of the modern prose writer simply to imitate it. True,
there were rebellions. Politian defiantly asserted, ‘Non
Ciceronem me tamen exprimo’—I am not expressing
Cicero but myself. This did not, however, prevent genera¬
tions of schoolboys for 300 years after Politian’s time
being assiduously trained in the pastiche of Latin prose
and Greek verses. When in seventeenth-century England
a new generation revolted against the elaborate periodic
syntax and the formal decorative rhythms of Ciceronian
prose, the alternative first arrived was not to write as one
spoke or as the occasion demanded, but simply to
switch to another classic model—Seneca instead of
Cicero. And 100 years later we find Swift believing that
the English language, having reached its stage of classical
perfection, should be fixed in that state, and all future
deviations should be prevented.
It must I think be admitted that this has been the
normal literary attitude to problems of language for most
of our history. In this as in other matters the watershed
is the nineteenth century. It is then that the scientific
study of language begins, and it participates in the his¬
torical and evolutionary attitude that pervades all
nineteenth-century thought. A century that saw the rise
of evolutionary biology, anthropology and modern
historiography was not likely to remain attached to the
idea of an immutable linguistic norm fixed somewhere
in the distant past. The first great development was that
of comparative philology, in practice the comparative
study of the Indo-European languages. The discovery of
Sanskrit and the observation that it had affinities with
the modern European languages was made by the
Orientalist William Jones (d. 1794); but the systematic
organization of these facts dates from 1816 when Franz
 /

LINGUISTIC STYLE-STUDY 23

Bopp published his work on the conjugation of Sanskrit,

which studies the relation of Sanskrit with Latin, Greek
and the Germanic tongues. This was the beginning of a
linguistic science that continued to develop throughout
the century. The great popularizer of these researches in
England was Max Muller (Lessons on the Science of
Languages, 1861). Determinedly evolutionary in its
methods, formed on a biological model (languages are
arranged in ‘families’, words have ‘roots’), comparative
philology traced the development of verbal and gram¬
matical forms through time and space, from the recon¬
structed ‘primitive Aryan’ to the vernaculars of today.
It is a late stage of these studies that Spitzer describes
in his account of the lectures of Meyer-Liibke; and he
puts his finger very precisely on their limitations. Isolated
linguistic forms were studied in their development through
time (diachronically, to use a later terminology): but
the whole state of a language at a given time was never
presented for inspection. Still less did comparative
philology, boring its deep and narrow tunnels into the
past, provide any apparatus for studying the actual state
of the language in the living present.
The great change, the second great development of
modern linguistics, came with Ferdinand de Saussure.
The work of Saussure published during his lifetime reveals
him as a distinguished comparative philologist of the old
school. His first important publication was a treatise on
the vowel system of the Indo-European languages (1879),
and he continued to labour in the same field. But between
1906 and 1911 he gave three courses in general linguistics
which so changed their objectives as to constitute a
veritable new departure in the study of language. He
never published these lectures, and it turned out after
his death that they had never been systematically written
down at all. They had, however, made such a deep im-
c
24 STYLE AND STYLISTICS

pression of originality that a group of his pupils took on

the difficult and laborious task of reconstructing them
from his own notes. The result was the celebrated Corns
de linguistique generate, published in 1915, from which
most modern linguistics is considered to take its origin.
The main point in Saussure’s work, and also the one
of most interest for our purpose, is that he makes an
absolute disjunction between diachronic and synchronic
linguistics. Diachronic linguistics is historical linguistics
—in effect, the old comparative philology that we have
briefly described. By synchronic linguistics is meant the
study of the actual state of a language at a given time,
conceived as a complete, interdependent system of com¬
munication, actualized in daily life and divorced entirely
from its history and origins. It is this kind of linguistic
study that Saussure succeeds in getting on its feet. It
is the study of language as a present and living organism,
as against the study of its fossil remains. When Browning’s
Renaissance grammarian ‘Gave us the doctrine of the
enclitic 8e,/Dead from the waist down’, the death of
his members is a symbol of Browning’s own almost
prophetic feeling about the deadness of such studies; and
‘The Grammarian’s Funeral’ joins hands with Spitzer’s
account of Meyer-Liibke’s lectures. It was Saussure’s work
that brought linguistics into relation with the living
language; and, quite unliterary as it was, this could hardly
fail in the end to have some influence on literary studies.
The second point in Saussure’s course was the distinc¬
tion between la langue and la parole. La langue for
Saussure is a definite element abstracted from the hetero¬
geneous facts of language in general. It is the public,
conventional aspect of language, the system established
by a sort of social contract among the members of a
community which alone makes it possible for them to
understand each other. It is la langue that is described
 LINGUISTIC STYLE-STUDY 25

in dictionaries and grammars; and they are only possible

because la langue exists, necessary and unalterable by
individual volition. For la langue is always external to
the individual; he inherits it, he is born into it as he is
born into a society; it is not a function of his individual
will. La parole, on the other hand, is individual utterance,
an act of will and intelligence, serving individual ends.
La langue is a code, and la parole is the way the code is
used in an actual situation, or the ways in which it is
habitually used by an individual speaker. It is only in
la parole that la langue is actualized; yet la parole would
be impossible without the public, social system of la
langue. The linguistics of la parole were not much dis¬
cussed by Saussure and were considerably more de¬
veloped by his pupil. Bally. But the mere distinction is
important for stylistics; it contains the germ of the idea,
so often appearing in discussions of style, that there is an
impersonal norm of which style is the specialized or
individual variant.

Bally and His Successors

Saussure is not much interested in the literary manifesta¬

tions of language. For him the primary object of the
linguist’s study is what he calls the ‘natural’ sphere of
language—the spoken language. The written language, the
literary language, the language of poetry are specializa¬
tions, more or less detached from this living reality.
These ideas were carried much further by Bally, and he
is no more inclined than his master to move in the
direction of literature. Bally is virtually the inventor of
the term ‘Stylistics’ but he does not mean by it the study
of literary style. At the base of Bally’s thought is the idea
of language in the service of life, language as a function
of life, soaked in human affections, mingled with human
26 STYLE AND STYLISTICS

strivings, existing only to fulfil the purposes of life itself.

Language so conceived is the only real and living language
that exists. Everything else is an abstraction made for
purposes of study, or a specialized use, belonging to a
particular activity or a particular milieu. Current col¬
loquial speech is as worthy of study as the most refined
literary utterance—for the linguist indeed more worthy
of study.

The natural language, such as we all speak, is neither in

the service of pure reason nor of art; it does not have in
view either a logical or a literary ideal; its primary and
constant function is not to construct syllogisms, to round
off periods, to conform itself to the laws of the alexandrine.
It is simply in the service of life, not the life of a few, but
of all, and in all its manifestations; its function is biological
and social (Le langage et la vie, 14).

What then is stylistics? We associate the word with

a more or less literary inquiry; but for Bally it is
the study of expressive effects and mechanisms in all
language—la langue de tout le monde. The conception
depends entirely on a distinction between the ‘logical’
and the ‘affective’ characters of language. The ‘logical’
aspect of language, the expression of pure ideas, the com¬
munication of facts in themselves, is an abstraction,
realized only in the artificial language of science, and
then imperfectly. Aucun homme ne vit par la seule
intelligence; il n’y a pas d’idee pure qui aide d vivre.
My ‘lived’ thought is of quite another material than pure
ideas; actual language is everywhere penetrated with striv¬
ings, affections, judgments of feeling and judgments of
value. The intellectual judgment The earth turns changes
into a judgment of value in the mouth of Galileo crying
before his judges: ‘E pur si muove!’ Saussure’s linguistics
had concentrated chiefly on the impersonal system, la
 LINGUISTIC STYLE-STUDY 27

langue. Bally’s stylistics studies all the ways in which this

impersonal system is converted into the stuff of living
human utterance.
His method is to consider all these living characters
of language as deviations from a norm. At first Bally
used the word ‘affective’ to describe such deviations, but
this proved too narrow, and later he talks of ‘affective
and expressive’ characteristics. The first norm proposed is
‘the logical or intellectual mode of expression, which one
might also call the language of the abstract, or the
language of pure ideas’. It is against this that the affective
characters of language are measured; the plain formula¬
tion serves as the standard, against which is set the for¬
mulation coloured by interest, feeling, pleasure or
displeasure, approval or disapproval. But there is another
class of particularities that require notice. These are not
primarily affective, but social. Certain modes of expres¬
sion suggest a certain social milieu—popular, refined,
learned, provincial or what not. Bally calls such effects
faits d’evocation; and these evocative effects are only
possible because there is also a norm against which they
are set—the common language, uncoloured by any special
social suggestions. Among these special dialects evocative
of a milieu Bally includes the written language, the
literary language, scientific language and familiar language.
The total result of such investigations applied to a single
linguistic communitity is to provide a description of the
entire expressive resources of the language as a whole.
Bally’s Traite de stylistique frangaise does this for
French; Marouzeau’s Traite de stylistique applique au
latin does the same for Latin. It is this study of the whole
expressive equipment of a language that is Bally’s princi¬
pal aim.
It is obvious that there is a great deal here that is of
the highest interest to the student of literature. As soon
28 STYLE AND STYLISTICS

as he turns his attention to the texture and detail of

literary utterance it is precisely with such transforma¬
tions that he is concerned. Yet as a linguist Bally ex¬
pressly refuses to make this transition:

There is an impassable gulf between the use of language by

an individual in the common, general circumstances im¬
posed on a whole linguistic group, and the use made of it
by a poet, a novelist, an orator. When the speaker is placed
in the same conditions as all other members of the group,
there exists by reason of this very fact a norm by which
one can measure the deviations of individual expression:
for the litterateur the conditions are quite different; he
makes a voluntary and conscious use of language...
secondly, and above all, he uses language with an aesthetic
intention; he strives to create beauty with words, as a
painter does with colours or a musician with sounds
(Traite de stylistique frangaise, 19).

But surely this can be questioned. Bally may personally

refuse to extend his consideration to literature; but it
does not seem nearly as certain as he supposes that
such an extension cannot be made. His contrast between
the spontaneous utterance of common life and the con¬
scious, voluntary utterance of literary composition is by
no means absolute. Some speakers choose their words
with care; some literary men write with great spon¬
taneity and freedom. His argument that aesthetic inten¬
tion introduces an entirely new dimension into language,
removing it from all ordinary considerations, is more
formidable. A measure of discontinuity is real enough.
All literature exists within, as it were, a parenthesis, dis¬
tinguishing it from actual discourse. But within the paren¬
thesis all the effects that have been observed outside
it are still active. Affective qualities, qualities evocative
of a milieu, occur in literature as much as they do in
common speech. It seems therefore that Bally’s limita-
 LINGUISTIC STYLE-STUDY 29

tion is unnecessary. The precise and detailed observation

of semantic and syntactical features that he applies to
common speech could equally be applied to literature,
and the analysis of literary works might well benefit
from it.
Some of Bally’s disciples indeed refuse to follow him
at this point. Marcel Cressot (Le style et ses techniques,
1947) expressly makes this dissociation:

In agreement with M. Bally up to now, we are going to

part company with him.... For us the literary work is
simply a communication, and all the aesthetics that the
writer puts into it are no more than the means of securing
the reader’s attention more firmly. This concern is perhaps
more systematic than in current communication, but it is
not of a different kind. We would even say the work of
literature is par excellence the domain of stylistics, precisely
because there the choice is more ‘voluntary’ and more
‘conscious’ (p. 3).

But although Cressot here takes a step into what Bally

regarded as an impassable gulf, although he makes use
of and discusses literary examples, he does not proceed
very far into the territory of the man of letters. He
analyses literary devices, but makes no attempt at the
analysis of a work of art. When all is said, the work of
this school belongs to linguistics rather than to literature.
What literary studies could learn from it is a technique of
accurate linguistic observation, a descriptive apparatus
with some pretensions to completeness. If Anglo-American
‘practical criticism’ had had any such foundations we
should have seen less of the merely capricious, less
pursuit of mere fortuitous crotchets, fewer attempts to
explain all literary phenomena by the use of some
fashionable catchword. Bally talks of stylistics as a science;
and literary style-study is never likely to be that. But
30 STYLE AND STYLISTICS

the attempt to give it some sort of scientific foundation,

some basis of ordered and demonstrable knowledge, might
have made it a less chancy and arbitrary affair than it
has been.
 3
Literary stylistics:
methods and problems

Stylistics and Literary Art

The field of literary stylistics is so wide that no attempt

can be made here to give a systematic record of its
evolution, especially as the developments have occurred
in several countries—parallel, overlapping or correlated.
In the next chapter I shall give some account of the
work of outstanding practitioners; in this one I shall out¬
line the topics and problems that literary stylistics at¬
tempts to deal with.
A marked contrast can be seen between the rigorous
and restricted methods of the Bally school of stylistics
and the relatively expansive procedure of those who were
more concerned with literature as an art. Bally and his
followers were concerned with establishing a general sys¬
tem of stylistic possibilities that could be applied to all
literary work as it could be applied to all types of utter¬
ance. A keystone of their system was the setting up of a
norm against which stylistic deviations could be measured.
From the literary point of view it is open to question
chiefly for taking uncoloured descriptive language as the
paradigm of all language. This may be a legitimate
methodological convenience, but there seems a certain
perversity in it, since on Bally’s own showing ‘natural’
32 STYLE AND STYLISTICS

language is hardly ever of this kind. The contrast can be

seen in the work of certain Germanic writers whose
primary point of reference is not a norm or a method but
the total being of a work of literary art. The difficulty
here is to distinguish anything that can be specifically
called ‘stylistics’. How does Stilforschung so conceived
differ from the vast unorganized field of literary criticism
—criticism of the ordinary historical, impressionistic or
belletristic kinds, which stylistics was supposed to im¬
prove on?
This can be illustrated in the work of Karl Vossler, a
writer of enormous and various learning. His vast work
on Dante (Die Gottliche Komodie, Heidelberg, 1907-10)
is mainly a profound and detailed inquiry into all the
various streams of medieval culture that contributed to
Dante’s thought. The American translation of the book
is simply called Medieval Culture, and this suggests what
is indeed the case, that the bulk of the work is cultural
history, not style-study at all. The long chapter on ‘The
Poetry of the Divine Comedy’ is an attempt to give a
stylistic analysis of the whole poem. But an attempt on
such a huge scale almost necessarily misses the specificity
and refinement that stylistic analysis demands. There is
nothing distinctive in the method, which indeed boils
down to a rehearsal of the incidents of the Commedia,
and a commentary on their emotional force, order and
arrangement—traditional literary commentary of a rather
pedestrian kind. Vossler seems to have been obsessed
with the idea of the internal analysis of a literary
work, without ever being able to decide on its precise
bearing and purpose. His book on Benvenuto Cellini is in
essence a psychological study; and in other works Vossler
is attracted to the study of national characteristics as
expressed through literary style (The Spirit of Language
is too much preoccupied with the German soul to be
 METHODS AND PROBLEMS 33
congenial to non-Teutonic readers). Vossler’s conclusion,
shared also by others, is that style-analysis simply coin¬
cides with literary criticism in its ‘objective’ form. It
is simply literary criticism with the element of arbitrary
personal preference purged away.
Another way of defining stylistics is found in Hatzfeld’s
clumsily titled essay, ‘Stylistic Criticism as Art-minded
Philology’. The distinction is indeed indicated in the title:
stylistic analysis is simply philology in the traditional
German sense (i.e. literary and linguistic study), with
the aesthetic dimension added. Bally’s stylistics, equally
with Meyer-Liibke’s comparative philology, severely re¬
nounced aesthetic considerations. If all or any of these
lines of investigation are re-directed towards aesthetic
objects we have literary stylistics. But these methodologi¬
cal considerations soon become extremely arid; let us
elucidate the situation a little farther by thinking of the
actual topics which literary stylistics has investigated.

Special Expressive Devices

Nearest to the methods of Bally and his school is the

study of particular expressive devices. In the work of
Cressot, Marouzeau and the linguists this occurs simply
as a part of a general stylistic organon, the total stylistic
tool-kit of a given language. Alter the context, study a
particular stylistic device as it is employed in an in¬
dividual work of art, or the work of an individual writer,
and we are at once within the literary sphere. We are
inquiring how a specific configuration of language is
used for a specific aesthetic purpose, or by what linguistic
means a particular aesthetic purpose is achieved. Studies
of special kinds of imagery, special choices of vocabulary,
special syntactical usages, all come under this head. I
shall not attempt to cover the extent and variety of this
34 STYLE AND STYLISTICS

kind of work. For that, as for all other kinds of study

referred to in this chapter, the reader must be referred
to Hatzfeld’s valuable Bibliography of the New Stylistics,
for the Romance languages; and the recently published
English Stylistics: a Bibliography, by Bailey and Burton.
Among the mass of bibliographical material on individual
authors and periods the distinctive existence of stylistic
work is not easily recognized. Let us take as illustration
the study of one particular stylistic procedure—the
so-called style indirect libre.
The ordinary grammatical distinction between direct
and indirect speech is known to everyone. It was in the
novel that the existence of another form was observed,
a form in which the context, the containing structure, is
that of indirect or reported speech, while a number of
elements (syntactic and lexical) of direct speech are also
allowed to remain. To the ordinary literary student this
may seem a trivial observation, but in the work of a
number of novelists it turns out to be a persistently em¬
ployed procedure, with its own distinctive flavour. It is
frequent in Jane Austen:

Scarcely had they passed the sweep-gate and joined the

other carriage, than she found her subject cut up—her hand
seized, her attention demanded, and Mr. Elton actually
making violent love to her: availing himself of the precious
opportunity, declaring sentiments which must be already
well known, hoping—fearing—ready to die if she refused
him; but flattering himself that his ardent attachment and
unequalled love and unexampled passion could not fail of
having some effect, and, in short very much resolved on
being seriously accepted as soon as possible (Emma,
Chapter 15).

Here the presentation is formally that of objective

narrative, narrative presented as fact; but the parts
 METHODS AND PROBLEMS 35

printed in italics diverge from this—they are a para¬

phrased version of what Mr. Elton actually said; the
expression with its absurdities and conventionalities is
neither Jane Austen’s nor Emma’s, but Mr. Elton’s. Syn¬
tactically, however, it is assimilated to the narrative:
‘She found Mr. Elton availing himself of the precious
opportunity, declaring sentiments’, etc. It could quite
easily have been presented in direct speech: ‘ “I avail
myself of the precious opportunity,” said Mr. Elton, “my
sentiments must already be well known to you. I flatter
myself that my ardent attachment cannot fail of having
some effect.” ’ In other places Jane Austen does present
such incidents in this way. On the other hand, it could
all have been done as objective narrative: ‘She found
Mr. Elton actually making violent love to her’, etc., and
the passage could have been continued in the uncoloured
objective manner. What we actually have is an inter¬
mediate form.
The existence of this form was first remarked in an
article by Bally in 1912, and, following his usual pro¬
gramme, he studied it simply as a current linguistic
device. Eight years later Proust, in an article in the
Nouvelle Revue Frangaise, remarked on Flaubert’s use
of this device, quite unconscious that it had been noticed
before. A. Thibaudet replied to him in a letter: two
years later he brought out his study of Flaubert, and in
the admirable chapter on Flaubert’s style the discussion
of this particular effect occupies a prominent place.
From then on a voluminous literature on the subject
has grown up (see R. Ullmann’s Style in the French Novel,
94-8). The form has been traced back as far as La
Fontaine, becomes more prominent in the late eighteenth
century, and almost a standard practice in nineteenth-
century fiction, alike in French, English and German.
Much of the discussion of this device has been a techni-
36 STYLE AND STYLISTICS

cal analysis of its various syntactic manifestations. Some

of these are trivial, but the method as a whole clearly
has literary significance. It is obviously connected with
the tendency to reduce the role of the omnicompetent
narrator, to incorporate the point of view of the
characters into the structure of the narrative. It is a
part, therefore, of a general movement in fiction from
the early nineteenth century on. It may be variously
motivated. It may spring from the desire to present
rather than merely to tell about the incidents of the
story. It may be a means of bringing in the subjectivity
of the characters, of portraying their inner life, while
preserving a greater measure of authorial control than
could be done by the use of simple direct speech. This is
often the case with Flaubert. It may be there simply to
give a flavour of liveliness and colour to passages of
merely functional narrative:

For half an hour Mr. Weston was surprised and sorry: but
then he began to perceive that Frank’s coming two or
three months later would be a much better plan, better
time of year, better weather; and that he would be able;
without any doubt, to stay considerably longer with them
than if he had come sooner (Emma, Chapter 18).

Flere the words italicized, with their abrupt, cheerful,

colloquial tone are obviously Mr. Weston’s; and they are
introduced much as a lively speaker will slip half-
unconsciously into mimicry in recounting the actions
and conversation of others. Perhaps more important than
any of these, the free indirect style, with its close colloca¬
tion of the author’s point of view and that of his
characters, becomes the vehicle of irony. It is therefore
one of the most important means by which the author
can convey his judgments and valuations without ob¬
viously intrusive commentary. The examination of
 METHODS AND PROBLEMS 37

Flaubert’s methods (by Proust, Thibaudet and Ullmann,

among others) has shown that the device has become
an integral part of the fabric of his prose, a procedure
by which some of his most characteristic effects are
obtained. An inspection of Jane Austen’s methods has
persuaded me that the same is true of her. And none
of this can be demonstrated without the detailed examina¬
tion of verbal technique that we call ‘stylistics’.
Still within the field of the novel, a more extended
kind of investigation is the study of narrative method
in general—first person as against third person; direct
narration by an ‘omniscient’ author, or the use of the
point of view of the internal characters; interior mono¬
logue, ‘stream of consciousness’ methods; and so on. The
impulse to these inquiries was largely given by the pre¬
faces of Henry James, and a recent example is Wayne
Booth’s admirable and comprehensive study. The Rhetoric
of Fiction. In poetry the tendency has been more to¬
wards the study of individual authors; but the isolation
of such qualities as ‘wit’ and ‘irony’ which has played
such a large part in recent poetical criticism is also an
investigation of particular stylistic devices. Again the
split between ‘stylistics’ and general criticism can be
observed. Formally stylistic studies attempt to be neutral
and objective; in more general criticism the observation
and analysis is often in the service of some other end.
Since there are always conventional judgments to be re¬
examined, and since every age makes its special de¬
mands, this must always be so. But it should be one of
the functions of stylistics in the narrower sense to give
partial and tendentious criticism some solid material to
work on.
Other stylistic devices that can be studied in similar
ways are almost innumerable: word-order, repetition,
rhythmical and musical patterns, metaphor, symbol and
38 STYLE AND STYLISTICS

imagery, local colour, synaesthetic effects. Some of these

are obviously of greater range and centrality than others:
but it will commonly be found that almost any stylistic
pecularity that is genuinely prominent and observable in
a particular writer can serve as a key to his artistic
procedure. Literary critics who are unaccustomed to
stylistic analysis tend to be sceptical about the significance
of such investigations. They often, indeed, look un¬
promising from the outside; it is only by actually using
such studies, or, better still, making them oneself, that
their value as a way into the work of literary art can be
experienced. It must be added that many of these
stylistic inquiries are designedly limited to observation,
analysis and record. To the critic of a comprehensive
turn of mind these give an impression of incompleteness.
They become authentic literary studies only when the
linguistic observation is used—when conclusions are
drawn from it that tell us something of importance
about the nature and meaning of the work as a whole.

Individual Style
The most familiar kind of style-study is the study of
individual style of a single author. The Romantic em¬
phasis on style as the expression of individual person¬
ality has brought this to the notice even of unliterary
readers, and post-Romantic depreciation of the expressive
outlook has put this particular approach somewhat under
a cloud. But there is no need to cross the line from
literary study to illicit biography, and in fact the study of
individual style is universally practised on various levels
of technical sophistication. A literary work is a verbal
structure and even, the critic who is primarily interested
in the history of ideas or the social implications of litera¬
ture can hardly proceed beyond generalities without pay-
 METHODS AND PROBLEMS 39
ing some attention to the way in which words are used.
It may again be asked therefore how style-study differs
from general literary criticism. The broad answer is that
in many kinds of criticism attention to verbal texture
may be intermittent, often unaware of itself and often
uncritical of its own methods. Many critics begin from
biography, from literary history or the history of thought
and arrive at close consideration of the literary work
itself only at the end. The characteristic feature of style-
study is that it begins from the literary work itself,
from words and the way they are combined in a par¬
ticular body of writing. There is no limit beyond which
the student of style is forbidden to go, but at least he
starts from a positive and identifiable point.
Whoever begins to look into these matters will im¬
mediately be struck by the immense body of such work
in the Romance languages and their relative paucity in
English. The tendency to separation between linguistic
and literary studies in the English-speaking world has
meant that comparatively few students of literature have
been willing to begin from close consideration of
language. Indeed, the consideration of a writer’s language
frequently comes as a sort of icing on the cake after
every other aspect of his work has been dealt with. The
claim of stylistics rests essentially on the proposition
that the farthest ranges of a writer’s art, the depths of his
emotional experience, the heights of his spiritual insight,
are expressed only through his words and can be appre¬
hended only through an examination of his verbal art.
Even if this claim is admitted it must also be admitted
that there is a genuine difficulty in making the transition
to these larger considerations from the particular features
of vocabulary and syntax with which the style student
generally starts, and that short-cuts of various kinds are
possible. The equipment of the linguist is frequently
D
40 STYLE AND STYLISTICS

different from that of the literary student, the one being

inclined to positive observation, the other to intuitive
perception and speculation; and apart from temperamen¬
tal differences of this kind there are real difficulties of
method. We are most of us incurably inclined to think
of ideas as the ultimate reality and words as their merely
accidental clothing, and over large areas this view may
have considerable justification. An expository writer
working within a well-established convention may exhibit
nothing particularly individual or characteristic in his
way of handling words and may afford very little material
for the student of style to work upon. And even when he
is considering an imaginative writer the literary student
may often find that he has had very little training in
observing the correlation between an intuitively observed
literary quality and the specific verbal means by which
it has been brought about. It is precisely this kind of
training that stylistic study professes to give. But it must
be confessed that it was some time before style-study
arrived at this point.
At one extreme we have the pure linguistic approach.
This tends to work by accumulation, by a complete inven¬
tory of the stylistic qualities of an author - vocabulary,
sentence structure, syntactic peculiarities, imagery and so
forth, listed according to some predetermined scheme.
In much early work of this kind no literary con¬
clusions were drawn. What we have is virtually an ac¬
cumulation of evidence on which such conclusions might
be based, but no more. A further effect of this procedure
is that in a complete inventory much of what is recorded
may virtually be waste matter. Many of the qualities
described have nothing particularly characteristic about
them and lead to no increase of literary understanding.
Much of what is presented is not the fruit of authentic
observation but results rather from the mechanical
 METHODS AND PROBLEMS 41
application of a set scheme. It is the prevalence of such
studies in which the aesthetic dimension has either been
renounced or has never been arrived at that has given
rise to a suspicion of stylistic work among many literary
students. Much will naturally depend on the nature of
the author under consideration. A pedestrian novelist
like Trollope may at first sight offer very little purchase
for a stylistic investigation. But it must be remembered
that his humour, his humanity, his understanding of the
society of his time, whatever qualities we find in him,
have only been materialized through verbal means and
they must in principle therefore be open to verbal in¬
vestigation. In the case of an intricate and difficult author
like Mallarme stylistic investigation is obviously re¬
quired. His syntax needs elucidation in order that it may
be merely understood. His vocabulary is often thought
to be excessively rarefied and we need to know from
what sources it is drawn. It is often said that his style
has been deeply influenced by his study of English. But
the student will want to know whether this is really so;
whether English syntax really does anything to explain
the extraordinary behaviour of Mallarme’s sentences. All
these matters have been investigated in an admirable
study by J. Scherer, which although it expressly re¬
nounces all consideration of meaning, all consideration of
the total bearing and significance of Mallarme’s art, yet
succeeds in being an indispensable preliminary to any
wider understanding. And since in the case of Mallarme
so much of his total significance is realized in his intricate
struggle with means of expression, even a study so de¬
liberately limited as this one can be of great literary
significance.
At the other extreme from these linguistic procedures
we have literary criticism which deals with stylistic
matters in an entirely unsystematic way. Such criticism
42 STYLE AND STYLISTICS

is apt to ask questions without suggesting any plausible

means of answering them, while on the other hand re¬
maining unaware that there are real questions to which
more or less verifiable answers can be found. At its best
such criticism may achieve great intuitive insight or have
programmatic value in changing the direction of literary
ideas. T. S. Eliot’s essays on Marvell, the Metaphysical
poets, and Dryden are of this kind. At its worst, criticism
of this type is a mere orgy of opinion. An example of
the worst would be F. R. Leavis’ celebrated essay on
Milton’s verse. That this curious paroxysm should for
long have been accepted as a serious contribution to litera¬
ture is an indication of how low the intellectual level of
literary discussion in England could sink during the
thirties and forties. Stylistic analysis must begin with an
act of submission to the work as it is in itself. Leavis’s
essay begins from a prefabricated judgment—a second¬
hand judgment, since his conviction that Milton has been
dislodged is merely a vulgarization of ideas that Eliot
had already worked out in the essays mentioned above.
The condemnation of Milton’s style is unsupported by
analysis or accurate description. Long quotations are
offered, but nothing is done with them. The old school¬
master’s dictum that Milton wrote English as though it
were Latin is repeated, but no syntactical demonstration
is offered. Otherwise the argument is sustained entirely
by rhetoric. The norm of comparison suggested for Milton
is Shakespeare—that is to say, an epic poet is compared
with a dramatic poet; a poet who writes almost entirely
of events and objects beyond this world is compared
with a poet who writes almost entirely of events and
objects on middle earth. If this is criticism, the distinction
between it and stylistic analysis will be clear. However,
this would be to reduce the idea of criticism to an absurdity.
And it is only fair to add that there are many examples of
 METHODS AND PROBLEMS 43
critical description, which start from parti-pris or the desire
to prove a point, as the stylistic analyst does not, but still
make a genuine attempt at understanding: and some of
these are by Leavis himself.
Effective style-study must lie somewhere between these
two—between hard-line linguistics and subjective criti¬
cism. It is a dogma in much recent criticism that de¬
scription involves evaluation, but this may be doubted.
Accurate and evaluatively neutral criticism is neither
impossible nor useless. Stylistic description almost in¬
evitably depends on comparison with some norm and
this norm should be seen to be a relevant one. The
language of a novelist might quite properly be viewed in
the light of the language of common life: but with an
epic poet in the classical tradition there would be very
little sense in talking in these terms. Style-study need not
involve pedantry or a self-conscious excess of system.
French critics in particular offer many brilliant examples
of style-study conducted with that immense sense of
responsibility towards language in which French critics
are on the whole so much superior to the English, and
conducted also with the elegance and lucidity of ordinary
literary discourse, entirely without pedantry or technic¬
ality. The same thing may, of course, be done in a more
scholastic and technical manner. The German writers in
particular tend to be devotees of system and to write
less for the general, educated reader. What all genuine
literary style studies ought to have in common is that
they are not mere catalogues of linguistic features, but
are directed to the understanding of a work of art.
Such studies are likely to abandon the complete inven¬
tory and to be more selective in their methods. But how
is the selection of stylistic features for examination in¬
itially made? There is always the danger of a mere arbi¬
trary selection of features which happens to have struck
44 STYLE AND STYLISTICS

a particular observer, for accidental or subjective reasons.

This can often be avoided by putting the individual style-
study in its historical setting. The individual utterance of
a novelist is compared with the common speech or the
general narrative style of his time. The individual utter¬
ance of a poet is brought into relief by seeing it as a diver¬
gence from the general practice of a school; as, for
instance, when we compare the language of Gerard
Manley Hopkins with that of the generality of Victorian
poetry. Obviously if this is to yield useful results it is
important to select the right frame of reference. But we
can argue in quite another way; it can be said that any
point of entry may serve as means of access to an
author’s work, and that any approach if carried far
enough will lead to its centre. An extremely interesting
essay by Sir George Rostrevor-Hamilton, ‘The Tell-tale
Article’, arrives at some penetrating observations about
modern English poets, simply by examining their use of
the definite article. To discover what stylistic features
will be revealing is largely a matter of experience and in¬
tuitive talent. We will reserve further consideration of
this matter until we discuss the work of Leo Spitzer in
the next chapter, since Spitzer has made this method his
own..
It is relevant to ask what the object of an individual
style-study ought to be. Much of the linguistic work
we have discussed seems quite untroubled by these con¬
siderations; it seems to regard style-study as an end in
itself, and to be quite content with a mere listing of an
author’s idiosyncrasies of expression. On this basis it is
hardly possible to distinguish mere mannerisms from
important expressive qualities, or trivial and accidental
habits from profoundly important imaginative means of
expression. Once we go beyond these limits, perhaps
the most frequent and obvious aim of individual style-
 METHODS AND PROBLEMS 45
study has been psychological. If we want to see the mere
bundle of expressive characteristics as part of a totality,
the totality that seems most obviously to present itself
is the personality of the author. The wild exuberance,
the mock learning, the profuse verbal coinages of
Rabelais; the marmoreal self-conscious coldness of
Landor’s style—these are easily seen as the expressions of
a particular type of character; and correlations of this
kind are not hard to make. Such studies may even be¬
come psychoanalytical, and they may very well end by
taking us outside literature all together. But the naive
equation of what a writer reveals in his work with his
historic personality is always very uncertain. The sculp¬
tural serenity of Landor’s literary attitude seems to have
been largely a defence against a passionate and disordered
temperament. A much more legitimate and much more
rewarding study occurs if we shift the terms a little—if
we transfer our attention from the historic personality of
a writer to his creative personality, to his imaginative
function. It is not enough to regard style, metaphorically
speaking, as the artist’s personal handwriting, and to treat
it as a graphologist would treat actual handwriting. The
literary style-analyst is studying works of art, not varieties
of human character. That these works are the outcome
of a particular temperament may have a peripheral interest;
but the real object of the student’s search is the organiz¬
ing aesthetic principle.
If the study is of a single work, or splits itself up into
studies of single works, we fairly obviously replace psy¬
chology by an examination of the artefact itself. Most
criticism begins by studying the major structure—plot or
character, thought or feeling. Style-study begins as it
were at the other end of the scale, with the precise verbal
manifestations; and this is not only a manner of method;
it enshrines a kind of faith—a faith that it is only by the
46 STYLE AND STYLISTICS

close and intimate examination of verbal texture that

the true being of a work of literary art can ever be
reached. Movements in this direction can be seen
everywhere in modern critical writing. Wilson Knight
sees a Shakespearian play as an expanded metaphor; T. S.
Eliot is apt to depreciate the originality and importance
of the poet’s thought and concentrate attention on his
verbal medium. The logical end of this process should
be a more exact attention to the verbal medium than
informal and intuitive criticism can generally attain. It is
this that style-analysis aims to supply.
The study of an individual style conceived in this way
may broaden out from pure linguistic description to in¬
clude almost anything found in general criticism. But
there are some exclusions. An essay like Leavis’s on
Milton’s verse is essentially polemical and tendentious in
its intention. It does not aim at an accurate description of
Milton’s language and rhythms; it is sketchy, suggestive
and highly coloured. Nor does it aim at penetrating with
sympathy and understanding to the heart of Milton’s
poetry; its purpose is to change the direction of taste. Ten¬
dencies of this kind are quite common in criticism,
though usually to a less extravagant degree. There are
various reasons, some creditable, some less so, for which
a critic may aim at the indoctrination of his readers rather
than at the increase of understanding. The criticism of
a poet is often disguised propaganda for his own creative
work or for that of his friends; the criticism of a man
deeply engaged in the literary controversies of the day is
propaganda for a sect, a party or a school of thought.
Stylistic analysis on the other hand aims at objectivity;
some have even said that its purpose is a scientific know¬
ledge of literature. All claims to turn literary study into
a science are extremely dubious; it is hard to imagine
a state of affairs in which the style-study of an individual
 METHODS AND PROBLEMS 47

author could arrive at the incontrovertible status of a

scientific demonstration. There will always be room for
disagreement about the distribution of emphasis, about
the relative importance of different features observed;
but it is possible to point out objectively the existence of
certain linguistic features; it is possible to arrange these
in a logically and psychologically compelling order; and
it is possible to bind these together into an argument
that can reach, if not certainty, at least a very high
degree of persuasiveness. No one would wish to wipe
out the partisan and speculative elements from criticism
altogether, but there is a great deal to be said for found¬
ing them on a basis of agreed, demonstrable analysis and
description.
Possibilities and limitations will be exhibited more fully
in the next chapter, when we discuss the work of in¬
dividual practitioners, but some illustrations may be given
here. Dr. Johnson as a prose writer has a very strongly
marked style, clearly differentiated from that of his im¬
mediate predecessors. An obvious way of beginning the
study of his style would be to distinguish it from that
of the age of Swift and Addison. Certain features of his
sentence-structure, such as his continual use of antithesis,
command immediate attention. When, however, we begin
to inquire into the function, the purpose of this balanced
and antithetical sentence-structure, we at once tread on
less certain ground. To a superficial observer it would
seem to correspond to something assured, commonsensi-
cal and logical in Johnson’s mind; but a rather closer
examination makes this seem less likely. Many of his
antitheses turn out on inspection to be decorative rather
than functional; they serve scarcely any logical purpose,
they seem to be there rather because Johnson enjoyed
them. From there we could go on to consider Johnson’s
theoretical views on prose composition, his belief that
48 STYLE AND STYLISTICS

the important truths are all known in advance and that

the function of the writer is chiefly to present the familiar
in a striking and effective manner. This accounts for some
of the features of his balanced and formal prose style
but not for all; there is a good deal left over that seems
to contain an element of sheer play, a delight in verbal
patterning for its own sake, a delight always kept in
check by a sober, responsible and morally oriented mind.
If we continued on these lines we should be moving
towards a description of the whole nature of Johnson’s
creative impulse.
It would not, however, be very profitable on the whole
to consider Johnson’s prose works as individual aesthetic
objects. Much of his work was occasional in its origin
and utilitarian in its purpose; almost all of it was part of
a continual process of musing, reflecting, informing,
moralizing rather than full imaginative creation. We
should be rather less inclined to discuss the style of, say,
a lyric poet in these wholesale and comprehensive terms,
for we expect the productions of the lyric poet each
to be a self-subsistent entity. If we were to take a major
lyrical reflective poem—say, Keats’ ‘Ode to a Nightingale’
or Wallace Stevens’ ‘Sunday Morning’—we should hardly
think of discussing it as part of the general stylistic pro¬
cedure of its author: we should rather wish, beginning
with its language, to arrive at a comprehension of its
total form and meaning. And we should be much less
inclined to make a transition from the work to the mind
of the author for the simple reason that the work itself
is sufficiently complex and complete to command our
whole attention.

Period Style

The most obvious development of the individual style-

 METHODS AND PROBLEMS 49
study is its extension to the style of a whole period.
This nevertheless presents certain difficulties. While we
are studying the work of one author we have a de¬
fined body of material to investigate; if we extend the
investigation to the work of a whole period we have
always far more material than can be dealt with. The
problem of selection immediately makes its appearance.
How is the selection to be made? If we choose to study
a period through the most strongly marked and
characteristic passages of its literature there is always a
danger of predetermining the state of the case. We choose
the evidence to suit an answer that we have already
in mind. This sometimes happens with the Romantic
poets. By focusing on illustrations drawn from, say, ‘The
Ancient Mariner’ and ‘Resolution and Independence’, we
tend to forget how much of the eighteenth century sur¬
vives in Wordsworth and Coleridge, and how readily
they could slip back into the eighteenth-century manner
at times when their imaginative pressure was low. On
the other hand, we would all, I suppose, agree that the
style of a period is constituted by its most original
literary minds in their most vigorous moments—the
‘Epistle to Arbuthnot’, the close of the Dunciad, rather
than the undistinguished mass of couplet-writing at that
time; Dickens rather than Wilkie Collins.
Vocabulary, image, sentence-structure, the proportion
of nouns to adjectives, of nominal phrases to verbs—all
these things go to make up the characteristic flavour of a
period style; and they can all be investigated on a quanti¬
tative and statistical basis. If this is done the result may have
some claim to be scientific, but it is unlikely to be of much
literary significance. These difficulties are seen in the
works of Josephine Miles on period styles in English
poetry, to which we shall return. That we all have a
vague composite idea of the style of Augustan prose.
50 STYLE AND STYLISTICS

of the Romantic poetic vocabulary, is undoubtedly true;

how far we can do anything to make this more precise,
and how far it is worth doing, is another question. In
individual style-studies we come to understand a work of
art, or an individual artist’s whole creative impulse; in
period studies we can only arrive at that ill-defined and
often self-contradictory abstraction, the mind of an age.
It is very noticeable that Continental students are far
more inclined to hypostatize that abstraction, the style
of a period, than we are in the English-speaking world.
Serious attempts are made to define the Baroque, the
Mannerist and the Rococo styles. It is noticeable that
the terminology is borrowed from the visual arts: and
many such studies seem to be an attempt to apply the
methods that Wolfflin used in his Vrinciples of Art
History (1915) to the study of literature. In this book,
of fundamental importance in its own field, Wolfflin dis¬
tinguishes between the linear and the painterly, between
plane and recession, between closed and open form, and
so forth; and on this basis attempts the distinction be¬
tween the styles of different periods. I cannot think of
any good reason why the same methods should not be
applied to literature, yet in practice the results of any
such attempt always seem to be both arid and uncertain.
It may be that in art history the illustrations can be
more immediately and rapidly apprehended; it may be
that in considering literature the isolation of particular
qualities in this way forbids all proper consideration
of meaning, and that style-study that fails to go on to
the consideration of meaning is doomed to sterility. Per¬
haps there is a concreteness and particularity in the
English literary mind that makes it radically unsympa¬
thetic to such studies. Certainly the best English work of
this kind seems to occur in passing and in the course of
doing other things; it also seems to move as rapidly as
 METHODS AND PROBLEMS S'

possible from the general to the particular. Geoffrey

Tillotson gives us not generalizations about Augustan
poetic diction, but a close and precise description of the
language of Pope’s descriptive and pastoral poetry; Helen
Gardner abandons generalizations about Metaphysical
poetry, and tends instead to show us the individuality
and variety among the poets of that school.
In our literary climate definition, classification and
labelling is apt to be thought of as a particularly barren
exercise. To establish a concept of the Baroque or of
Mannerism in literature, and then to assign Milton to
the first and Donne to the second, seems neither con¬
vincing nor satisfying. But we may be wrong. I am often
haunted by the suspicion that the art historians, with
their schools and styles and periods, have a command of
their material and of its development that literary his¬
torians seem to lack. And to many Continental students
our unwillingness to deal in such ideas would seem but
another example of our well-known incapacity for going
beyond the barest empiricism.
All that has been said about the attempt to find the
style of a period can be said with even more force about
the attempt to define the style of a whole national litera¬
ture. There was a time when it was seen as a virtue in an
English writer to be characteristically English, but this
belongs to a simple and comfortable patriotism rather
than to literary intelligence. It has on the whole been
left to the Germans to bring the study of specifically
national styles into the range of serious intellectual
endeavour. Examples of this occur in the work of Vossler,
Curtius, and even occasionally in that of Spitzer. We
tend to regard these endeavours with suspicion or in¬
difference. Yet it cannot be doubted that there are
national styles, and they may be legitimate objects of
curiosity. There is a Frenchness about French literature;
52 STYLE AND STYLISTICS

we do feel in reading French poetry the presence of a

complex of characteristics quite different from those
that we experience in reading English poetry, and we are
quite willing to talk about it, on a loose and conversa¬
tional level. Yet both the desire and the equipment to
bring this into the area of serious discussion are lacking.
So far as these exercises are mere outbreaks of literary
nationalism we may willingly dispense with them; yet
we can imagine a serious study of the separate contribu¬
tions of individual national literatures, seen as constituent
parts of that great organism which Goethe conceived
of as ‘world literature’; and we can imagine that this
might not be without its value. It does not, however,
seem to be within the range of present possibilities.

History of Style

If it is difficult to arrive at a characterization of the

style of a particular period, to write any consistent his¬
tory of style seems more difficult still. We can trace
the history of a single stylistic device, see how it changes
in form and develops in function. We can, admitting the
difficulties already mentioned, arrive at some sort of pic¬
ture of the prevailing style of a particular epoch. But
to move through the centuries on a broad and compre¬
hensive front is virtually impossible; and all such attempts
tend either to split up into unrelated fragments or to
spread out into a swamp of generalization. To use
Saussure’s terminology, we have not yet found a way of
combining the synchronic with the diachronic point of
view, and perhaps we never shall. The best that can be
done is take a series of cross-sections, or a series of typical
examples, and place them side by side. This is not history,
but it is perhaps the nearest we can get to it. There have
 METHODS AND PROBLEMS 53

been some notable examples of this, the pre-eminent one

being Auerbach’s Mimesis.
Apart from the difficulty of marshalling a mass of
material relating to one period and keeping it all moving
together through successive developments, there is the
still greater difficulty of anomalous and irregular survivals.
The style of imaginative prose in the early nineteenth cen¬
tury goes through many and rapid changes: but the style
of expository writing hardly alters in any distinct way
from that of the later eighteenth century, until the
threshold of the modern period is reached. As we have
remarked, there are similar survivals in poetry. The most
popular and influential criticism is often misleading. The
propaganda for a new classicism in the early part of this
century would lead us to suppose that there was a com¬
plete break with Victorian stylistic tradition about 1914.
And this has tended to obscure the presence of decidedly
Tennysonian elements in the verse of T. S. Eliot, and the
strong continuity between the verse of Browning and that
of Pound. We have been led to suppose that Gerard
Manley Hopkins was one of the founding fathers of the
modern movement in poetry: yet actual traces of his
stylistic influence are few. It looks very much as though
such matters can be properly discussed only by tracing
the development of individual poets.

Statistical Methods

Something should be said at this point about statistical

methods of investigation, now becoming much more pro¬
minent with the use of computers. They are probably repug¬
nant to most students of literature, Luddites by nature,
as Lord Snow has observed. It is felt on the one hand that
insensitive and inappropriate methods are introduced into
literary scholarship by such means; on the other, that
54 STYLE AND STYLISTICS

literary students are unwilling to submit their observa¬

tions to positive verification. Demarcation disputes are
not usually very rational affairs, but there are some real
questions to be asked and some answers can be given.
If it is asked whether statistical information is ever rele¬
vant to style-study the answer must be a qualified ‘Yes’.
Nearly all criticism, even that most stigmatized as im¬
pressionistic, employs it in a loose and informal way.
If a critic remarks on Johnson’s antithetical style, he
means in the end that more antitheses will be found in a
typical passage of Johnson than in a typical passage of,
say, Addison of similar length. This is finally a statistical
matter. By choosing a number of suitable passages and
doing the necessary calculation, the results could be ex¬
pressed numerically as a statistical average—Johnson so
many antitheses per 2,000 words, Addison so many. And
this is only to put the original observation into numerical
form.
Has anything been gained by performing this opera¬
tion? Here the answer is less simple. If there is any doubt
whether Johnson’s style is actually marked by a free use
of antithesis, this is obviously the way to settle it. A
statement about the frequency of a particular stylistic
device will then have been verified by appropriate statisti¬
cal methods; and these are the only methods by which it
could have been verified. Sometimes this is a real gain.
Received ideas about the style of a particular period or a
particular author sometimes turn out to have no basis in
fact when subjected to this kind of investigation. If snarks
can be shown to be boojurns only by statistical analysis,
why then we must use it. In the Johnson case, which I
take to be an ordinary and typical one, I cannot see that
we are likely to gain anything. No one is in serious doubt
that Johnson did make frequent use of antithesis; and
for most literary purposes the common observation can
 METHODS AND PROBLEMS 55

stand without further precision, just as the observation

that it is a cold day can stand without consulting the
thermometer. And the questions of literary importance
are likely to be how he used antithesis, what he did with
it, what part it plays in the total economy of his work—
none of these open to statistical investigation.
There are people who talk as though any gain in
numerical precision is valuable in itself. From the literary
point of view this is nonsense. Suppose (to keep to our
painfully over-simplified example) someone worked the
matter out and decided that the proportion of antitheses
in Johnson to those in Addison was 2:1. From a literary
point of view we should say. ‘Yes Johnson uses far more’,
as we had always seen’. Suppose now that a more pain¬
staking investigator did the job again and discovered that
the count has been wrong, and that the proportion was
really 3:1 or 3:2. What difference would it make? None
at all. We should still say, ‘Yes, Johnson uses far more,
as we had always known.’ Nothing that the literary
judgment can make use of is contributed by these figures.
Some sort of statement about frequency of occurrence
is nevertheless necessary. If the use of a particular stylis¬
tic feature in a novel is being discussed, it makes a
difference whether it appears on every page, about once
in a chapter, or five times in the whole book. But it is
usually enough, as Stephen Ullmann has remarked in an
admirable passage on this matter, to note the significant
recurrence of some linguistic feature without inquiring
into precise numerical details. Insignificant figures, even
though they may in themselves do no particular harm,
tend to vitiate the quality of a whole argument by
implying claims to precision, or to a kind of precision,
that literary inquiry simply does not admit of.
The principal use of numerical criteria, as Ullman goes
on to point out, is in cases where style is examined not
E
56 STYLE AND STYLISTICS

for its own sake, but for some ulterior purpose—to settle
questions of authorship, or to establish the chronology
of an author’s work. Statistical examination of sentence-
lengths, frequency of certain words in the vocabulary,
etc., can by comparison with an author’s known work
go far to determine the authenticity of doubtful works.
These methods always fall short of complete certainty,
but they can pile up an impressive body of inductive
evidence. Such investigations deal only with positively
identifiable characteristics, and commonly with charac¬
teristics which would not be apparent to ordinary direct
observation; they deal with large numbers and large
bodies of material; and there is no substitute for accurate
statistical procedure. Formerly very laborious, many of
these tasks—the making of concordances, word indexes,
etc.—can now be done by computers; and it would be the
merest superstition not to use them. Though it is notable
that one of the best of these studies is also the earliest,
well before the computer period: G. Udney Yule’s statis¬
tical studies in the authorship of The Imitation of Christ.
But the data provided by computers have only a small
and peripheral place in literary study; and it is important
that these machine-minding activities should be kept peri¬
pheral and small. Since they use elaborate apparatus,
cost a great deal of money and can be pursued by per¬
sons of no literary culture whatever, they naturally ac¬
quire great prestige in forward-looking universities.
Older statistical methods were frequently uncertain of
what they were counting and therefore liable to count
wrong. From the latter part of the nineteenth century
metrical tests of a statistical kind were employed in
attempting to date Shakespeare’s plays. The proportion
of rhyme, of light endings and weak endings, were
counted and used as criteria of date, as in a general way
they quite properly can. But the precise nature of the
 METHODS AND PROBLEMS 57
data to be examined was often ill defined, and the statisti¬
cal methods were extremely rough and ready. Even in
such small and positive matters as these, no two ob¬
servers seem to come up with the same results. No doubt
we have reformed this indifferently with us: but there
are some daunting examples of misplaced statistical
fetishism even in quite recent years. Josephine Miles in
her book Eras and Modes in English Poetry sets out to
distinguish period styles by the proportion of verbal to
substantival elements—roughly the proportion of verbs
to nouns and adjectives. Those styles with a high pro¬
portion of verbs are called ‘clausal’, those with a high
proportion of nouns and adjectives are called ‘phrasal’,
and those in between are called ‘balanced’. On this basis
an elaborate historical scheme is evolved, with all the
centuries behaving in a miraculously symmetrical manner.
At the back of the book is a table of the English poets
showing the proportion of adjectives, nouns and verbs
in their work. This is set out in such a way that no easy
comparison can be made between them—Dryden, io;
19; 10; Prior, 12; 15; 9; Rossetti, 12; 15; 8. (It is not clear
whether adjectives have any separate significance from
nouns, but it is the proportion of verbs to nouns and
adjectives taken together than is taken account of in
the argument.) When by some tedious arithmetic we
reduce the figures to a common denominator we find
that the allegedly balanced Dryden has almost precisely
the same proportion of verbs to nouns and adjectives
as the allegedly phrasal Prior; and that the clausal
Rossetti, who ought to have more verbs, has actually a
higher proportion of nouns and adjectives. Keats and
Shelley in a supposedly clausal period have a high pro¬
portion of nouns and adjectives. And so on. The figures
are almost unreadable as they stand, and if reduced to an
intelligible form make nonsense of the general argument.
58 STYLE AND STYLISTICS

The book as a whole is very far from nonsense: there

is a real difference between dominantly verbal and domin¬
antly nominal styles, and many interesting and important
observations flow from this distinction. It reveals more
about kinds than about periods—Pope and Thompson,
coincident in time, show strikingly different proportions;
but then they were writing different kinds of poetry:
and the bare figures tell us almost nothing. Dryden’s
formula, io; 19; 10, is exactly the same as that of Gerard
Manley Hopkins: they are both balanced. Are we any
the wiser for this? Does it show that they had anything
important in common? Not at all. Everything that is
valuable in the book comes in the qualitative discus¬
sions, which are frequently acute and perceptive. The
method as a whole is a classic case of misplaced
quantification.
The first question for literary students who feel tempted
to embark on statistical methods is whether the numeri¬
cal data that are looked for can contribute anything to
the qualitative interpretation of style. If the answer is
‘Yes’, three simple rules might be added: (i) make a clear
identification of what it is that is being counted; (ii) get
the help of someone who can count right, or use a
machine; (iii) make sure that the argument is actually
supported by the figures. The application of these prin¬
ciples would banish most (not all) statistical arguments
from literary studies to some other field, where we need
not follow them.
 4

Some practitioners

Leo Spitzer

In this chapter I shall consider some eminent modern

exponents of style-study in order to see how the general
principles already discussed work out in practice. The
choice of examples is necessarily selective, and with more
space many more might have been included. But I hope to
illustrate the principal trends which ought to be of in¬
terest to the English reader.
I shall begin by discussing the work of Leo Spitzer,
because he has a particularly well-developed method of
style-study and has expounded it in considerable detail.
Spitzer was one of the most distinguished of the German
literary scholars who were expelled by the Nazis, and
after the war found a new home in the United States. His
earlier work on Romance philology was entirely in the
German tradition; after he settled in America some con¬
cessions, though not many, were made to an Anglo-
Saxon audience. He was a man of immensely wide learn¬
ing, not without pedantry; polemical and contentious
in his manner of writing, but with a passion for literature
and a profound and generous understanding of the roots
of Western culture. Himself a Jew and a Viennese, his
chosen field was the literature of the Romance languages.
6o STYLE AND STYLISTICS

and his intellectual allegiance was to a Hellenic-Christian

tradition that he saw as a single entity. I have chosen a
passage from his confession of faith as the epigraph to
this book, both for its intrinsic appropriateness and as a
tribute to his memory. He started, as we have seen, from
a dissatisfaction on the one hand with the old compara¬
tive philology, on the other hand with purely external
and positivist literary history. The first he found to have
no connection with literature at all, while the second
accumulated facts about literature—dates, attributions, in¬
fluences and so forth—without ever penetrating to the
heart of literature itself. These disillusionments might
almost have led to the abandonment of literary study
altogether; but they did not. They led instead to a grow¬
ing conviction of the essential unity of literary and lin¬
guistic study; a belief that the study of language should
lead to an understanding of the greatest achievements of
language—works of literary art; and that works of literary
art can only be understood by a minute study of the
language in which they are realized. This is the theme
of his book Linguistics and Literary History, for the
English reader the best introduction to his work.
Spitzer remains always a philologist, in the extended
German sense of the word; that is to say, a literary
scholar, with equal emphasis on both members of the
phrase—literature and scholarship. He had an equal con¬
tempt for dilettante or impressionist literary study and
for the kind of scholarship that accumulates facts with¬
out adding anything to literary understanding. Although
his work was extremely wide-ranging and his conclu¬
sions often more speculative and transcendental than is
congenial to the English mind, he insisted strongly on the
relative autonomy of literary studies. He disliked the
tendency to disregard the philological character of the
discipline of literary history, which is concerned with
 SOME PRACTITIONERS 6l

ideas couched in linguistic and literary form, not with

ideas in themselves (this is the field of the history of
philosophy). He distrusted the self-assurance with which
students of literature undertake to treat complex sub¬
jects of a philosophic, political or economic nature and
so turn their proper field into ‘the gay sporting-ground of
incompetence’. ‘Only in the linguistic literary field,’ he
added, ‘are we philologians competent qua scholars.’ He
even remained, to a large extent, a philologist in the
narrower sense: much of his work was on historical
semantics, and he was always anxious to start his literary
inquiries from linguistic and etymological observations,
even if their literary relevance seems on the face of it very
slight.
The essential of his method is what he calls the philo¬
logical circle—a method for which he adduces an elabor¬
ate genealogy in German hermeneutics that need not
concern us here. The procedure is to argue from an
observed detail to the central core of a work of art, and
then to proceed outward from the centre in search of
further confirmatory detail. The process can then be re¬
peated as often as necessary until the limits of under¬
standing have been reached. Two features in this method
are especially to be observed. In the first place, it is
opposed to the mere linear accumulation of observations,
in the vague hope that some pattern will finally emerge
from them. Secondly, the process is intuitive at two
points. The initial observation is a spontaneous intuitive
insight; it cannot be manufactured or enforced by a sys¬
tem. Some authentic connection between the observer and
the work of art must establish itself spontaneously. If a
particular student finds that this connection is absent,
no amount of system or training will help him; he had
better give up literature and do something else. And
again, the passage from the observed peripheral fact to
62 STYLE AND STYLISTICS

the central core is at first an intuitive leap, a hypothesis.

A peculiar grammatical detail suggests a certain central
orientation of the author’s mind; and this is an informed
surmise rather than a demonstration. We are rescued,
however, from mere impressionism and subjectivity by
returning outward from this central point and finding
other details which will confirm and, if possible, refine the
original hypothesis. In this way, a body of evidence is
built up which will provide a view of the author’s central
creative attitude, or as the informing principle of a single
work.
This is a good deal easier to explain than it is to illus¬
trate in practice, for it is a description of a general
approach rather than a rigid series of operations, and in
an actual literary inquiry the paradigm is not often ex¬
hibited completely. Spitzer gives a simple illustration in
the work of an elementary student of Latin construing
a Ciceronian period: he can only do it by moving
from a detail (the form of a particular word) to an
assumed whole (the structure of the whole sentence) and
then back to further details in order to confirm and build
up the picture of the whole. I have little doubt that this
represents truly both the aim and the method of literary
inquiry. The aim is understanding, not new information:
and the method is a continual to-and-fro movement be¬
tween observable details and the hidden centre which we
wish to reach, possess and comprehend. The end is an
act of contemplation rather than the safe delivery of a
package of facts.
It is not necessary that the starting-point should be a
linguistic observation: it may be an observation about
imagery or plot. Spitzer says somewhere that it was his
own philological training that led him to adhere to the
linguistic approach. As a matter of fact, he does not always
adhere to it; but it remains his favoured method and he
 SOME PRACTITIONERS 63

does seem to feel that the linguistic detail is the irreducible

demonstrable fact which affords the securest anchorage
for later speculations. And this conforms also to his con¬
viction of the essentially philological character of literary
study—that it cannot properly be short-circuited into
history of ideas, psychology or a loose kind of philoso¬
phizing. As for value judgments, comparative or absolute,
he is chary of them. For him, the primary task of the
student is the critique des beautes, involving the prior
acceptance of the work of art as a value in itself, to be
apprehended in its totality:
Indeed, any explication de texte, any philological study,
must start with a critique des beautes, with the assumption
on our part of the perfection of the work to be studied
and with an entire willingness to sympathy; it must be an
apologia, a theodicy in a nutshell.... A criticism which
insists on faults is justifiable only after the purpose of the
author has been thoroughly understood and followed up
in detail. The glibness with which critics, especially great
German critics (Lessing, Schiller and Schlegel), have
slandered French classical drama is only to be explained
on the basis of premature judgments drawn from a quite
extraneous comparison with Shakespeare. (Linguistics and
Literary History, 128).
It is relevant to ask what kind of understanding
Spitzer’s work seeks to arrive at. In his earlier writing he
tended to look for the meaning of the work of art in
the mind of its author; his interpretation was largely
a psychological one. It was the guiding spirit, the central
spring of the author’s imagination, that the observation
of linguistic details was supposed to lead to; and there
was a distinct and acknowledged Freudian influence on
his investigations. Later, perhaps under the influence
(never acknowledged) of a more modern and more sophis¬
ticated Anglo-American criticism, he came to distinguish
6\ STYLE AND STYLISTICS

more clearly between the creative imagination and the

historic personality of the author, and between the work
of art and the mind of its creator. His object then became
rather to explain the inner being of the work itself; but
for him this always remained quite clearly what the
author meant to put into it. He was severe about the
incorporation into hermeneutics of later accretions of
meaning and was quite willing to talk in terms of the
author’s purpose. I never heard him speak of those op¬
ponents of the ‘intentional fallacy’ whose views were
much in the air in his later days: but the whole tenor,
expressed and unexpressed, of his work was to think of
the author’s intention as something specific, definite and,
in principle, discoverable. To discover it was indeed the
business of stylistics.
Spitzer’s method has been the object of a great deal
of criticism, and his philological circle was frequently
denounced as a vicious one. The general tendency can
be indicated by a quotation from Professor Harold
Cherniss—though this passage was actually directed at
scholars of the Stefan George school rather than at
Spitzer himself:
The intuition which discovers in the writings of an author
the natural law and inward form of his personality, is proof
against all objections, logical and philological; but while
one must admit that a certain native insight, call it direct
intelligence or intuition, as you please, is required for
understanding any text, it is all the same a vicious circle
to intuit the nature of the author’s personality from his
writings, and then to interpret those writings in accord¬
ance with the inner necessity of that intuited personality
(ibid., 34).

Spitzer’s reply was to disembarrass the idea of intuition

from any transcendental or mystical overtones: intui¬
tion, as he understands it, is simply a gifted and informed
 SOME PRACTITIONERS 65

perception. He then goes on to remark that though the

method is indeed circular the circle is not vicious, but
benign. The intuitive leap from the surface detail to the
assumed centre is checked by a return to the surface and
the observation of further details; and so on until the
whole has been grasped. It is this repeated series of
movements that rescue the process from mere subjective
and unverifiable private judgment.
A more serious criticism to my mind is that the link
between the linguistic starting-point and the literary in¬
sight is often a tenuous one. In some of his writing the
preliminary linguistic work looks too much like a series
of conjurer’s passes, designed to mislead the spectator
about what is actually going on. After we have been
suitably dazzled by a display of linguistic sleight of hand,
a black cloth is held up in front of the top-hat, and
when it is withdrawn an admirable but wholly irrelevant
rabbit is pulled out. And this, I think, points to one
of the cardinal difficulties of stylistics. Stylistics claims—
must claim—that the understanding of a work of literary
art is continuous with the understanding of its language,
and that the close, even technical, study of language is the
only sure way to literary understanding. Many critics,
while paying lip-service to this view, actually tend to
take short-cuts, and proceed to interpret works of litera¬
ture in moral or ideological terms, without considering
linguistic texture in any detail at all. And it sometimes
seems that Spitzer’s conclusions are those of the scholarly
man of letters, that they depend more on wide literary
experience, a large range of comparisons, deep know¬
ledge of cultural history, rather than on any expertise
that could be called philological in the narrower sense.
It is a not uncommon dilemma; stylistics stays firmly
within its own terms of reference, and is condemned
66 STYLE AND STYLISTICS

to triviality: or it proceeds to ideas of great scope and

generality by deserting strictly stylistic methods.
But much of Spitzer’s work bears witness that this need
not be so. The brilliant study of the style of Diderot, for
example, begins from some observations on rhythm—a
breathless, accelerating, almost feverish rhythm to be
found in a passage of Diderot’s writing, almost as though
it were patterned on the rhythm of the sexual act. This
observation is fused with a consideration of the subject-
matter of the passage; other passages of a different tenor
are brought into the argument until a vision of Diderot’s
imaginative activity as a whole is reached, quite transcend¬
ing common critical objections to its fragmentary and
incoherent quality. In the study of Don Quixote the
starting-point is the variability and uncertainty about the
naming of the characters. The various names for a single
character are then seen as reflections of various points
of view from which he can be regarded. From this we
pass by links which I confess are sometimes not easy
to follow, to the prevalence of varied, shifting and un¬
certain viewpoints in the whole work: so that in the
end the essence of the book is seen to be characterized
by this perspectivism.
It is sometimes the case that the stylistic observations
lead to a conclusion not very different from that reached
impressionalistically by common literary opinion. Spitzer
is prepared for this:

Sometimes it may happen that this etymology leads simply

to a characterization of the author that has long been
accepted by literary historians (who have not needed
apparently to follow the winding path I chose), and which
can be summed up in a phrase which smacks of a college
handbook. But to make our way to an old truth is not only
to enrich our own understanding: it produces inevitably
 SOME PRACTITIONERS 67

new evidence of objective value for this truth—which is

thereby renewed (ibid., 38).

Sometimes the initial observation is not of a peculiarly

linguistic kind. The magnificent interpretation of a pas¬
sage from Claudel’s Cinq Grandes Odes begins by an¬
alysing the structure of what at first seems a rhapsodic
poem, singularly resistant to analysis. It is true that
Spitzer finds the clue to the structure in the punctuation
provided by the word grand six times repeated at critical
points. But this verbal clue is of the slightest: and what
we arrive at in the end is a superb and deeply felt in¬
terpretation of the five odes as a whole, seen as a single
great poem. But here the difficulty was to find a way into
this daunting and almost impenetrable piece. Spitzer’s
claim is that the slightest linguistic observation may
serve as a point of entry. It does not matter what it is as
long as the observation is genuine and original. No sys¬
tem or predetermined scheme can serve as a guide. It is
here, I believe, that the freshness and liberating power
of Spitzer’s recommendation is to be found. The student
or aspiring interpreter is free to make his way into the
work by any route that native wit or temperamental
inclination may offer. His duty is then to pursue his
course with rigour, pertinacity and complete submission,
not to any predetermined code of rules, but to the work
of art as it presents itself, secure in the belief that the
work is an organic whole and that the extended study of
any part of it must lead to an understanding of the
whole.

It is my firm belief, corroborated by the experience of

many exercises practised in seminars with my students,
when I chose to start from any particular point suggested
by one of the group, that any one good observation will,
when sufficiently deepened, infallibly lead to the centre of
68 STYLE AND STYLISTICS

the work of art. There are no preferential vantage-points

(such as the ideas, the structure of the poem, etc.) with
which we are obliged to start: any well-observed item can
become a vantage-point and however arbitrarily chosen
must, if rightly developed, ultimately lose its arbitrariness
(ibid., 198).
And the end of the whole process? Here Spitzer’s
claims are essentially modest. He is not out to build a
critical system or expound a philosophy, but simply to
understand the work on its own terms—an understanding
that could ideally be arrived at without any exegesis at
all. He concludes the Claudel study by saying:

Stylistics as I conceive it is an exclusively auxiliary science.

Just as, according to Pascal, for him who knows truth no
style, no art de persuader is needed, so stylistics must
abdicate once the true nature of the work of art has been
perceived. A study of the kind we have attempted could
have been made entirely unnecessary from the start by a
simple recital of the poem, if the performer were able, by
various pauses and intonations, to suggest the main motifs
we have taken pains to distinguish, and to show within the
crystalline ball of the work of art the play of the conflict¬
ing forces in the equilibrium which Claudel has been able
to establish.

Erich Auerbach
Next to Spitzer it is appropriate to consider the work of
Erich Auerbach. His formation and personal history were
closely similar to Spitzer’s. After already distinguished
careers in Germany, both worked in Turkish universities
during the Nazi period, and after the war both settled
in the United States. In Auerbach’s case it is largely this
period of relative isolation in Turkey that brings him
within the scope of our study. Also a Romance philo¬
logist by vocation, his orientation was on the whole
 SOME PRACTITIONERS 69

more historical than Spitzer’s, and he was less exclusively

devoted to explication de texte. His magisterial essay
‘Figura’, for example, begins as a semantic investigation
into the significance of the word, and develops into a
profound study of Christian typology, its motives and
its procedures. But it is not tied to any particular literary
text. Working in Istanbul with limited library facilities,
he was increasingly forced back to concentrate on the
texts themselves; and with this stimulus he began to put
together, under the domination of a single idea, a number
of studies of individual texts, of widely differing periods,
varying from Homer and the Old Testament to the Gon-
courts and Virginia Woolf. The result of this is his great
book Mimesis, probably the most profound, most learned,
and most wide-ranging work of style-study that has ever
been written. The dominant idea is given in the sub-title
of the book, ‘The Representation of Reality in Western
Literature’. Its scope is enormous—nothing less than the
various ways in which men’s actual experience, histori¬
cal, social, moral and religious, has been represented in
literary form in all the various phases of Western culture.
It was only the limits imposed by his situation that made
this study possible. As he says himself, ‘it is quite possible
that the book owes its existence to just this lack of a rich
and specialized library. If it had been possible for me to
acquaint myself with all the work that has been done
on so many subjects, I might never have reached the
point of writing.’
As it is, the range of learning is formidable; and it is
hard to believe that the educational situation of post-war
Europe will ever produce anything like it again. But it is
all concentrated on the study of particular passages, all
of them comparatively short. Auerbach has this to say
about his procedure:
The method of textual interpretation gives the interpreter
70 STYLE AND STYLISTICS

a certain leeway. He can choose and emphasize as he

pleases. It must naturally be possible to find what he claims
in the text. My interpretations are no doubt guided by a
specific purpose. Yet this purpose assumed form only as I
went along, playing, as it were, with my texts, and for
long stretches of my way I have been guided only by the
texts themselves. Furthermore, the great majority of the
texts were chosen at random, on the basis of accidental
acquaintance and personal preference, rather than in view
of a definite purpose. Studies of this kind do not deal with
laws, but with trends and tendencies, which cross and com¬
plement each other in the most varied ways. I was by no
means interested merely in presenting what would serve
my purpose in the narrowest sense; on the contrary, it was
my endeavour to accommodate multiplex data and to
make my formulations correspondingly elastic (Mimesis,
49i)-
The result is probably unique in its combination of
the synchronic with the diachronic approach, in doing
full justice to the individual being of individual works,
yet at the same time giving a valid and substantiated pic¬
ture of historical development. The only pity is that so
few scholars are likely to have the equipment to emulate
it. Most style-studies have been, perhaps naturally, studies
of individual works, individual authors or limited ten¬
dencies. Spitzer, for example, seems to have flitted from
flower to flower in the immense meadow of Romance
literature without thinking of tracing a connected route;
and his books are simply assemblages of separate and
separately published papers. Each of the studies that go
to form Mimesis has the density and particularity of an
individual learned article, but the whole is directed by a
single purpose, and from the great mass of varied critical
scholarship an unforced but coherent pattern can be
seen to emerge. Furthermore, the connection between
the linguistic observations—remarks on vocabulary or
 SOME PRACTITIONERS 71

syntax—and the wider considerations to which they lead

is always clear. Auerbach is ultimately a historian of
culture, and he arrives in the end at conclusions of great
scope and generality. But he always founds these wider
considerations securely on a linguistic base. If, as I have
suggested, the methods of stylistics are sometimes open
to question, we have at least in Mimesis a triumphant
vindication of their proper use.
It would be impossible to summarize the content of
this book, as it contains twenty separate essays, on
topics ranging over a span of 3,000 years. All that can
be done here is to indicate the procedure in one or two
of the separate studies, and the main general argument
that emerges from the whole. The first essay contrasts a
passage of Homer—the recognition of Odysseus by
Euryclea, the old nurse—with a story from the Old
Testament, the sacrifice of Isaac. In the Homeric narrative
everything is in the foreground, uniformly illuminated;
the place, the setting, the identity of the characters
clearly indicated. When a passage of reminiscence is
introduced, as it is in the lines describing how Odysseus
came by his scar, this in its turn becomes foreground,
equally clear, present and objective. With this Auerbach
contrasts the biblical narrative, beginning with a dialogue
between Abraham and God. It is located nowhere; God
has no physical being. We know nothing of his purpose;
He has not, like Zeus, discussed it in Council with the
other gods. Yet a sense of purpose and significance, that
yet remains obscure and mysterious, enfolds the whole
narrative. This is only the starting-point of a full and
elaborate differentiation; but it is also the starting-point
for a contrast that runs through the whole book:

The two styles, in their opposition, represent basic types:

on the one hand fully externalized description, uniform
F
72 STYLE AND STYLISTICS

illumination, uninterrupted connection, free expression, all

events in the foreground, displaying unmistakable mean¬
ing, few elements of historical development and of psy¬
chological perspective; on the other hand, certain parts
brought into high relief, others left obscure, abruptness,
suggestive influence of the unexpressed, background
quality, multiplicity of meaning and the need for inter¬
pretation, universal historical claims, development of the
concepts of the historically becoming, and preoccupation
with the problematic (ibid., 19).

In succeeding essays, later styles are discussed—•

Petronius, Tacitus, Ammianus Marcellinus—to illustrate
how a formal and elaborate rhetoric is employed on
increasingly irrational, crude and violent material, leading
up to the crisis of style in which this tension finally
culminated. One of the central points of the book is the
strict separation of styles in classical literature: the high
style for noble subjects, the low style for comic and
vulgar material. Where is it that common men, speaking
in common language, without the graces of rhetoric, are
nevertheless represented as engaging in affairs of the
greatest weight, seriousness and moment? Not anywhere
in classical literature: it is in the Bible, especially in the
Gospels, that we find this mixing of styles, always hetero¬
dox from the classical point of view. The reverberations
of this contrast are clearly felt throughout the whole of
Western literature and Western culture; and this must
serve as a brief illustration of the way in which a
stylistic observation can lead to cultural and historical
considerations of a fundamental kind.
The classical, patristic and medieval periods, and the
intellectual links which bind them together, are probably
Auerbach’s preferred field; but he passes on, via studies
of Rabelais, Montaigne, Cervantes, Saint-Simon and many
others, to consider modern realism. This he finds to be a
 SOME PRACTITIONERS 73

distinct new development, different from the Christian

realism of the Bible and the Middle Ages. It is the self-
consciousness and rapid acceleration of social change
inaugurated by the French Revolution that gives rise to
modern realism; and Auerbach traces its varieties in
Stendhal, Balzac, Flaubert and the Goncourts. He con¬
cludes with a brief glance at the disintegration of tradi¬
tional methods of representation in modern literature,
illustrated from a passage of Virginia Woolf; but it is
only a brief glance; and he does not disguise his convic¬
tion that the history of Western culture, as he and his
generation have understood it, is fast drawing to an end.
He may be right. However that may be, it is notable
that neither he nor Spitzer has had much dealing with
distinctively modern literature, nor with literature in
English. Spitzer can be positively wrong in interpreting
even traditional English literature, as he is in his reading
of Donne’s ‘Ecstasy’; or naively eager to make a gesture
of accommodation to a new world, as in his essay on
‘American Advertising as Popular Art’. Auerbach, so
vigorously at home in earlier centuries, seems to work
under a slowly darkening elegiac cloud as he approaches
modern times. It is possible, therefore, to admire the
methods of these two writers, to be awed by their
scholarship, and still to feel that there is plenty left to do
on similar lines by those who could not hope to compete
with them on their own ground.

Damaso Alonso

The work of Damaso Alonso is of a rather different

character from that of Spitzer and Auerbach, for it has
been confined to Spanish poetry, particularly the poetry
of the Siglo d’Oro. Of this my ignorance is almost com¬
plete, so what I have to say about his writing must be
74 STYLE AND STYLISTICS

partial and external. But his book, Toesia Espanol, has

its sub-title, ‘Ensayo de Metodos y Limitos Estilisticos’,
and the theoretical considerations interspersed among the
particular studies are of great general interest. I should
like to believe that they have not been without effect on
my own thinking, such as it is, since I first made their
acquaintance nearly twenty years ago; and they should
certainly be better known to English readers. Many links
will be observed with the methods of Saussure and
Bally, against which Alonso’s work is a reaction; and with
the methods of Spitzer, to which it shows some
similarities.
Alonso begins by accepting a classification from
Saussure’s terminology, but rejecting its clear-cut opposi¬
tions. Saussure calls the complete linguistic phenomenon
a ‘sign’; and the sign unites two elements, the signified
and the signifies which he defines as a concept and an
acoustic image respectively. Alonso says that an insalvable
abismo separates him from this way of looking at the
matter. For Saussure the signified is a concept; the signi-
fiers are simply conveyers or transmitters of concepts.
This, Alonso says, is an idea ‘as aseptic as it is poverty-
stricken’, far from the three-dimensional reality of
language. In fact, the signifiers do not transmit concepts,
but delicate functional complexes, including associations,
synaesthesias, affective charges. We cannot consider the
signified as merely conceptual. Indeed, we cannot isolate
it. When a mother calls ‘Jackie!’ to her child she may
do so affectionately, or angrily, or apprehensively because
he is about to run under a lorry. In these cases what is
the essential signified—the tone, the intensity, the speed,
or what? In a line of verse the stress may be a signifier,
the value of a single vowel sound may be one. The
whole line with its rhythm, its accents, its vowel pattern
and its conceptual contents is a complex signifier which
 SOME PRACTITIONERS 75
arouses in us a complex signified. The signified is not
essentially a concept; it is an intuition which produces
an immediate modification of some elements, possibly
every element, of our psyche. One of the elements will be
conceptual, but others will not. Besides the departure
from Saussure, it will be observed how clearly Alonso
separates himself from Bally, with his basic distinction
between the logical and affective aspects of language.
For stylistic purposes Alonso would substitute a dis¬
tinction between outer form and inner form. Outer form
is the relation of signifier to signified from the point of
view of the signifier. The inner form is the same relation
seen from the point of view of the signified. The study
of outer form is the easier, for it starts from concrete
phonetic realities. The study of the inner form is more
difficult, for, as Alonso sees it, it deals with the psy¬
chology of the moment of creation, the moment of
internal formation of the signified, and its immediate
embodiment in a signifier. The stylistics of the future,
he says, if there is to be any, will tend to an equal
emphasis on both perspectives, outer and inner. It is
easy to see that this has something in common with
Spitzer’s method of working from an exterior linguistic
detail to the internal form of a whole work.
For Alonso there are three modes of understanding
literary work, marked by an increasing degree of pre¬
cision. The first is the understanding of the common
reader, who seeks neither to analyse nor to exteriorize his
impressions. It is a totalizing intuition, which forms itself
in the process of reading and comes to reproduce the
totalizing intuition which gave rise to the work—that
is, the intuition of the author. Like Spitzer, Alonso is quite
clear that the object of literary study is to recover the
author’s original purpose, though he perhaps expresses
himself less in terms of conscious purpose. This ‘reader’s
76 STYLE AND STYLISTICS

understanding’ is immediate and intuitive, and it is the

purer the fewer extraneous elements have come between
reader and author. This means, I take it, that pure literary
understanding is independent of external information—
biographical, bibliographical, historical, etc. And this
‘reader’s understanding’ is intransitive; it does not at¬
tempt to express itself. It is a simple relation between
reader and work; its primary object is delight, and in
delight it terminates. It is, of course, the indispensable
foundation of all other kinds of literary understanding.
The second degree of understanding is that of the critic.
There is a type of man in whom the qualities of the reader
are exceptionally developed: his receptive capacity is
both more intense and more extended than the ordinary.
This exceptional creature is the critic. Ideally the range
and intensity of his reactions should correspond with
those of the poets themselves; and reading should incite
in him profound and luminous intuitions, comprehending
the work in its totality. Above all, the critic is a receptive
apparatus of the greatest delicacy and amplitude.
But the critic has also, as a natural tendency of his
personality, an expressive activity. He must communic¬
ate, compactly and rapidly, images of the intuitions he
has received. This is his mission. He transmits his re¬
actions, but they are not in themselves a problem for
him. In general, it does not interest him to establish how
or why these reactions have been produced. He is con¬
tent with some general classificatory scheme that will
enable him to communicate his intuitions to a reader of
the same poem. Strictly speaking, these aesthetic intuitions
are ineffable; all the same, the critic does his best to
express them in a creative and poetic fashion. Criticism
is an art.
It will be evident from this that Alonso severely limits
the positive and scientific pretentions of criticism, and
 SOME PRACTITIONERS 77
is more than willing to find room for the impressionist
criticism that our century has often been inclined to
condemn. What he does not find room for, any more
than Spitzer does, is conventional literary history. There
are true literary works and false; the only true literary
works are those which have something to say and say it
directly to the heart of man. The true literary work is
ahistoric and cannot be the object of history, but there
is a monotonous mass of secondary works which say
nothing to the heart or mind of man; these form a vast
necropolis full of dead imitations. And this is the object
of what is commonly called literary history. But it is not
really literary history; it is the history of literary culture,
which is a different thing. Criticism is not concerned with
it. The discrimination of the true from the false literary
work has always been considered the principal function of
criticism.
In practice a distinction must be made between
criticism of the literature of the past and criticism of
contemporary literature. The criticism of past literature
is a continuous collective process. Sudden devaluations
and denigrations are an impossibility. Humanity will not
suddenly abandon the estimate of centuries for an upstart
fashion. The collective valuation has become itself an
intuition for humanity, and no individual judgment can
overturn it. The first function of criticism is not to re¬
direct literary judgment, but to remove the rust of cen¬
turies, to explain changes of language, meaning, customs,
allusions, etc. And for this the intuitive insight of the
critic is not enough; he needs also a reasonable erudition.
The criticism of contemporaries is a different matter:
it is always and necessarily an unreliable affair, especially
criticism of the work of an immediately preceding genera¬
tion. However, from an infinite rosary of judgments, all
78 STYLE AND STYLISTICS

inexact, that men make on their immediate circumstances,

Dios integra su verdad, God composes his truth, the single
criticism on which no one can cast doubt.
But there is a third stage in the understanding of
literary works which goes beyond that of the critic. If
we begin to ask, ‘Why does this poem, or verse, move
me? What is it, whence does it originate this emotion
that passes through me, whence does it proceed?’ we
begin to pass to this third kind of understanding. Criti¬
cism answers these questions only in the vaguest and most
general fashion because the problem does not really in¬
terest it: it is enough for the critic to make a rapid
evaluative survey of his intuitions. His conclusions are
intuitive and unscientific.
But we can at least begin to go farther, to our third
stage, and consider the possibility of a scientific under¬
standing of artistic facts. This scientific approach is stylis¬
tics. Typological classification solves nothing: it is the
individual work that must be examined; and the final
act of apprehension must always be an intuition. But
stylistics offers the possibility of a precise and demon¬
strable analysis.
The poem presents itself to us on the one hand as a
temporal succession of sounds—the signifiers; on the
other as a spiritual content—the signified. The signified
is a modification of our spiritual life, difficult to penetrate
or to analyse. The total signifier A is linked to the total
signified B by numerous partial links:

A Qg Qg Qg ... Qn

III I
B bi bg bg . . . bn

The first function of stylistics is to investigate the rela¬

tion between the two wholes by investigating the relation
between all the partial elements. The complete relation
 SOME PRACTITIONERS 79
will be arrived at by integrating all these partial relations.
These separate elements are very numerous—far too
numerous for complete study. A selection must be made
of those which are most relevant and most revealing.
There is no cut-and-dried method available here; the selec¬
tion can only be intuitive. This is evidently the same as
Spitzer’s principle of the personal intuitive observation as
the point of entry into the work, and it shows the same
objection to uniform and mechanical methods.
We begin stylistic investigation with the outer form—
the signifier—because this is the concrete fact presented
to our sense-perception. We consider the inner form—
the signified—as a complex of elements: conceptual,
affective, synaesthetic and image-producing; and we assert
that the same complexity must exist in the signifier. The
passage from the outer to the inner form is difficult, but
it is precisely the object of stylistics to make it. Like
Spitzer, Alonso sees the understanding to be attained
partly in psychological terms. Indeed, he says that the
literary investigator must double his role with that of the
psychologist. He has to classify and study all the elements
that have touched the spirit of the poet, all the elements
which may have determined a certain reaction in him.
We may add, parenthetically, that it is not easy to see
how this can be done without passing outside the stylistic
realm. However, Alonso insists that the aim of stylistics
is to establish a rigorous and concrete link between
signifier and signified, and so reach a full and accurate
understanding of the total sign—that is to say, the total
literary existence of the work. Stylistics so considered is
the science of literature, and it is the only possible route
to a true philosophy of literature. At present this science
is limited and immature. All sorts of mixtures and com¬
binations of criticism and stylistics are possible in prac¬
tice and the one does not supersede the other. No one can
8o STYLE AND STYLISTICS

be an investigator in stylistics who has not first been a

passionate reader and secondly a devoted critic. We must
always remember the three stages—the readers’ know¬
ledge which leads to an intuitive pleasure; the critics’
knowledge, which has a pedagogical intention, and stylistic
knowledge which leads to the solution of a problem. And
for the final stage, the first two are always necessary.

I. A. Richards

When we pass from the work of the Continental Euro¬

pean scholars to work in English, certain distinguishing
features become very evident. The first, and to my mind
the most important, is a sense of ecumenical range in
this Continental scholarship, even when it is dealing with
a particular national literature. Auerbach sees the whole
of Western culture from the Old Testament to the
threshold of modern times as a single object of con¬
templation; Spitzer says that philologists are in essence
theologians, studying an eternal truth; Damaso Alonso
says that out of a series of partial literary judgments, all
inexact, Dios integra su verdad. We should perhaps not
make too much of the theological reference in the last
two utterances; what is striking is the sense of a whole
culture with a single and indisputable authority, trans¬
cending temporary fashion and individual opinion. This
seems to me a more informed and authentic commit¬
ment than the appeal to tradition common in the Anglo-
American writing of the thirties, which was often used
merely polemically in support of quite untraditional
attitudes. And in the work of I. A. Richards and his suc¬
cessors the sense of long-established authority is wholly
absent. The appeal is to a science that is supposed to be
quite novel, to semantics or to a new experimental psy-
 SOME PRACTITIONERS 8l

chology, an enlightenment that dawned only yesterday,

sometimes to mere bright ideas that serve an immediate
purpose—to solve a local problem, to cast a new light on a
particular text, to dazzle the sophomores, quite without
integration into any comprehensive scheme of know¬
ledge. I do not want to say this, and for myself I should
wish to work for a state of affairs in which for England
it would no longer be true, but as far as literary study is
concerned the belief of General de Gaulle seems to be
only too well founded—les Anglo-Saxons do seem to
stand outside Europe.
In this matter of range and authority the advantage
seems clearly to be with Europe. In another respect I
am less certain. Spitzer and Alonso both see the ultimate
goal of their research and the ultimate authority for their
judgment in the intention of the author, which they
regard as something final and, in principle, discoverable.
Anglo-American work, whether criticism or stylistics in
Damaso Alonso’s sense, tends to be more sceptical about
this possibility, and tends to allow far more for the
accretion of meaning to a work of art with the passage
of the centuries. It could justify this attitude by saying
that any later accretion of meaning must in some sense
have been latent in the original. The Anglo-American
procedure would also make far more concessions to his¬
torical relativism. Whatever is true about a work of art
is true only for the state of culture and society that
sees it to be so. And this tendency has been accentuated by
the Empsonian conception of ambiguity, of multiple mean¬
ing, even though the concept was not arrived at via histori¬
cal considerations, but by a strictly stylistic, internal
examination of the way that poetry works. The ideas of
multiple meaning and accretion of meaning obviously give
room for subjectivism, fantasy and eccentricity; but this
may be a justifiable risk. Great liberty of interpretation
82 STYLE AND STYLISTICS

may in the end do more justice to the depth and multi¬

fariousness of complex works of art.
Finally, the distinction between criticism and stylistics,
made explicitly by Damaso Alonso and at least implied
by Spitzer, is much less clear in English writing. The
critics have made considerable use of stylistic analysis;
and literary stylistics with a formal linguistic base is
hardly found among writing in English. With these pre¬
liminaries, we can turn to the work of I. A. Richards.
A large part of Richards’s writing does not concern us
here; his Principles of Literary Criticism and his writing
on literary theory in general are outside our field. The
important work is Practical Criticism, a book which has
the distinction of setting stylistics on a new footing al¬
most without reference to hereditary literary disciplines.
Its results are not repugnant to the consensus of former
literary judgment; they would be self-condemned if they
were. But they have been arrived at by new methods and
were announced in a new tone. Indeed, the tone of
Practical Criticism often contrives to suggest that literary
interpretation had never seriously been undertaken be¬
fore, or that all previous attempts at it were quite
negligible. This is not true; but it is true that Richards’s
work is just about contemporary with new procedures
in stylistics in Europe, and so it is part of a general phase
of innovation. Richards’s Principles first appeared in 1924,
Practical Criticism in 1927; Bally’s Traite de stylistique
frangaise first appeared in 1920, Le langage et la vie
in 1925; and in some respects, though by very different
methods, they were pursuing the same ends. Both assumed
that the logical and grammatical aspects of language had
received a good deal of study already and both wished
to turn their attention to the affective element of linguis¬
tic expression; but Bally approached this from orthodox
linguistics, Richards from semantics and psychology—as
 SOME PRACTITIONERS 83

such subjects were understood at that now rather distant

period.
They were also addressing a different audience.
Saussure, Bally and the Continental writers on stylistics
generally were addressing their peers, scholars in language
and literature. They did not feel it necessary to make
concessions to ignorance or inexperience. Richards made
his experiment in controlled reading with a class largely
composed of undergraduates, and seems to be addressing
his book to a general audience. He says he wrote Vractical
Criticism ‘for those who were interested in the con¬
temporary state of culture, whether as critics, as philoso¬
phers, as teachers, as psychologists or merely as curious
persons’. The wide spread of his potential readers means
that he cannot presume in them any real body of shared
experience or any high degree of literary sophistication.
This sometimes gives his writing an air of naivete, of
which it would not be hard to find examples. However,
I do not think we should be too eager to search them
out, for they are perhaps necessary consequences of
another endeavour, immensely valuable in itself and
quite unrepresented in Continental writing. That is the
endeavour to find out from the beginning what the pro¬
cess of reading and interpreting poetry is really like; what
actually goes on in the reader’s mind; what are the ob¬
stacles to better understanding. This involves asking a
number of awkward questions and exposing a number
of unwelcome truths; and even if the later consequences
have not been as uniformly salutary as might have been
hoped, we have every reason to be grateful that the basic
elementary difficulties of literary interpretation have been
so fully exposed.
The procedure of Practical Criticism is by now familiar
to everyone who is concerned with these matters. A
number of short poems or extracts from longer poems
84 STYLE AND STYLISTICS

were presented to the members of a large university class,

without any indication of authorship or date, and they
were asked to comment on them. The comments (which
Richards, for some impenetrable reason, calls ‘protocols’)
were collected, classified and analysed; and from this body
of information an extended series of deductions was
made about actual reading habits, about the methods and
criteria that readers actually employed, and as a conse¬
quence about the state of literary education. As everyone
knows, the results were disturbing. Incomprehension, cap¬
ricious and eccentric judgment, and sheer, helpless be¬
wilderment were found to be frequent. But I shall not
enlarge on this, as it is a matter of general educational
history rather than of stylistics. It must be realized from
the start, however, that the conditions of the experiment
were severely limited. Historical knowledge, the founda¬
tion of traditional opinion, the sense of an accepted
literary tradition were as far as possible excluded by the
dateless and anonymous character of the extracts pre¬
sented. This was a necessary condition for the particular
inquiry in hand: it was not a recommendation that such
knowledge should be excluded for ever, or an assertion of
its irrelevance. It is necessary to emphasize this, for the
wide extension of ‘practical criticism’ methods as an
educational procedure has led to the absurd idea that there
is some special virtue in approaching literature divorced
from its historical setting, that no knowledge of this is
necessary, or, in more extreme cases, that there isn’t
really anything to know. Of course, Richards did not
intend anything of this kind. Both Spitzer and Alonso at
a later date have emphasized that the contact between
the reader and the work must be effected by a kind of
intuitive click; it is not a matter of imported external
knowledge. Richards, much earlier, is setting himself to
inquire how this intuitive click takes place. What are the
 SOME PRACTITIONERS 85

connections that form themselves between the reader and

the work, and what ought they to be?
Many of the obstacles to true interpretation disclosed
by Richards’ experiment were below the level of literary
skill on which our Continental scholars were operating.
Incomprehension of the plain sense, misjudgment of the
feeling, half-baked and misinformed notions about metrics
and poetical kinds, all discovered to be widespread, are
part of educational pathology. What is of interest for
stylistics is the ensuing analysis of literary meaning. Like
Bally in Le langage et la vie and Damaso Alonso in
Voesia Espahola, Richards insists that meaning cannot
be reduced to conceptual meaning; the affective and ex¬
pressive sides of language are also a part of meaning,
and a part that has been damagingly neglected. No appar¬
atus for examining it has ever been provided, and this
Richards aims to supply.
He divides meaning into four aspects, which he calls
sense, feeling, tone and intention. Sense is conceptual
meaning; Bally’s logical or intellectual aspect of lan¬
guage. ‘We use words to direct our hearers’ attention
upon some state of affairs, to present to them some
items for consideration, to excite in them some thought
about these items.’ The three other kinds of meaning
represent a classification of Bally’s affective and expres¬
sive aspects of language; and they are a clear improve¬
ment on Bally’s very general terminology. Feeling is the
emotional attitude towards the subject presented by
sense. Tone directs our attention differently; it is the
attitude not towards the subject, but towards the person
addressed, actually or in imagination. Intention is the
purpose, conscious or unconscious, of the whole utter¬
ance, the effect that the writer intended to promote. Sense
and feeling probably need little further comment. Of tone
we may say that it is obviously powerful. Richards re-
86 STYLE AND STYLISTICS

marks acutely that Gray’s Elegy is largely a triumph of

exquisitely adjusted tone. Intention presents a difficulty
in that it can often be subsumed under one of the other
three functions. The intention of a scientific paper is al¬
most entirely a matter of sense; the intention of a love-
letter is to convey feeling. But in many cases, particularly
in literature, we need to consider intention as a distinct
controlling factor. There is an obvious class of external in¬
tentions, such as that of the political speech that is made
for the purpose of getting a certain party elected. Any
interpretation of such a speech must take this controlling
purpose into account. And in literature such extra¬
curricular intentions are not absent. But characteristically
in literature we need to consider intention in another
way: Is the passage under consideration addressed directly
from author to reader, or is it a dramatic utterance
spoken by a fictional character with whom the author
does not identify himself? Or, to introduce examples
that Richards does not, is it governed by certain over¬
riding conventions, such as the decorum of neo-classical
tragedy, or the traditional praise of a lady in courtly
love poetry? Obviously, both the choice and ordering
of material is profoundly affected by such intentions as
these, and obviously any stylistic analysis will go astray
if they are not taken into account.
Literary meaning is characteristically a fusion of these
four functions. By abstraction they can be considered
separately, and in the process of stylistic inquiry they
must be so considered. But the total meaning is the inte¬
gration into a unity of all four. By this piece of analysis-
and-synthesis, Richards has provided a simple and
workable instrument of stylistic inquiry—so simple that
it can be employed by students of literature at a very
elementary stage, so fundamental that it cannot be neg¬
lected by even the most complex and sophisticated
 SOME PRACTITIONERS 87

literary interpretation. It can be objected that the analysis

does not go far enough. Richards shows a strong tendency
to set conceptual meaning on one side, and to bundle
all other sorts of meaning together as emotive. This is
more evident in the Principles than in Practical Criticism,
but it is a pervasive element in his thinking, and, looked
at more closely, the emotive label turns out not to be
satisfactory. It often calls a halt to further inquiry just
where inquiry is needed. Much of what Richards classi¬
fies as feeling and tone can be analysed into different
aspects of sense. This objection has been taken up by
William Empson, and I think it is well founded; but for
all that, Richards’s approach is clear, compact and usable.
It is, no doubt, a formalization of what the skilful in¬
terpreter has always done; but there is an immense gain
in having it set forth in this lucid and serviceable form.
Much of Practical Criticism is concerned with the
obstacles to just interpretation—doctrinal prejudice, the
domination of received ideas, partial and ill-informed
notions about literary form. These are of great interest
and undeniable educational importance, but rather aside
from our present theme. Much of its message is conveyed
in detailed discussion of the poems and in discussion of
the comments on them; and this it is impossible to sum¬
marize. It remains to say something about the limitations
of Richards’s approach and the lines on which it has been
developed in the hands of others. Richards makes much
of the contribution of psychology to literary interpreta¬
tion. But the psychology employed is not of a very ad¬
vanced order, and I think it will seem to most readers
today that the real novelty of Richards’s method is a
willingness to consider the internal processes of reading
and interpreting as they actually are, rather than any
importation of esoteric psychological knowledge. This
means a continual emphasis on what goes on in the
G
88 STYLE AND STYLISTICS

reader’s mind; and at first sight it seems as though this

takes us away from the poem towards the psychology
of the audience. On this there are two things to be said.
First, this orientation is a necessary consequence of
Richards’s particular purpose, which was not, as we have
seen, a purely stylistic investigation. Secondly, and with¬
out going into any theoretical considerations about the
mode of being of a work of literary art, the report of a
reader’s reaction to a poem, if sufficiently detailed, re¬
fined and controlled, is in practice tantamount to a de¬
scription of the poem. And the effect of Richards’s
method, if rightly understood, is indeed to bring about,
at least initially, that submission to the poem on its own
terms that is recommended by a more traditional in¬
terpreter like Spitzer.
However, there is a difference. In Richards this sub¬
mission is only initial, and his procedure is decidedly
oriented towards judgment. His selection of pieces in¬
cluded bad and trivial poems as well as good ones, and
he is much concerned with the means of discriminating
between them. In general, stylistics has not thought much
along these lines. It has been content to accept the his¬
torical literary canon, and to examine only works that it
assumed to have excellence in their own kind. Spitzer,
for example, is on principle extremely chary of adverse
and limiting criticism, and would never dream of pulling
a bad poem to pieces for the sake of giving his powers of
discrimination an airing. This difference has come about
by reason of a subsidiary social and educational purpose
in Richards’s work—the purpose of providing a defence
against the flood of trash and corrupted values that now
besiege culture on every side. Heaven knows the defence
is needed; but it is a matter of some difficulty to know
how far literary studies can go in this direction without
losing their own integrity. We cannot discuss this ques-
 SOME PRACTITIONERS 89

tion here; but it is true that in the large body of writing

in English that has followed directly on Richards’s work
we find consequences of this attitude that are not par¬
ticularly admirable. The embattled stance towards bad
writing tends to become habitual, tends to extend itself
to the good, and to preclude a more generous acceptance.
Scrutiny’s inveterate hostility towards most of the creative
literature of its own time is a case in point. To use
Richards’s own terminology, doctrinal adhesions and
stock responses of a more or less sociological kind find
it too easy to obtain a lodgement in literary criticism
conducted under these auspices.
Quite evidently this was far from Richards’s intention.
But it was accidentally encouraged by the non-historical
nature of his inquiry. The contributors to the Practical
Criticism experiment were required to spin their judg¬
ments out of their own entrails, without any help from
the sense of cultural tradition, so strong in the best
Continental style-study. And this leads us to ask how
far literary judgment and interpretation, divorced from
history and cultural tradition, are possible at all. George
Watson has remarked that all that was really proved by
Practical Criticism is that unhistorical reading is bad
reading. And making all allowance for the special nature
of the investigation, I think we must give some weight
to this. We are invited to shake our heads over the
misinterpretations of Donne’s sonnet ‘At the round
earth’s imagined corners’; but a good deal of it is due to
ignorance that was forced upon the commentators by
the circumstances of the inquiry. True, it would not
happen now; a comparable audience would probably
know the poem; but at a period when seventeenth-
century religious poetry had been little studied, and was
certainly not part of the equipment of the ordinary
student, no adequate set of comments could have resulted.
9o STYLE AND STYLISTICS

It is not, as rather seems to be suggested by the tone of

the discussion in Practical Criticism, a matter of such
information as a footnote could supply—about beliefs
concerning the Last Judgment, etc.—though indeed a
good deal of such information is needed. It is at least
some understanding of a whole state of mind and climate
of feeling that is required. To read the poem out of its
context is not to read it at all. And it cannot be denied
that Richards’s work has in subsequent educational prac¬
tice often been misinterpreted, and has inadvertently en¬
couraged ill-prepared reading of this kind, and even led
to the belief (more evident amongst students a few years
ago than it is now) that literary study is a mere orgy of
rootless opinion, sanctioned by a few procedural tricks.
But even so the balance is heavily on the credit side.
The sense of mere helplessness in stylistic inquiry has
disappeared; some notion of how to tackle questions of
literary interpretation is widespread; and this can be seen
everywhere, from the work of students, through general
criticism, up to studies of the highest scholarly kind. In
fact, Richards had a lesson to teach and a great number
of people have learnt it. That it has not proved a cure for
all our cultural ills is another matter.

William Empson

Practical Criticism is obviously a prolegomena to style-

study rather than an example of it. The work of William
Empson both extends Richards’s methodological inquiries
and affords some massive examples of the method in ac¬
tion. The exemplification comes first. Empson’s first
book, Seven Types of Ambiguity, was written directly
under Richards’s influence, much of it indeed while
Richards was still teaching him in Cambridge; and it is
 SOME PRACTITIONERS 91

essentially the application of Richards’s method to a

particular literary problem. It is what we called in the
previous chapter the investigation of a special expressive
device, but it is a very extended one. By the time Empson
has finished with it, the special device becomes almost a
fundamental feature of all poetic language. Richards’s
analysis of meaning obviously leans towards the idea of a
composite reaction to a single utterance. The sense, for
instance, may seem to lead us in one direction, the feeling
in another, and this may be enriching, not confusing.
Under the heading ambiguity, Empson greatly develops
this idea. ‘I propose’, he says, ‘to use the word in an ex¬
tended sense and shall think relevant to my subject any
verbal nuance, however slight, which gives room for
alternative reactions to the same piece of language.’ The
result is a rich and detailed exemplification of the possi¬
bilities of multiple meaning in poetic language. It caused
a good deal of alarm when it first appeared. Empson is
more ingeniously intelligent than any of his critics; he
takes a slightly fantastic delight in verbal analysis; and,
as he says in the Preface to the Second Edition, he erected
the ignoring of tact into a point of honour. But when the
dust settled some years later it turned out that his main
principles had been quietly absorbed into the canon of
modern literary criticism. The words ‘tension’, ‘paradox’,
‘irony’, which have done such heavy duty in the last
thirty years are all adaptations of the Empsonian
‘ambiguity’.
The common objections to the procedure were that the
ingenuities were perversely pursued for their own sake,
and that the analysis left a passage or a poem in pieces
without reconstituting its totality. The first is false, the
second sometimes true, but perhaps inevitable. The book
is a commando raid on the ineffable, or on what con¬
ventional criticism has been content to regard as in-
92 STYLE AND STYLISTICS

effable. Again, a few sentences from the Preface to the

Second Edition explain perfectly justly what is going on:

I was frequently puzzled in considering my examples....

I felt sure the example was beautiful and that I had broadly
speaking reacted to it correctly. But I did not at all know
what had happened in this reaction; I did not know why
the example was beautiful. And it seemed to me that I
was able in some cases partly to explain my feeling to
myself by teasing out the meanings of the text. Yet these
meanings when teased out (in a major example) were too
complicated to be remembered together as if in one glance
of the eye; they had to be followed each in turn, as possible
alternative reactions to the passage; and indeed there is no
doubt that some readers sometimes do only get part of the
full intention.

Space forbids any illustration of the working of his

method; and in any case the book is familiar. I want to
point out here, however, that for all the difference of
tone and manner the method is essentially the same as
that of Spitzer and Alonso—an intuitive grasp of the
work, followed by an analytical attempt to show how
the intuition was arrived at. Of course, the methods of
analysis are very different from those of the Continental
scholars. Auerbach, Spitzer and Alonso make far more
systematic use of the history of the language and of
general cultural history. Empson remains more closely
focused on the reader’s mind. But, like Alonso, he be¬
lieved that the reaction in the reader’s mind, if it is
correct and not merely capricious, reconstitutes the
reaction in the mind of the author at the time of creation.
‘If critics are not to put up some pretence of under¬
standing the feeling of the author they must condemn
themselves to contempt.’ Empson’s reactions it must be
confessed have sometimes been capricious, and historical
 SOME PRACTITIONERS 93

scholars have sometimes been able to catch him out.

This does not, however, invalidate his procedure.
The Structure of Complex Words is clearly continuous
with the earlier book. It contains a detailed critique of
some of Richards’s views, offers another approach to re¬
place them, and gives some admirable fully developed
examples of this apparatus in action. The point in
Richards’s work which Empson takes up has already been
mentioned. It is the tendency to bundle all aspects of
meaning that are not sense under the head of ‘emotion’.
Empson takes exception to this. To describe complex
metaphorical and figurative language in poetry as having
‘purely emotive value’ is to give an inadequate account
both of what it achieves and how it works. He proceeds to
analyse most of what Richards calls ‘emotive’ as modi¬
fication or subsidiary direction of sense. What Richards
calls the ‘pseudo-statements’ of poetry are in fact ways
of stating something, not merely attempts at manipulat¬
ing the reader’s feelings. Empson does not deny the role
of feeling, but he sets himself to answer the difficult yet
elementary question of how the feelings get into words,
if they do get in; or how the words indicate their pre¬
sence, if the feelings merely accompany the words.
He first attempted to separate various entities in the
habitual uses of a single word—particularly senses, impli¬
cations, emotions and moods. He then goes on to show
how these separate entities interact with each other—
how a word can apparently carry a doctrine or direct
opinion. It does this, he says, by various kinds of equa¬
tions, which are then described, illustrated and com¬
mented on. Here I abandon the attempt to summarize;
the inquirer must read Chapter 2 of the book for himself.
Empson devises a symbolism to present these entities
and relations; and the system develops numerous off¬
shoots and parentheses. I do not doubt for a moment
94 STYLE AND STYLISTICS

that he is describing something which really goes on,

and that he accounts for many ways of conveying mean¬
ing that have not been accounted for before; but as it is
presented I find the argument difficult to follow. The
symbolism is, I think, ill-chosen and unmemorable, there
are too many aspects to be considered, and what needs
to be sub-divided or presented in tabular form is done as
a continuous argument, so that it is too easy to lose the
thread. Perhaps I am exceptionally stupid or exceptionally
lazy, but I cannot keep so many balls in the air at once.
There are hardly any local obscurities; the writing is al¬
ways lucid in detail, but I lose the sense of the whole. I
suspect I am not alone in this; and it does raise the
question of how far a literary argument can depart from
ordinary literary methods without denying itself to pre¬
cisely those readers who are likely to seek illumination
from it.
However, the theoretical chapters are only part of the
book. The rest consists of extensive essays in which the
method is applied. Key words in certain works are ex¬
amined and their various permutations and combina¬
tions are seen in the light of the whole work in question.
Examples are: ‘wit’ in the Essay on Criticism, ‘honest’
in Othello, ‘sense’ and ‘sensibility’ in various contexts,
‘sense’ in Measure for Measure. These inquiries are stylis¬
tics in the strictest sense; they start from close verbal
analysis and end in controlled literary interpretation. We
may agree or disagree, but the argument is generally close
and continuous and we are not left with the awkward
feeling common in reading Continental stylistics that
there is a gap between positive linguistic observation
and highly speculative literary conclusions. One cannot
suppose that this Empsonian shorthand is ever likely to
become popular reading; but there is no reason why the
method in a less concentrated form should not be per-
 SOME PRACTITIONERS 95
fectly available to others; and it certainly leads into wide
stretches of unexplored territory. We often hear about
how semantics ought to assist criticism; these brilliant
and deeply thought essays of Empson’s are almost the
only places I know of where it has really done so to much
purpose.
So I believe that Empson’s work is immensely important
and that it has not yet had its full effect. He has to a
certain extent become the victim of his own endowments.
Trained as a mathematician before he turned to literature,
with a considerable interest in the sort of problems that
symbolic logic investigates, a poet himself, he brings to
his critical work an equipment that the average literary
student cannot easily cope with. Add to this a few
passionate convictions, occasional bouts of ferocity, a
vigorous comic sense, and we have a mixture that the
academic appetite finds fairly indigestible. Yet it is too
much like hard work to have a very wide appeal outside
the academy. I suspect that literary studies must go on
digesting it for some time yet.

John Holloway

English style-studies do not show much continuity; they

are rather a succession of fresh starts. An entirely new
direction is found in John Holloway’s book. The
Victorian Sage. It does not take off from the work of
Richards and Empson, and it owes nothing to Continental
stylistics. Yet it clearly is stylistics—a study of the de¬
tailed verbal organization by which larger literary effects
are produced. As Richards approached literature from
semantics and experimental psychology, so Holloway ap¬
proaches it from philosophy—Oxford philosophy of the
1940’s. Richards’s point of view has an evident kinship
with early logical positivism. Early logical positivism was
96 STYLE AND STYLISTICS

concerned to distinguish meaningful statements from

statements that were strictly nonsense (not nonsense =
rubbish, but nonsense = non-sense). By Holloway’s time
the direction of interest had changed. The realm of non¬
sense continued to be thickly populated, and it obviously
included most of the things in which men take a passion¬
ate interest. Even if statements of this class were not
subject to verification, they continued to be made, were
made for some reason and presumably had some sig¬
nificance. So a later kind of verbal analysis set itself to
inquire what these significances are. This is roughly the
change from early to late Wittgenstein; and it is from
this later viewpoint that Holloway’s work starts.
The great Victorian sages, Carlyle, Newman, Arnold
and some of the novelists, all express notions of great
scope and generality about the world, man’s situation in
it and how he should live. Yet they all seem to feel
themselves dispensed from the burden of proof. Their
arguments are not logical or empirical demonstrations;
to a greater or less degree, all of them deal in vast
oracular assumptions, large assertions, ‘broad sweeping
gestures, hints thrown out, suggestions which leave us
quite uncertain about their detailed import’. They exercise
and have continued to exercise a pervading power. The
question that Holloway sets himself to answer is how this
power is exercised, what methods supply the place of
proof, of logical argument, of empirical demonstration
in writing of this kind. Obviously this is a matter of
great literary importance. The case of the Victorian sage
is particularly well marked, but it is not unique; and
these characteristic techniques of persuasion are liable
to be found in literature at most times. Holloway con¬
fines himself to a particular phase of a particular histori¬
cal period; but the questions he raises and the answers
he provides have a wide application; they concern the
 SOME PRACTITIONERS 97
status and the procedure of that vast body of speculative,
persuasive and hortatory literature, neither formally philo¬
sophical nor strictly scientific, that is found at all periods.
Again, it is impossible to summarize the book at all
comprehensively; the discussion is very close and im¬
possible to generalize. Since the sage or thinker is not
working to the rules of any formal discipline, the direct¬
ing force behind his writing tends to be his own tempera¬
ment and his own experience. This is particularly marked
in the case of Matthew Arnold, of which Holloway gives
a subtle and penetrating analysis. Arnold’s favourite
method is to exhibit his own personality as an example
of the virtues that he is recommending; and he uses
irony partly indeed as a weapon against his opponents,
but partly as irony directed against himself, to disinfect
this presentation of his own personality from com¬
placency and self-importance. If disinterestedness, ur¬
banity and amenity are to be pressed upon the British
public, the very texture of all Arnold’s writing must
exhibit these virtues, and the process of persuasion is
more like a face-to-face encounter with a person in actual
life than a formal argument, the capacity for which
Arnold is always disclaiming. Another of his methods
is repeated and cumulative quotation from his adversaries,
by which process a particular response, this time un¬
favourable, to their attitudes and temperaments is built
up. An extension of this is the creation of imaginary
personalities to be the bearers both of his own attitudes
and those of his adversaries. And here Arnold joins hands
with Carlyle, who exhibits even more strongly this ten¬
dency to dramatize an argument. In his case it is a
function of his overwhelming sense of movement and
activity, of a world full of variety, development and
burgeoning life. His vocabulary and his metaphors, both
of which Holloway illustrates very fully, exhibit this
98 STYLE AND STYLISTICS

quality at every turn. Newman is examined in the same

way—his forms of argument, his use of suggestive ex¬
amples and his imagery are all discussed.
These three essays do something new. They give an
analysis of informal unprofessional philosophical writing
from the point of view of a linguistic philosopher.
Holloway is not much interested in the validity of the
arguments (had he been, he would have been obliged to
say that some of them are remarkably slippery): it is
their methods that interest him. And he demonstrates the
existence of a method that can be examined and de¬
scribed, in literary areas that had formerly been ex¬
tremely resistant to analysis. The three remaining chapters
of the book are on novelists, Disraeli, George Eliot and
Hardy, whom Holloway sees also as sages—quite appro¬
priately within his terms of reference. But here his own
method tends to take a different direction. As is usual in
criticism of a novel, it is the larger features—character,
incident, description—that engages his interest, and the
linguistic structure is less attended to. There is some
examination of images and descriptive passages, but a
good deal of these essays is close and intelligent general
criticism rather than stylistics.

Stephen Ullmann

We can contrast Holloway’s treatment of the novel with

that in Stephen Ullmann’s work—Style in the French
Novel and The Image in the Modern French Novel.
Ullmann writes in English, but his procedure is closely
allied to that of Continental stylistics. Much of his work
has been on semantics. The introductory chapter of
Style in the French Novel (together with a later book.
Language and Style) gives a valuable compendium of
current stylistic methods and problems; indeed, it is the
 SOME PRACTITIONERS 99
best general introduction to the subject in English. In
general, Ullmann starts from linguistic observations and
tries to indicate how the gap between linguistic and
literary study can be bridged. Unlike Spitzer, he advances
no comprehensive theory of the relation between lin¬
guistics and literary history, but he has made a number
of separate studies of particular stylistic questions. For
instance, the essay on reported speech in Flaubert takes
up the question of the style indirect libre, gives a report
of preceding discussions of the matter and carries them a
stage further. Other essays in the same volume deal with
the attempt of French novelists of the nineteenth cen¬
tury to introduce local colour into their work, and with
certain syntactical constructions in the work of the
Goncourts which can be seen as parallel to impressionist
techniques in painting. These are evidently parts of a
large tendency in the nineteenth-century French novel—
to introduce a more colourful and expressive style into
prose fiction—a tendency which has gone much farther
in our own century. Ullmann confines himself pretty
much to specific and limited questions, but his books also
give extensive and valuable bibliographical indications,
often to works that are not generally known to English
readers. We can learn from them how much more there
is to be done in this field, and in what directions further
inquiries can be pursued.

Donald Davie

Last, I want to refer to two books by Donald Davie,

unique and isolated pieces of work, both dealing with
matters of great general significance. They are Vurity of
Diction in English Verse and Articulate Energy. The first
deals with the diction and the second with the syntax
of English poetry. These works are isolated because their
IOO STYLE AND STYLISTICS

starting-point is neither linguistics nor conventional stylis¬

tic criticism; they arise rather from Davie’s own concerns
as a poet. Even when they deal with matters that have
often been discussed in a special technical vocabulary,
they make very sparing use of linguistic terminology.
They show indeed what in the view of many people, my¬
self included, needs to be shown—that linguistic matters
can be properly discussed in the traditional language of
literary criticism.
Purity of Diction has as its point of departure a par¬
ticular phase in the history of modern English poetry—
a reaction (in itself probably a minor affair) against the
self-consciously concrete, image-packed style that was the
English heritage of symbolism. As a result of what can
roughly be called the imagist propaganda of Pound and
Eliot, it had become almost dogma in the thirties and
forties that the language of poetry must be inveterately
concrete, that the prime weapon of poetry is the image.
Davie has no difficulty in showing that historically this is
a good deal less than certain; and he sets out to illustrate
the virtues of another currently neglected kind of diction,
a diction characterized by a cool exactitude, often ab¬
stract, rather than by the desire to hand over immediate
physical sensation. His intention is partly polemical and
to that extent aside from stylistics; but in the course of
pressing his argument he gives extremely sensitive and
original analyses of poetic language from the seventeenth
to the twentieth century.
It was also a received idea that the first necessity for
poetry in the earlier part of this century was the establish¬
ment of an appropriate modern poetic diction. The effect
of Davie’s slightly sceptical inquiries was to remove some
of the emphasis from diction. The question was then
where the emphasis was to be put. The answer to this
comes in Davie’s next book. Articulate Energy, which is
 SOME PRACTITIONERS IOI

about poetic syntax. If we begin to doubt the absolute

primacy of a vivid and sharply-coloured diction, it be¬
comes clear that the principal agent of poetic power is
syntax—syntax in the widest sense. The effect of imagist
doctrine had been to depreciate syntax, even in extreme
cases to abolish it altogether. A mere parataxis of striking
images was to take its place. As Davie uses the word, it is
not simply syntax in the grammatical sense, not simply
ordonnance in the interests of lucid and effective arrange¬
ment. It means the whole skeletal and muscular structure
of poetic language, as the only possible instrument of
poetic energy. The word ‘poetic’ is an important quali¬
fication, for the variety of syntax possible in poetry is
far wider than in prose. Davie illustrates this with an
impressive range of discussions both of current theory
and past practice. Again there is a polemical purpose;
but even if this should come to seem less relevant than
it once was, the book remains immensely illuminating.
These two books illustrate very clearly that whatever
contribution linguistics can bring to literary studies there
is no substitute for literary understanding and literary
ability. And that is not a bad note to end this chapter on.
 5
Conclusion:
limits and possibilities

It will be apparent from what has been said already that

the contribution of linguistics to style-study is strictly
limited. It is virtually confined to semantics and syntax.
Many of the positivist linguists of a now slightly out¬
dated school regarded semantics as only doubtfully a
part of their discipline, and semantics has therefore been
rather apart from the main body of modern linguistics.
Much of the modern study of syntax is remote from
literary interests, and for many of its practitioners
(Chomsky, for instance) literature is only an epipheno-
menon.
There is a fundamental reason for this. Modern struc¬
tural linguistics sets out to study language as a general
human phenomenon, and to evolve a set of concepts
that will describe any language. A linguist who is study¬
ing, say, modern English has to keep this whole ex¬
planatory scheme in mind. Stylistics is in quite a different
position. It studies particular works in a particular
language. A student of style in the modern French novel
is not at all helped by remembering the linguistic habits
of the Hopi Indians. And even the French language is of
interest to him within narrow limits. La langue, the public,
shared system of the language, its phonemic, morphemic
H
104 STYLE AND STYLISTICS

and syntactical structure, is for him a datum, not an ob¬

ject of investigation. He is concerned only with la parole,
a series of individual communicative acts, individual
applications of the code. Great harm is done to fruitful
collaboration between linguistics and literary studies by
linguists who wish to foist on literature a whole battery
of apparatus and a whole array of accomplishments that
are quite irrelevant to its purpose.
In fact. Continental stylistics arose out of an older
school of historical linguistics. This was indeed the
linguistics of a particular language, or group of languages;
and so in an obvious sense it was closer to style-study.
Even here much of it was irrelevant. It is frequent to
find, for example, in the work of Spitzer fossils of histori¬
cal philology that serve him as a starting-point, but have
little real connection with what he ultimately wants to
say. And the school of Richards shows that the starting-
point could be quite different. But the fact remains that
literature is a linguistic structure, and some awareness of
the nature of language would seem to be essential to the
student of literature. There is an obvious value in even
so much (or so little) linguistics as would remind the
literary student that all literary communication is
achieved by linguistic means, that these means can be
described, and that linguistic techniques can alert him
to features that would otherwise have passed unnoticed.
The absence of this awareness in Practical Criticism and
its progeny has undoubtedly been damaging. The em¬
phasis on uninspectable events in the reader’s mind, and
his ‘attitudes’ and emotions, has led much literary study
into an orgy of unsubstantiated opinion—amateur philo¬
sophizing, amateur sociology, and propaganda for what¬
ever code of morals the individual critic happens to
support. This is not to say that such work can have no
value: it approximates to the activity of the Victorian
 conclusion: limits and possibilities 105

sage, and it requires long experience and great general

powers. But it cannot be a progressive study; it is per¬
suasion rather than knowledge; and it is a poor educa¬
tional foundation. One cannot really run a School of
Prophecy. Stylistics makes more modest claims: that it
can be the systematic (I will not say scientific) study of
literary expression; that within limits it can increase
knowledge; that this knowledge can be consolidated, and
within limits communicated to others.
What are these limits? One is the limit of natural
aptitude, the capacity for receiving intuitions from works
of literature. About this there is nothing to say. The
other is that stylistics, however extended, can never cover
the whole field of literary study. Much that literary
students are interested in consists of larger units than
style-study can cope with—plot, character, and the
ordonnance of ideas; and without becoming ‘the gay
sporting-ground of incompetence’, in Spitzer’s phrase,
literary study can attack these larger structures direct,
short-circuiting the approach through language and style.
The methods are different, but failure of communication
between them is not necessary. Empson, who has ap¬
proached Shakespeare through language, has said that the
Shakespearian critic he most admires is Bradley.
It seems likely that modern linguistics is developing in
ways that will take it farther away from literature—
hence the old-fashioned air of most of the linguistic
references in this book; but mention must be made to
one move in the contrary direction. I mean the holy
alliance between linguistics and anthropology associated
with the work of Levi-Strauss. Levi-Strauss is an anthro¬
pologist, and of this, his real field, I can say nothing.
But the essence of his work is to discover a common
principle in all human thought—in myth, primitive re¬
ligion, kinship patterns, equally with the thinking of
106 STYLE AND STYLISTICS

modern man. The emphasis is on thought; and the affec¬

tive element, the psychology of the passions that we
associate with Freud, is depreciated. Part of Levi-Strauss’s
structural organon is derived from Saussurean linguistics,
with additions from one of the most distinguished lin¬
guists of the next generation, the Russo-American Roman
Jakobson. A particularly important principle is that of
binary opposition, the choice between a pair of alterna¬
tives as the fundamental principle of human thinking.
(It is often difficult to tell whether this is put forward as
part of the nature of things, or simply as a model, a heu¬
ristic convenience.) And as anthropology has borrowed
from linguistics, so literary thinking in its turn is in¬
clined to borrow from both. There has been a wide ex¬
tension of quasi-anthopological ideas, an enthusiastic and
often quite irresponsible application of methods drawn
from one discipline to others with which it is not evi¬
dently related. In French criticism ‘structuralism’ has
become a magic word. Roland Barthes, its principal
spokesman, professes a science of semiology, a key to all
cultural codes, whether they are linguistic or not. But he
is also a critic and writer on style. His Essais Critiques
appear to the outsider a rather belated discussion of
modern critical methods already more fully worked out
in Anglo-American writing, together with a little com¬
munication theory—the whole boiled up into a highly
flavoured stew. His book on Racine displays under the
name of ‘structuralism’ an assortment of familiar but
erratically employed Freudian ideas. It has been effec¬
tively, but rather stuffily, criticized by Raymond Picard. In
Le degre zero de Yecriture Barthes is nearer to stylistics
with an analysis of neutral or ‘styleless’ writing. There
is a ferment here that recalls the controversies of the
forties about the New Criticism in America. Some of
it is mere froth; but there is also a host of new ideas.
 conclusion: limits and possibilities 107

though they are more likely to prove fruitful in philo¬

sophical criticism than in stylistics.
A more serious instance of structuralism applied to
literature, and one much nearer the fountain-head, is a
joint study of Baudelaire’s sonnet ‘Les chats’ by Jakobson
and Levi-Strauss. Such a collaboration cannot be without
importance. The point of view is made clear in the intro¬
ductory paragraph:
It may seem surprising that an anthropological review
should publish a study devoted to a French poem of the
nineteenth century. However, the explanation is simple;
if a linguist and an ethnologist have thought fit to unite
their efforts to understand what a sonnet of Baudelaire
was made of, it is because they have found themselves
independently confronted with complementary problems,
in poetical works the linguist discerns structures that bear
a striking analogy with those that the analysis of myths
reveals to the ethnologist.

There follows a minute formal analysis of the poem,

comprehending the stanza-form, rhyme-scheme, the
phonological structure and the semantic structure. The
closeness of the observation, and its justness, cannot be
doubted, as is never the case with Barthes. There is an
obvious gain in perception of detail which the normal
literary student would hardly have noticed; but it re¬
mains an open question how great is the gain for literary
understanding. All that is demonstrated by this in some
ways impressive piece of analysis is that many structures
actually present in a work of literature are without
literary importance, and that some things of literary im¬
portance cannot be represented in structural terms at all.
Every study has its own end, and therefore a relative
autonomy within its own sphere. In literary study there
can be no substitute for literary intelligence. But every
study has its external relations: it may be part of a larger
108 STYLE AND STYLISTICS

whole; it may have multifarious links with other ac¬

tivities. It is hard to see literary study as part of a larger
whole; and its links with other activities have probably
been too many and too unregulated for its own good.
The literary philosophy and the literary sociology that we
have mentioned are cases in point. One contact that is
indispensable is that of literature with language, of criti¬
cism with linguistics; and it obviously needs more dis¬
interested investigation than it has had. This is not, or
should not be, a take-over bid by either side; it is the
exploration of a common frontier. Stylistics is the border
area between these two studies, and as yet it is neither
adequately mapped nor firmly settled. It would be a
mistake to underrate the difficulties.
The first is that every branch of study in modern
conditions makes such huge demands. The literary canon
presented to a student of English literature is vastly en¬
larged since the old days, when it ran from Chaucer to
Wordsworth. Even if we neglect transitory novelties (and
we probably do not want to, for they may speak to our
condition) the range of necessary critical ideas is far
wider than it used to be. And no one can be satisfied
with a literary education confined to a single language.
To attain any real competence as what was once called
‘a man of letters’ is a long and exacting process; and it
can hardly be combined with a long and exacting process
of quite a different kind. Modern linguistics is a highly
developed and complex science, with many branches,
with its own intellectual apparatus, and a frame of
reference that only overlaps with literary study in a
small and peripheral area. What is more, the state of
mind in which the two studies are pursued is radically
different. The linguist aims to describe the object of his
investigation as fully and explicitly as possible, without
any ambiguity or appeal to intuition. The literary student
 conclusion: limits and possibilities 109

finds complete description superfluous or stultifying, often

values the suggestive rather than the explicit, and is
tolerant of diverse interpretations. For the linguist value
resides in the completeness and exactitude of his descrip¬
tions; the actual material he works on may be a mere
corpus vile used for experimental purposes. For the
literary student value resides in the work of art under
consideration; description and interpretation are ancillary
and subservient to a kind of contemplative understanding
that is essentially independent of these activities.
So I think it quite idle to suppose that literary criticism
and linguistics can ever be a united field, though they can
form brief alliances for special purposes. The special
purposes are likely to be those of the student of litera¬
ture. I have never, in such reading of linguistics as I have
done, noticed any occasion when the linguist has wanted
to avail himself of literary knowledge. There are many
occasions when the literary student, especially the student
of style, wants to make use of linguistic techniques and
linguistic knowledge. But he will hardly be equipped
with the full range of linguistic skills, and he will prob¬
ably make brief raids on whatever parts of linguistics suit
his purpose. There is no reason to suppose that style-
study will be willing to follow wherever the technical
exigencies of linguistics happen to lead. As time goes on,
it is just as likely that literary thinking about style will
be influenced by a general ‘science of signs’, a semiology
less eccentric and trivial than that of Barthes. It may be
more affected by communication theory, some of whose
concepts seem admirably adapted to literary use. Works of
literature are expressed in language; they are systems of
signs; they are communications. Ideas drawn from the
sciences that preside over these several areas can all con¬
tribute to literary study. But literary works are also works
of art of a unique kind, and their proper study has its
no STYLE AND STYLISTICS

own methods and its own ends. It even has its own
public, different from that of the more specialized
sciences. However esoteric literary studies may become,
they must fail of their object unless their results ultim¬
ately filter through to the intelligent common reader,
and unless they are expressible in something like the
language of common life.
Select Bibliography

Short notes are given on works not discussed in the text.

alonso, a., ‘The Stylistic Interpretation of Literary Texts’,

Modern Language Notes, lvii (1942), 489-96.
A liberal-minded manifesto for stylistics as a literary activity.
alonso, d., Toesia espahola; ensayo de metodos y limites
estilisticos, Madrid, 1950.
auerbach, e.. Mimesis: the Representation of Reality in
Western Literature, trans. Willard Trask, Princeton, 1953.
bailey, r. w. and burton, d. m., English Stylistics: a Biblio¬
graphy, M.I.T., 1968.
The only full critical bibliography of English stylistics. Much
needed.
bally, c.. Traits de stylistique frangaise, 3rd edn., Paris, 1951.
- Le langage et la vie, 3rd edn., Geneva, 1951.
barthes, r„ Writing Degree Zero, trans. Annette Lavers and
Colin Smith, 1967 (Le Degri Zero de l’ecriture, Paris,
1953)-
- Essais Critiques, Paris, 1964.
-Sur Racine, Paris, 1963.
bloomfield, l.. Language, New York, 1953.
A standard comprehensive handbook of linguistics, now
partly superseded by Hockett.
booth, w. c.. The Rhetoric of Fiction, Chicago and London,
1961.
A pioneering study of the expressive resources of the novel,
especially of the role of the narrator.
112 SELECT BIBLIOGRAPHY

cressot, M.,Le style et les techniques: precis d’analyse

stylistique, Paris, 1947.
davie, d., Turity of Diction in English Verse, London: Chatto
& Windus, 1952; rev. edn., 1967.
- Articulate Energy, London : Routledge & Kegan Paul, 1955.
doubrovsky, s., Tourquoi la nouvelle critique?, Paris, 1967.
A defence of Barthes and the French ‘new criticism’ gener¬
ally.
empson, w.. Seven Types of Ambiguity, London: Chatto &
Windus, 1930; rev. edn., 1947.
- The Structure of Complex Words, London: Chatto &
Windus, 1951.
enkvist, n. e., spencer, j., and Gregory, m. j.. Linguistics
and Style, Oxford, 1964.
Useful opening essay on defining style.
fowler, r. (ed.). Essays on Style and Language, London:
Routledge & Kegan Paul, 1966.
A collection of essays attempting to explore ‘an area where
linguistics and literary criticism overlap’. Not notably suc¬
cessful from the literary point of view.
gourmont, r. de, Le problbme du style, Paris, 1902.
An early classic in this field. The principal essay in the col¬
lection is a polemic against mechanical and external ideas
of style.
Hamilton, g. r., The Tell-tale Article: a Critical Approach
to Modern Voetry, London: Heinemann, 1949.
hatzfeld, h., ‘Stylistic Criticism as Art-minded Philology’.
Yale French Studies, 11 (1949), 1-9.
- A Critical Bibliography of the New Stylistics, applied to
the Romance Languages, 1900-52; Chapel Hill, N.C.,
1953-
A comprehensive bibliography, with very full discussion, of
stylistic work on the Romance languages. This is not only
an indispensable tool, but amounts to a critical work in its
own right.
hockett, c. f., A Course in Modern Linguistics, New York,
1958.
The best and most up-to-date general handbook.
 SELECT BIBLIOGRAPHY 113

holloway, j.. The Victorian Sage: Studies in Argument,

1953 -

jakobson, r., and levi-strauss, c., ‘ “Les Chats” de

Charles Baudelaire’, L’homme: revue frangaise d’anthro-
pologie, n, i (1962).
lodge, d.. Language of Fiction: Essays in Criticism and Verbal
Analysis of the English Novel, 1966.
Particularly useful introduction on the novel as a linguistic
structure.
miles, j„ Eras and Modes in English Voetry, Berkeley, 1964.
murry, j. m„ The Vroblem of Style, 1922.
A general discussion, rather loose and unsystematic by later
standards, but containing some valuable observations.
nowottny, w.. The Language Toets Use, 1962.
A sensitive and original account of poetic language, using
linguistic insights in the interests of literary understanding.
picard, r., Nouvelle critique oil nouvelle imposture, Paris,
19 65.
An attack on Barthes and the French ‘new criticism’ in
general.
Richards, 1. a., Vrinciples of Literary Criticism, 1924.
-Vractical Criticism, 1929.
- Interpretation in Teaching, 1938.
saussure, f. de. Corns de linguistique generale, 5th edn.,
Paris, 1955.
sayce, r. a.. Style in French Vrose: a Method of Analysis,
1953 -

A systematic procedure for the analysis of prose—too

systematic, probably, for most literary purposes.
scherer, j., L’expression litteraire dans l’ceuvre de Mallarme,
Paris, 1947.
sebeok, t. a., (ed.), Style in Language, New York and London,
i960.
Report of a joint conference of linguists and literary critics:
a few articles of merit, but on the whole profoundly de¬
pressing, and contains more nauseous jargon than any similar
work known to me.
spitzer, l.. Linguistics and Literary History, Princeton, 1948.
114 SELECT BIBLIOGRAPHY

-Essays on English and American Literature, Princeton,

1962.
-Stilstudien, Marburg, 1928,
Style in Prose Fiction: English Institute Essays 1958, New York,
1959-
A collection of essays: valuable introductory essay by R. N.
Ohmann on the concept of style.
thibaudet, a., Gustave Flaubert, 1821-80; sa vie, ses romans,
son style, Paris, 1922.
Admirable chapter on Flaubert’s style.
ullmann, s.. Style in the French Novel, 1957; 2nd edn., 1964.
Useful introduction on stylistics in general, a sound working
bibliography, and further very extensive bibliographical in¬
dications in the notes.
- The Image in the Modern French Novel, London: Cam¬
bridge University Press, i960.
vossler, K., Medieval Culture: an Introduction to Dante and
His Times (Die gottliche Komodie), trans. W. C. Law-
ton, London: Constable, 1929.
- The Spirit of Language in Civilization, trans. O. Oeser,
London: Kegan Paul, 1932.
wimsatt, w. k„ The Prose Style of Samuel Johnson, New
Haven, 1941.
yule, g. u.. The Statistical Interpretation of Literary Vocabu¬
lary, London: Cambridge University Press, 1944.
Concepts of Literature
General Editor, William Righter
Lecturer in English
University of Warwick

This series is designed to provide both

the student and the general reader with
a new approach to the study of
literature. Instead of considering in¬
dividual authors, works or periods the
series will explore in depth literary
forms, ideas and issues. Each con¬
tributor will take a major concept such
as Symbolism, Style or Realism and
provide a clear description and critical
evaluation of the term. The treatment
throughout is introductory and each
volume will be as concise and up to
date as possible. The volumes in the
series are all written by leading
scholars in the field, and will, it is
hoped, enable the reader to add a new
dimension to his understanding of
literature and its criticism.
 CONCEPTS OF LITERATURE

All books available in two editions, library and paperback.

Books already published :
COMPARATIVE LITERATURE
Henry Gifford, Professor of English, University of Bristol

SBN 7100 6417 9

ROUTLEDGE & KEGAN PAUL