Stylistics routledge English Language Introductions

Language and literature: Roger Fowler
Mr Bateson’s attempt to disqualify linguistics as a discipline of relevance to litera-
ture comprises a very cunning and apparently substantial argument inter-woven with
a misleading line of polemic. In his usual masterly fashion, he has constructed a case
against which it is very difficult to argue in an organised way. I will try to show that
his argument (1) rests on premises which cannot provide an adequate aesthetic for
literature, (2) is ineffective as a disqualification of linguistics because of misconcep-
tions about the nature of linguistics, (3) is motivated by a set of prejudices which
inhibit constructive discussion of the present issue. [. . .]
(1) Mr Bateson’s case is basically a somewhat complicated version of ideas he has
offered linguists and medievalists for at least the last ten years. He compounds 
(a) Coleridge’s ‘homely definition’ of poetry as ‘the best words in the best order’;
(b) I. A. Richards’ distinction between scientific and emotive language; and (c) 
de Saussure’s distinction between 
langue
and 
parole
. These formulae are jumbled
together to support an assertion of a unique and exclusive aesthetic so con-
structed as apparently to make literature inaccessible to objective study [. . .]
(a) ‘The best words in the best order’ is, except in Mr Bateson’s first book, an
empty catch-phrase. Coleridge used it to distinguish verse from prose, and
his requirement that ‘the words, the media, must be beautiful, and ought
to attract your notice’ becomes impertinent when the phrase is employed
to mark off literature from nonliterature, as Mr Bateson uses it currently.
It is obviously at best incomplete when applied to most kinds of poetry, and
quite wrong for narrative verse and for drama and the novel. [. . .]
(b) I. A. Richards’ distinction between emotive and scientific language is gener-
ally agreed to be without reasonable foundation: it is much too simple to
serve as a linguistic theory or to provide an isolation of the characteristics
of ‘literary language’. [. . .]
(c) I do not think the reference to de Saussure is merely a sop to the linguists.
Unfortunately, taken as a serious proposal it doesn’t make sense. 
Langue
and 
parole
together, according to de Saussure, make up the 
faculté du
langage
. 
Langue
is the abstract system of rules which enables a speaker to
communicate with others. 
Parole
is a particular concrete act of speech, or a
corpus collected from many such acts. Chomsky has recently revived and
adapted the distinction using the terms 
competence
and 
performance
. Mr
Bateson, like de Saussure and Chomsky, says that linguists study 
langue
,
which is true in so far as linguists write grammars of languages and enquire
into the linguistic mental capabilities of speakers. But, Mr Bateson says, 
[. . .] the study of literature is the study of 
parole
– or rather, the study of
11
111
11
111
L A N G U A G E A N D L I T E R A T U R E
153

---

a special kind of 
parole
of which the 
langue
is not 
langue
but ‘style’ [. . .]
This is an illegitimate and meaningless adaptation of de Saussure. It is ille-
gitimate because the original scheme works only if it applies in the same
way for all uses of language; it is meaningless because of the poverty of Mr
Bateson’s definition of the linguistic components of style and because of his
misunderstanding of the interests and procedures of grammatical study in
relation to 
langue
and 
parole
.
(2) [. . .] we should not regard the linguistic description of a text as the discovery
of structure. A linguist ought only to claim to make articulate what a native
speaker tacitly knows about a text he understands. In this way, linguistic descrip-
tion is a representation of (part of) the process of understanding: it shows how
a native speaker encounters and interprets sentences generated by the grammar
(
langue
) that he knows. The ‘linguistic analysis’ of literature is an attempt to
make explicit part of the process of reading by the use of terms and concepts
which have psychological reality (are humane even if they are scientific) through
being appropriate to the reader’s individually internalised yet culturally shared
grammar of the language. Some such description is necessary because speakers
do not 
know about
their language anything like as efficiently as they 
know
it. I
see no reason derived from either the supposed opposition of science and liter-
ature, or the theoretical potential of linguistics (as distinct from the claims of
many linguists and some linguist critics), why such a linguistics should be intrin-
sically alien to certain parts of literary study.
(3) The ‘misleading line of polemic’ derives from Mr Bateson’s feelings about the
old and disgraceful ‘language-literature controversy’. There are still large
numbers of people in English departments who, for reasons of academic poli-
tics, feel obliged to parcel the world into two intractably opposed groups. This
is not the proper place to make a fuss about that issue, since only by associa-
tion does it affect the dispute between Mr Bateson and me. Modern linguistics
is only accidentally the ‘successor’ of the old philological medievalism. Linguistics
lays fair claim to be an autonomous discipline in its central concern, the theory
of 
langue
; it provides a theoretical status for 
parole
, but needs augmenting by
other disciplines in the study of 
parole
. In places where 
parole
is studied (e.g.
English departments) linguistics has obvious uses. But its value cannot be appre-
ciated while people’s words and votes are governed by oversimplified and
historically irrelevant ‘sets’ towards imagined opponents. Reciprocal ignorance
about literary criticism and linguistics is also, as I have several times argued,
deeply inhibiting.
We can now attempt to rebuild the theory of style. I will argue that linguistic theory
provides a definition for a level of language which we may call ‘style’. Since no work-
able theory has ever been constructed which uniquely isolates a category of language
‘literature’, there seems no good reason why the concept of style should be exclu-
sively literary. (Whether or not a particular text is regarded by its readers as being
literature is an important cultural fact, but an irrelevant fact from our point of 
view.) The linguistic theory which provides a definition of style by the same definition
154
E X T E N S I O N

---

establishes the contribution, and the limitation of contribution, of grammar to the
study of style. [. . .]
I hope the general strategy of my reorientation is clear. The distinction between
competence and performance provides the only fruitful way of understanding the
special formal qualities (style among them) of texts. Wellek and Warren speak of 
‘the contrast of the language system of a literary work of art with the general usage
of the time’ (1949: 177). The contrast attempted is evident from the mention in 
the next line of ‘common speech, even unliterary speech’: the false norm of non-
literary language is being invoked to provide background for a spurious category
‘literary language’. But in reality the sense in which there is a gap between the ‘normal
language’ and any distinctive text is the same for all texts: there is the grammar
(langue, competence) and what you do with the grammar (parole, performance, 
e.g. texts identified stylistically). In the light of this fundamental tenet of linguistic
theory there is no formal category ‘literature’. In my opinion no study of texts which
recognises the critical importance of ‘style’ can afford to define ‘style’ against any
background other than that provided by a linguistic theory of this kind.