Stylistics routledge English Language Introductions

Literature and linguistics: a reply by F. W. Bateson
Roger Fowler is in effect proposing – both here and in the collection of essays he
edited (Fowler 1966) which was reviewed by Mrs Vendler – an academic alliance
between post-Saussure linguistics and post-I. A. Richards criticism. And why on earth
shouldn’t he? 
A priori
, as it were, it sounds a good idea. After all, even if a literary
masterpiece is more than the sum of its words, still when the words are taken away
what is left? Blank pages! To invite the reader to look hard, really hard, at the words
on the page is indeed what the modern critical doctrine of close reading amounts 
to, when it is reduced to its simplest terms. And, since a similar concentration is 
the initial premise of modern descriptive linguistics, some degree of amicable co-
operation between the two approaches should not be impossible.
Unfortunately it doesn’t work. Or rather, when it does work, it is only at the most
elementary level. In this reply to Mr Fowler I shall try to indicate rather more fully
than was possible in my short postscript to Mrs Vendler’s review why the sort of 
co-operation that Mr Fowler and his colleagues are pleading for is a vain hope.
[. . .]
Descriptive linguistics is always at least headed towards total description – a
detached, objective, universally available discipline (whatever the user’s age, sex,
nationality, or culture). Literature, on the other hand, has its ineradicable subjective
core, which tends to define the range and effectiveness of its uses. The point of depar-
ture between the two specialisations from the vulgar tongue can also be put in strictly
linguistic terms. It is a matter either of breaking down the sentence into its separable
parts – or else of taking the sentence as the unit and building up larger units as the
sentences accumulate. The close reader conscientiously intent on the words in front
of him, can opt either way. If he is a natural grammarian he will divide and subdi-
vide the verbal material; if he has been born a literary critic he will synthesise and
amalgamate it. What is it, then, in the words of literature that encourages the literary
reader to amalgamate and not to subdivide? The answer to this question is a crucial
one in my argument with Mr Fowler, though to prove it as a case is proved in a
court of law would require a book instead of a short article. But I can at least
summarise the theoretical objections to the mating of the language of description
and the language of evaluation. Grammar, for one thing, is essentially logical in 
its linguistic presuppositions, and as such it is governed by the principle of non-
contradiction; literary criticism, on the other hand, assumes in the verbal material
criticised the presence of opposite and discordant qualities whose provisional balance
and reconciliation the common reader will agree under certain circumstances to
accept. Those circumstances, considered linguistically, can be summed up in the word
‘style’ – a term that includes the whole armoury of rhetorical devices, phonetic and
semantic, with their larger structural extensions such as tragedy and comedy. The
function of style is to unify – or at least encourage the reader to attempt to unify –
literature’s disparate linguistic parts. As such it is the exact opposite of grammar,
whose function is not primarily to unite but to divide. (A sentence is grammatical
when its separate parts have been found subject to classification, the ‘parsing’ process,
and then shown to cohere.) Although some grammaticalness certainly survives in
literature, it is as it were accidentally and incidentally, a left-over of logic from the
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
151

---

common speech of which the language of literature is one derivative. The reader is
scarcely aware of it. What he is aware of – especially in poetry but also in prose with
any literary pretensions – is the style (in the wider sense already indicated). If my
attention is drawn to breaches of grammar in a work of literature, I can always invoke
the magic word ‘ellipsis’ – a term apparently invented to save grammar’s face when
we really ignore it.
May I offer Mr Fowler a definition of literature? A work of literature is successful
linguistically, the best words in the best order, when appropriate stylistic devices 
co-operate to unify humane value judgments, implicit or explicit, on some aspect of
life as it is lived in the writer’s own society. As for the reader of such a work, he 
will only be successful if he registers, consciously or at least semiconsciously, the
unifying stylistic devices that enable him to respond to the human situation avail-
able to him in it. In a word, the role played by grammar in description is comparable
to that of style in evaluation. But if comparable they are also mutually incompatible,
because grammar is primarily analytic in its methods and premises, whereas style is
essentially synthetic.
To Mr Fowler’s optimistic escape-clause that, because ‘some mathematicians and
physical scientists are fine musicians and poets’, therefore a linguistic training will
sometimes be useful for the literary critic, the answer is simple and obvious: musical
physicists do not improve as physicists by learning to play the piano. I am not sure
who Mr Fowler’s mathematicians and physicists are who are also fine poets, but these
ambidextrous geniuses certainly don’t grow on blackberry bushes. I can’t think of
one. On the other hand, it is common knowledge that most structural linguists don’t
write particularly good English prose. Why indeed should they as long as they are
intelligible? The temperamental predisposition that results in Smith becoming an
eminent grammarian is normally very different from that which turns Brown into a
good critic. Let us agree to be different. [. . .]
Stylistic discrimination is the one indispensable prerequisite for the aesthetic
appreciation of great literature. That some knowledge of linguistics, historical and
descriptive, has certain minor uses in literary studies is not to be denied, but for the
native speaker of English this additional knowledge is, as it were, supplementary either
in eking out one’s birthright by the help available in the OED for an unusual word
or idiom, or in saving one from incidental errors outside one’s immediate range of
linguistic experience. Such information may be compared to the odds and ends of
social and political history with which the scholar-critic will also have to equip
himself. A little learning of this kind will go a long way, though with none at all the
reader’s fingers can sometimes be badly burnt. But for the literary beginner the best
way to acquire such information – including the sign-posts to the topic’s more sophis-
ticated levels – is 
ad hoc
; in other words, by consulting a glossary or an editor’s notes
only when he needs them. What he naturally resents is the traditional compulsory
spoon-feeding with grammar or history most of which he will never need. Of course,
if he is interested in either descriptive linguistics or the history of the language or
social history for its own sake, that is another matter. My real quarrel with Mr Fowler
– or rather with the cause for which he is pleading – is that he is presenting the study
of language as a necessary concomitant to the study of literature. For the native
152
E X T E N S I O N

---

speaker, except occasionally and superficially, this is simply not true. It is not true
even for the reading of Chaucer.
This has been a fighting retort and I hope the words I have used will not seem
unnecessarily offensive. But 
pace
Mr Fowler, the things I have said do need to be
said.