Language and Literature
http://lal.sagepub.com/content/5/3/157
The online version of this article can be found at:
DOI: 10.1177/096394709600500301
1996 5: 157
Language and Literature
Adrian Pilkington
Introduction: relevance theory and literary style
Published by:
http://www.sagepublications.com
On behalf of:
Poetics and Linguistics Association
can be found at:
Language and Literature
Additional services and information for
http://lal.sagepub.com/cgi/alerts
http://lal.sagepub.com/subscriptions
http://www.sagepub.com/journalsReprints.nav
http://www.sagepub.com/journalsPermissions.nav
http://lal.sagepub.com/content/5/3/157.refs.html
ARTICLE
Introduction: relevance
theory
and
literary
style
1
Adrian
Pilkington,
Royal Holloway
University
of
London,
UK
Relevance
theory
is
a
theory
of
verbal
communication
grounded
in
a
theory
of
cognition
and,
as
such,
has
influenced
and
contributed
to
research in
a
number of
related
disciplines,
most
notably
linguistics,
psychology
and
philosophy. 2
It also
offers
a
new
theoretical
perspective
on
the
study
of rhetoric and
style,
a
perspective
that
should
encourage
new
lines of research for those
working
at
the
’interface*
of
language
studies and
literary
criticism.
In this introduction I wish
to
suggest
some
ways in which relevance
theory
may
contribute
to
a
better
understanding
of
literary style,
and,
more
particularly,
to
the
understanding
of
poetic
effects.
Before I do that
I
will make
a
few
general
remarks about relevance
theory
as a
pragmatic theory.
It
should be
recognised
that the
account
I
offer here is very
simplified
and
incomplete.
Clark,
in
this
issue,
provides
his
own
brief summary and
more
comprehensive
accounts
of the
theory
are
given
by
works cited in
the
reference.
3
Relevance
theory rejects
both code models of verbal communication and
purely
inferential
accounts.
Utterance
interpretation,
it is
claimed,
involves
two
distinct
phases.
A
context-independent decoding phase yields
semantic
representations
which
provide
a
schematic but
radically incomplete
representation
of the
thoughts
that
are
communicated.
An
inferential
phase
brings
non-linguistic
contextual
information
to
bear upon the
output
of
decoding
to
arrive
at
the
fully-fledged thoughts
that
are
communicated. This inferential
phase
involves
fleshing
out
the semantic
representation
by
resolving ambiguities,
assigning
reference and
enriching
the
content
of
concepts
that contribute
to
the
proposition expressed.
It
also determines the addresser’s attitude
to
the
proposition
expressed
and
leads
to
the derivation of
implicatures.
In
this view
thoughts
are
considerably
richer in
meaning
than the
meanings
of the
linguistic
expressions
that
are
used
to
communicate them.
A
pragmatic theory
has
to
explain
how
context, in the form of
particular
concepts
and contextual
assumptions,
is
accessed
or
constructed,
before it
can
play a
role in
inferencing.
It
should be noted that for relevance
theory
context
construction
is
a
dynamic
part
of
utterance
interpretation:
context
is
not
static
and
pre-given,
as
it is in earlier
code
model
accounts.
The
concepts
that
contribute
to
the
thoughts
communicated
are
also
addresses
at
which
lexical,
logical
and
encyclopaedic
information
are
stored. The
encyclopaedic
entry
attached
to
a
conceptual
address
contains
assumptions carrying
information
about the extension of the
concept
and
culturally
shared information
relating
to
the
concept.
These
assumptions
become available for
use as
contextual
assumptions
in the
inferencing
stage
of
utterance
interpretation,
together
with
assumptions
constructed
on
the
basis
of
perceptual information
derived from the
physical
environment. In the
case
of
literary
communication
they, together
with
158
assumptions
derived
from earlier
parts
of
the text,
are
the
exclusive
source
of
contextual
assumptions.
Utterance
interpretation
is
guided
by
the
search
for
an
interpretation
that is
consistent
with the
principle
of
relevance.
According
to
the
principle
of
relevance every
utterance
carries the
presumption
that its
interpretation
will
provide
a
satisfactory
range of contextual effects for
no
unjustifiable processing
effort.4
Contextual effects
are
achieved
when
new
information interacts with
a
context
of
existing assumptions
in
one
of three ways:
by causing
a
relatively
weakly
held
existing
assumption
to
be
strengthened,
by
contradicting
and
eliminating
an
existing
assumption,
or
by combining
with
an
existing assumption
to
yield
a
contextual
implication.
In the latter
case
the
contextual
assumption
is
a
logical
implication
that is derivable neither from the
new
information
alone,
nor
from
the
context
alone,
but from the
new
information
and
the
context
combined.
New
information
is relevant when it achieves contextual effects in that
context;
its relevance
is
relatively
greater,
the
greater
the contextual effects. The
other
factor
affecting
relevance is the
processing
effort
required
to
decode
an
utterance,
to
access
context
and
to
compute
contextual
effects
in that
context.
In
this
case
the smaller the
processing
effort involved in
interpretation
the
greater
the
relevance.
According
to
this view
one
of the
main
factors
guiding interpretation
is
the
relative
accessibility
of
assumptions:
the
more
accessible
they
are
the easier
they
are
to
process.
The addresser in
fashioning
his
or
her
utterance
takes
into
account
what he
or
she considers
to
be the
concepts
and
assumptions
that
are
most
accessible
to
the addressee. The addressee follows
a
route
of least effort in
using
the
most
accessible
concepts
and
assumptions
until
a
range of contextual
effects
that
the addresser
could
rationally
have
intended is derived. Context is
extended until
such effects
are
achieved.
These
effects
then
constitute
the
interpretation.
An
addresser
might
achieve
particular stylistic
effects
by creating
special
kinds
of
processing
difficulties
for
the addressee. He
or
she
might,
for
example,
encourage the construction of
a
certain context,
which,
once
constructed,
has
to
be
rejected
and
replaced
with
another. Such is the
case
with
jokes,
or
with
humorous
utterances
more
generally.
(See
Jodlowiec
1991 and
Curc6
1995 for
more
detailed
accounts.)
The
pragmatic garden-pathing
involved here is linked
to
a
special
kind of
physiological
response,
culminating
perhaps
in
laughter,
and
to
a
special
kind of
qualitative
response
or
experience -
what it
feels like for
something
to
be
funny.
An addresser may
cause
the addressee
to
engage in
a
more
extensive search
through
context
than is
normally
the
case.
Metaphorical
utterances
typically
communicate
a
range of
assumptions
simultaneously.~ This
set
of
assumptions
constitutes
a
complex thought
that
the addresser wishes
to
communicate.
The
contextual
assumptions
used
to
interpret
a
standard
or
conventional
metaphor
are
relatively easily
accessible. Because the
addressee
is
fairly
confident that the
159
implicatures
these
give
rise
to
are
intended
by
the addresser
they
are
said
to
be
strongly
communicated. Even
so,
it is
usually
not
possible
to
find
adequate
paraphrases
for
such
metaphors
in
terms
of
a
list
of
what
might
be
taken
as
implicatures.
The
metaphor
seems to
lose
some
of
its
expressive
power in the
paraphrase.
This
suggests
that
a
further range of
implicatures
are
more
weakly
communicated.
Where
the addressee is
reasonably
certain that the addresser
wishes
to
communicate
a
certain
implicature,
then it is
said
to
be
a
strong
implicolllre.
Where the
addressee is
given
some
encouragement
to
explore
context
further,
but is less certain
as
to
whether the addresser wishes
to
communicate
the
resulting implicatures,
then these
are
said
to
be wean
iniplicatitres.
The notion of weak
implicature,
developed
within
relevance
theory,
is
crucially important
for
explaining
poetic
effects. Poetic
metaphors,
for
example,
are
characterised in
terms
of
the communication of
a
wide array of
weak
implicatures.
This notion of weak
implicature helps
to
explain
the
indeterminacy
of
poetic
effects
(the
problem
of
saying exactly
what range of
implicatures
are
communicated).
It also
helps
to
explain why
poetic
metaphors
and
other
sources
of
poetic
effects
are
difficult,
if
not
impossible,
to
translate.
As
in
the
case
for
humour,
in the
communication
of
poetic
effects
a
special
kind of
pragmatic
processing
is
encouraged,
in
this
case
involving
a
wide-ranging
search
through
the
encyclopaedic
entries of certain
concepts
for
assumptions
that
might
be used
in
the
interpretation
process. This kind
of
pragmatic
processing might
similarly
be linked
to
a
special
kind of
physiological
response, and
to
a
special
kind
of
qualitative
response
or
experience -
an
aesthetic response
or
experience.
It
should
be noted that
stylistic
effects,
as
described
here,
refer
to
assumptions
communicated
(in
various combinations and
with
varying
degrees
of
strength).
In
terms
used
by
Leech
and
Short
(1981,
Chapter
1),
stylistic
choices
are,
in this
view,
choices
of
matter
or
content
rather
than choices of
manner or
expression.
The
assumptions
communicated
also affect the reader in
specific
ways. A
certain
kind
of
pragmatic processing
is
encouraged,
leading
to
a
certain
kind of
qualitative
response.
This is because the addresser has
a
certain
kind of
complex
thought,
linked
to
certain
qualitative
properties,
that
he
or
she
wishes
to
communicate.
It
should
also be noted that
no
special literary reading strategies,
conventions
or
maxims
are
needed
to
explain
how
poetic
effects
are
communicated,
or
how
literary
communication
works
more
generally.
There
is the
same
on-line
search
for
intepretations
consistent with the
principle
of relevance. What
distinguishes
poetic
effects
is that
extra
processing
effort is
required
because
of
the
lack
of
readily
accessible contextual
assumptions
that
might
lead
to
a
range of
satisfactory
contextual
effects. A
more
extensive
search
through
context
is
encouraged.
Where
addressees find
such
context
rich
enough,
the
extra
processing
effort is
compensated
for
by
a
wider range of contextual effects
than
would
normally
be the
case.
Addressees may find
contexts
rich in different
ways,
of
course,
or
they
may
not
find them rich
at
all.
160
Poets
(and
novelists and
playwrights)
may
spend
a
considerable
amount
of
time
making
sure
that the
right
words
are
placed
in the
right
order.
They
take
this trouble because the
thoughts
that
they
wish
to
communicate
are
extremely
rich and subtle.
Only by adopting
a
theoretical
approach
of the kind that
relevance
theory
offers,
concentrating
on
thoughts
as
complex
sets
of
assumptions
of
varying degrees
of
strength
and
weakness,
interacting
dynamically
on-line with other
sets
of
assumptions,
can
justice
be done
to
the
complexity
of the
thoughts
that may be communicated. As I have
argued
elsewhere
(Pilkington
1992)
the notion of literariness
might
more
satisfactorily
be defined in
terms
of mental
representations
and mental processes, of the
specific
kind described
here,
than in
terms
of the
linguistic
properties
of
texts.
Linguistic foregrounding
does
not
necessarily,
of
itself,
lead
to
poetic
effects.
This
special
issue of
Language
and Literature
is
devoted
to
articles
which
explore
the
implications
of relevance
theory
for the
study
of rhetoric and
literary
style.
In the
case
of rhetoric relevance
theory
has
reanalysed
tropes
and offered
new
accounts
of how
they
are
interpreted
in
the
light
of
the
more
general theory
of communication and in
a
way that is consistent with
current
psycholinguistic
research. This
reanalysis rejects
the Gricean view that
tropes
deviate from
a
literal
norm.6
Relevance
theory
uses a
number
of
new
concepts
to
explain
them,
most
importantly
the notions of
irrterpretative
use
and,
in
the
case
of
irony,
echoic
rrse.
These
concepts
are
explained
and
illustrated in the articles that follow.
Papafragou
and Vicente tackle theoretical issues
relating
to
metonymy
and
metaphor
respectively. They
are
both
concerned
with
a
cognitive
explanation
of
how these
tropes
are
interpreted. Papafragou
shows
that
there
is
a
stylistic
continuum from conventional
through
to
creative
uses
of
metonymy.
She
develops
an
account
of creative
uses
of
metonymy
using
the
notion of
interpretive
use.
She also discusses the process of semanticisation that
metonymy
may
undergo.
In
particular
this article raises
interesting
theoretical issues
for the
field of lexical semantics and for the notion of
interpretative
use
within
relevance
theory.
One incidental feature of
Papafragou’s
article,
which I would like
to
draw
attention
to
here,
is her criticism of Lakoff and Johnson’s
(1980)
account
of
metonymy.
Lakoff and associates
are
mainly
famous for their
account
of
metaphor.
Lakoff and Turner
(1989),
in
particular,
is
a
key
text
that has been
extremely
influential for research into
poetic metaphor.
Given the influence of
Lakoff’s
work,
it is
important
to
point
out
the differences between the
two
approaches,
as
Papafragou
does
(indicating problems
that the Lakoff
account
has
in
explaining
some
of the
data),
and
to
develop
a
debate between the
two
accounts.
One of the main differences would appear
to
be that Lakoff and
associates
are
concerned with structural
questions
about how
conceptual
domains
map
onto
each other rather than with the process of
interpretation.
Vicente’s article
contrasts
a
relevance-theoretic
account
of
metaphor
with
semantic and Gricean
accounts.
The
main interest of this article is the
consideration it
gives
to
alternative
accounts
of
metaphor
that may be
consistent
161
with
the
general
relevance
theory
account
of
verbal communication. In
particular,
arguments
are
considered for the view that
metaphors
contribute ad
hoc
concepts
directly
to
the
proposition
expressed.
Both of these articles
are
concerned with
general
accounts
of
metonymy
and
metaphor.
But both take
note
of the
stylistic
differences between standard and
creative
uses
of these
tropes
and offer
some
explanation
for these differences.
These
discussions,
although they
do
not
focus
directly
on
literary style,
are
of
fundamental
importance
to
an
understanding
of how these
tropes
are
used
to
achieve
poetic
effects.
The
articles
by
MacMahon
and Clark
are more
directly
concerned with the
application
of
relevance
theory
to
literary
texts.
MacMahon
discusses the
use
of
voice
in
literature
in
terms
of the relevance-theoretic notion
of echoic
interpretation.
The
article
stresses
that
the
flow
of
ideas
and
insights
between
pragmatic theory (specifically
relevance
theory)
and
literary
studies
is
potentially
a
highly productive
one.
Clark’s
article discusses
one
way in which relevance
theory
may
suggest
ideas
for
the
teaching
of
literary stylistics,
as
well
as
for the
teaching
of
semantics and
pragmatics.
He also
presents
an
argument
in favour of
developing
an
’interface’
between
relevance
theory
pragmatics
and
literary
studies and
suggests,
in
particular,
that the
study
of
literary
texts
should
raise useful
questions
and
be
a
good testing ground
for
relevance
theory.
There
is
clearly
room
for
disagreement
with
regard
to
the theoretical
accounts
developed
for
tropes
and for ideas
concerning
the
possible applications
of
relevance
theory
to
literary
studies.
The
articles
here
(including
this
introduction)
in
no
way
represent
an
orthodoxy.
While the contributors
to
this issue
might
want to
question
various
points argued
for in the other articles
presented
here,
there is nevertheless
a
basic
agreement
that relevance
theory
does
present
a
radically
new
theoretical
perspective
on
the
study
of rhetoric and
style
that
not
only
offers
extremely
rich
insights,
but also allows
questions
to
be asked which
were
previously
either
impossible
or
difficult
to
pursue.
Notes
1.
I
am
very
grateful
to
Robyn
Carston,
Billy
Clark and Katie
Wales for discussion
and
comments.
2.
For
a
good
bibliography
see
Sperber
and
Wilson
(1995).
3.
The
key
text
is
Sperber
and Wilson
(1995).
This
second
edition,
to
quote
the
blurb
on
the back
cover,
includes
a
new
Postface
outlining developments
in
Relevance
Theory
since
1986,
discussing
the
more
serious
criticisms
of the
theory,
and
envisaging
possible
revisions
or
extensions’.
It also
provides
a
comprehensive
bibliography
of work
in relevance
theory. Sperber
and
Wilson
(1987)
provides
a
useful
precis
of
relevance
theory, prior
to
the
envisaged
’possible
revisions
or
extensions’. This is followed
by
peer commentaries and
a
response
to
the
commentaries. Wilson
(1994)
provides
another useful
summary of the
theory
and Blakemore
(1992)
provides
a
textbook
account.
A
special
issue
of the
journal
Lingua
(1992)
contains
articles
on a
variety
of
aspects
of relevance
theory,
including
several
on
rhetoric and
style.
For work
more
specifically
in the
areas
of
rhetoric
and
style
see
in
particular Sperber
and
Wilson
(1985/6)
on
metaphor,
Wilson and
Sperber
(1992)
on
irony,
and
Sperber
and
Wilson
162
(1990)
on
rhetoric in
general.
Blakemore
(1989)
and
(1993)
discuss
general
aspects
of style.
Forceville
(1994)
discusses
pictorial
metaphor
in
advertising.
Reboul
(1992)
uses
relevance
theory
to
discuss
aspects
of
literary
style.
Fabb
(forthcoming)
makes
use
of relevance
theory
in
chapters
on
metaphor
and
irony
in literature.
4.
What
is
now
referred
to
as
the second
principle
of relevance
applies
to
all
acts
of
ostensive-inferential
communication.
5.
See
Sperber
and
Wilson
(1985/6)
for
a
much
fuller
account
than
can
be
given
here.
Vicente
in
this issue deals in greater
detail
with
metaphor.
6.
An
example
of
psycholinguistic
research that
focuses
on
tropes
and
rejects
’the literal
meaning
hypothesis’
can
be
found,
for
example,
in Gibbs
(1994).
7.
See Carston
(forthcoming)
for
a more
detailed discussion
of
the
view
that ’loose uses’
(including
metaphor)
contribute
to
the
proposition
expressed
by
an
utterance.
References
Blakemore,
D.
(1989)
Linguistic
form
and
pragmatic interpretation:
the
explicit
and the
implicit,
in
L.
Hickey
(ed.)
The
Pragmatics
of Style,
Routledge, London,
pp. 28-51
Blakemore,
D.
(1992) Understanding
Utterances: An Introduction
to
Pragmatics,
Biackwell,
Oxford
Blakemore,
D.
(1993)
The relevance
of
reformulations,
Language
and Literature
2.(2):
101-20
Carston,
R.
(forthcoming)
Enrichment
and
loosening:
complementary
processes in
deriving
the
proposition expressed,
UCL
Working Papers
in
Linguistics,
London
Curcó,
C.
(1995)
Some
observations
on
the
pragmatics
of
humorous
interpretations:
a
relevance
theory
approach,
UCL
Working Papers
in
Linguistics,
7:
27-47
Fabb,
N.
(forthcoming)
Linguistics
and
Literature, Blackwell,
Oxford
Forceville,
C.
(1994)
Pictorial
Metaphor
in
Advertising, Vrije
Universiteit,
Amsterdam
Gibbs,
R.
(1994)
The Poetics
of Mind: Figurative
Thought, Language
and
Understanding,
Cambridge University
Press,
Cambridge
Jodtowiec,
M.
(1991)
What makes
jokes
tick,
UCL
Working Papers
in
Linguistics
3: 241-53
Lakoff,
G. and
Johnson,
M.
(1980)
Metaphors
We
Live
By, Chicago University
Press,
Chicago
Lakoff,
G. and
Turner, M. (1989)
More
than Cool
Reason: A
Field
Guide
to
Poetic
Metaphor,
Chicago University
Press,
Chicago
Leech,
G.
and
Short,
M.
(1981)
Style
in
Fiction:
A
Linguistic
Introduction
to
English
Fictional
Prose,
Longman,
London
Pilkington,
A.
(1992)
Poetic
effects, Lingua
87: 29-51
Reboul,
A.
(1992)
Rhetorique
et
Stylistique
de
la
Fiction,
Presses
Universitaires
de
Nancy,
Nancy
Sperber,
D. and
Wilson,
D.
(1985/86)
Loose talk,
Proceedings of
the Aristotelian
Society
LXXXVI:
153-71.
Reprinted
in S. Davis
(ed.)
(1991) Pragmatics:
A
Reader,
Oxford
University
Press,
Oxford,
pp.
540-9
Sperber,
D. and
Wilson,
D.
(1986)
Relevance:
Communication
and
Cognition,
Blackwell,
Oxford
iSperber,
D.
and
Wilson,
D.
(1987) Presumptions
of
relevance,
The
Behavioural
and Brain Sciences
3(1):
179-84
Sperber,
D.
and
Wilson,
D.
(1990)
Rhetoric
and
relevance,
in J. Bender and D.
Wellbery
(eds)
The
Ends
of Rhetoric:
History, Theory,
Practice,
Stanford
University
Press,
Stanford, CA,
pp.
140-56
Sperber,
D.
and
Wilson,
D.
(1995) (2nd edn)
Relevance:
Communication
and
Cognition,
Blackwell,
Oxford
Wilson,
D.
(1994)
Relevance and
understanding,
in
G. Brown et
al.
(eds) Language
and
Understanding,
Oxford
University
Press,
Oxford,
pp.
35-58
Wilson,
D,
and
Sperber,
D.
(1992)
On verbal
irony,
Lingua
87: 53-76.
Reprinted
in J.J. Weber
(ed.)
(1996)
The
Stylistics
Reader: From
Roman Jakobson
to
the
Present,
pp.
260-79