Supplementary Readings

Chapter 9

Language and Literature

Text 1. Rhetoric, Stylistics and Modern Criticism (Bradford)

Text 2. Development in Stylistics (Simpson)
Text 3. Foregrounding and Interpretation (Leech)
Text 4. How to Analyze English Fictional Prose (Leech and Short)
Text 5. Foreword to Cognitive Stylistics (Semino and Culpeper)

Text 1

Rhetoric, Stylistics and Modern Criticism

Richard Bradford

Rhetoric
The academic discipline of stylistics is a twentieth-century invention. It will be the purpose of this
book to describe the aims and methods of sty1istics, and we will begin by considering its
relationship with its most notable predecessor -- rhetoric.
The term is derived from the Greek techne rhetorike, the art of speech, an art concerned with the
use of public speaking as a means of persuasion. The inhabitants of Homer's epics exploit and,
more significantly acknowledge the capacity of language to affect and determine non-linguistic
events, but it was not until the fifth century BC that the Greek settlers of Sicily began to study
document and teach rhetoric as a practical discipline. The best-known names are Corax and Tisias
who found that, in an island beset with political and judicial disagreements over land and civil
rights, the art of persuasion was a useful and profitable profession. Gorgias, one of their pupils,
visited Athens as ambassador and he is generally regarded as the person responsible for piloting
rhetoric beyond its judicial function into the spheres of philosophy and literary studies. Isocrates
was the first to extend and promote the moral and ethical benefits of the art of speech, and one of
Plato's earliest Socratic dialogues bears the name Gorgias. It is with Plato that we encounter the
most significant moment in the early history of rhetoric. In the Phaedrus Plato/Socrates states that
unless a man pays due attention to philosophy 'he will never be able to speak properly about
anything' (261A). A real art of speaking, which does not seize hold of truth, does not exist and
never will' (260E). What concerned Plato was the fact that rhetoric was a device without moral or
ethical subject matter. In the Gorgias he records an exchange between Socrates and Gorgias in
which the former claims that persuasion is comparable with flattery, cooking and medicine: it
meets bodily needs and satisfies physical and emotional desires. Rhetoric, he argues, is not an 'art'
but a 'routine', and such a routine, if allowed to take hold of our primary communicative medium,
will promote division, ambition and self aggrandizement at the expense of collective truth and
wisdom, the principal subjects of philosophy Plato himself, particularly in the Phaedrus does not
go so far as to suggest the banning of rhetoric; rather he argues that it must be codified as
subservient to the philosophers search for truth.
Aristotle in his Rhetoric (c. 330 BC) produced the first counter-blast to Plato's anti-rhetoric thesis.
Rhetoric, argues Aristotle, is an art, a necessary condition of philosophical debate. To perceive the
same fact or argument dressed in different linguistic forms is not immoral or dangerous. Such a
recognition -- that words can qualify or unsettle a single pre-linguistic truth -- is part of our
intellectual training, vital to any purposive reconciliation of appearance and reality. Aristotle meets
the claim that rhetoric is socially and politically dangerous with the counter claim that the
persuasive power of speech is capable of pre-empting and superseding the violent physical
manifestations of subjection and defence.

The Plato--Aristotle exchange is not so much about rhetoric as an illustration of the divisible
nature of rhetoric. It is replayed, with largely Aristotelian preferences, in the work of the two most
prominent Roman rhetoricians, Cicero and Quintillion; it emerges in the writings of St Augustine
and in Peter Ramus's Dialectique (1555), one of the founding moments in the revival of classical
rhetoric during the European Renaissance. Most significant1y, it operates as the theoretical spine
which links rhetoric with modern stylistics, and stylistics in turn with those other constituents of
the contemporary discipline of humanities: linguistics, structuralism and poststructuralism.
Plato and Aristotle did not disagree on what rhetoric is; their conflicts originated in the
problematical relationship between language and truth. Rhetoric, particularly in Rome and in
post-Renaissance education, had been taught as a form of super- grammar It provides us with
names and practical explanations of the devices by which language enables us to perform the
various tasks of persuading, convincing and arguing. In an ideal world Aristotle's thesis) these
tasks will be conducive to the personal and the collective good. The rhetorician will know the truth,
and his linguistic strategies will be employed as a means of disclosing the truth. In the real world
(Plato's thesis) rhetoric is a weapon used to bring the listener into line with the argument which
happens to satisfy the interests or personal affiliations of the speaker, neither of which will
necessarily correspond with the truth. These two models of rhetorical usage are equally valid and
finally irreconcilable. Lies, fabrications, exaggerations are facts of language, but they can only be
cited when the fissure between language and truth is provable.
For example, if I were to tell you that I am a personal friend of Aristotle, known facts will be
sufficient to convince you (unless you are a spiritualist) that I am not telling the truth. However, a
statement such as, Aristotle speaks to me of the general usefulness of rhetoric' is acceptable
because it involves the use of a familiar rhetorical device (generally termed catachresis, the
misuse or mis-application of a term): Aristotle does not literally speak to me, but my use of the
term to imply that his written words involve the sincerity or the immediate relevance of speech is
sanctioned by rhetorical stylistic convention. What I have done is to use a linguistic device to
distort pre-linguistic truth and to achieve an emotive effect at the same time. My reason for doing
so would at the same time be to give a supplementary persuasive edge to the specifics of my
argument about the validity of Aristotle's thesis. Such devices are part of the fabric of everyday
linguistic exchange and, assuming that the hearer is as conversant as the speaker with the
conventions of this rhetorical game, they are not, in Plato's terms, immoral or dishonest. But for
Plato such innocuous examples were merely a symptom of the much more serious consequences
of rhetorical infection. The fact that Aristotle lived more than two millennia before me cannot be
disputed, but the fabric of intellectual activity and its linguistic manifestation is only partly
comprised of concrete facts.
Morality, the existence of God, and the nature of justice: all of these correspond with the verifiable
specifics of human existence, but our opinions about them cannot be verified in direct relation to
these specifics. The common medium shared by the abstract and the concrete dimensions of
human experience is language and, as a consequence, language functions as the battleground for
the tendentious activity of making the known correspond with the unknown, that speculative
element of human existence that underpins all of our beliefs about the nature of truth, justice,
politics and behaviour. Plato and Aristotle named the conditions of this conflict as dianoia and
pragmata (thought and facts, otherwise, known as res or content) and lexis and texis (word choice
3

and arrangement, otherwise known as verba or form), and the distinction raises two major
problems that will occupy much of our attention throughout this book.
First of all it can be argued that to make a distinction between language -- in this instance the
rhetorical organization of language-- and the pre-1inguistic continuum of thought, objects and
events involves a fundamental error. Without language our experience of anything is almost
exclusively internalized and private: we can, of course, make physical gestures, non-linguistic
sounds or draw pictures, but these do not come close to the vast and complex network of signs and
meanings shared by language users. The most important consequence of this condition of
language dependency is that we can never be certain whether the private world, the set of private
experiences or beliefs, that language enables us to mediate is, as Plato and Aristotle argue, entirely
independent of its medium. The governing precondition for any exchange of views about the
nature of existence and truth -- a process perfectly illustrated by Plato's Socratic dialogues -- is
that language allows us to disclose the true nature of pre-linguistic fact. However, for such an
exchange to take place at al1 each participant must submit to an impersonal system of rules and
conventions. Before any disagreement regarding a fact or a principle can occur the combatants
must first have agreed upon the relation between the fact/principle and its linguistic enactment. An
atheist and a Christian will have totally divergent perceptions of the nature of human existence,
but both will know what the word 'God' means.
The twentieth-century alternative to Aristotle's and Plato's distinction between dianoia/pragmata
and lexis/taxis has been provided by Ferdinand de Saussure, a turn-of-the-century linguist whose
influence upon modern ideas about language and reality has become immeasurable. Saussure's
most quoted and influential propositions concern his distinction between the signified and the
signifier and his pronouncement that 'in language there are only differences without positive terms'.
The signifier is the concrete linguistic sign, spoken or written, and the signified is the concept
represented by the sign. A third element is the referent, the pre-linguistic object or condition that
stands beyond the signifier/signified relationship. This tripartite function is, to say the least,
unsteady. The atheist and the Christian will share a largely identical conception of the relation
between 'God' (signifier) and 'God' (signified) but the atheist will regard this as a purely linguistic
state, a fiction sustained by language, but without a referent. For such an individual the signifier
God relates not to a specific signified and referent, but to other signifiers and signifieds -- concepts
of good and bad, eternity, omniscience, omnipotence, the whole network of signs which enables
Christian belief to intersect with other elements of the human condition. In Saussures terms, the
signified 'God' is sustained by the differential relationship between itself and other words and
concepts, and this will override its correspondence with a positive term (the referent). Plato and
Aristotle shared the premise that it is dangerous and immoral to talk about something that does not
exist, and that it is the duty of the philosopher to disclose such improper fissures between language
and its referent. Saussure's model of language poses a threat to this ideal by raising the possibility
that facts and thoughts might, to an extent, be constructs of the system of language.
The relation between Classical philosophy/rhetoric and Saussurean linguistics is far more
comp1icated than my brief comparison might suggest, but it is certain that Saussure makes explicit
elements of the divisive issue of whether rhetoric is a potentially dangerous practice. And this
leads us to a second problem: the relationship between language and literature. Plato in The
Republic has much to say about literature -- which at the time consisted of poetry in its dramatic or
4

narrative forms. In Book 10 an exchange takes place regarding the nature of imitation and
representation: the subject is ostensibly art, but the originary motive is as usual the determining of
the nature of truth. By the end of the dialogue Socrates has established a parallel hierarch of media
and physical activities. The carpenter makes the actual bed, but the idea or concept behind this act
of creation is God's. The painter is placed at the next stage down in this creative hierarchy: he can
observe the carpenter make the bed and dutifully record this process. The poet, it seems, exists in a
somewhat ambiguous relation to this column of originators, makers and imitators.
Perhaps they [poets] may have come across imitators and been deceived by them; they may not have
remembered when they saw their works that these were but imitations thrice removed from the truth,
and could easily be made without any knowledge of the truth, because they are appearances only and not
realities. (1888: 3l2)

In short, the poet is capab1e of unsett1ing the hierarchy which sustains the clear relation between
appearance and reality. Poets, as Aristotle and Plato recognized, are pure rhetoricians: they work
within a kind of metalanguage which draw continuously upon the devices of rhetoric but which is
not primarily involved in the practical activities of argument and persuasion. As the above quote
suggests, they move disconcertingly through the various levels of creation, imitation and
deception, and as Plato made clear, such fickle mediators were not the most welcome inhabitants
in a Repub1ic founded upon a clear and unitary correspondence between appearance and reality.
Plato's designation of literature as a form which feeds upon the devices of more practical and
purposive linguistic discourses, but whose function beyond a form of whimsical diversion is
uncertain, has for two millennia been widely debated but has remained the dominant thesis.
During the Eng1ish Renaissance there was an outpouring of largely practical books on the proper
use of rhetoric and rhetorical devices: for example R. Sherrys A Treatise of Schemes and Tropes
(1550), T. Wilsons The Arte of Rhetorique (1553), R. Rainolde's A Book Called the Foundation of
Rhetorike (1563), H. Peachams The Garden of Eloquence (l577) and G. Puttenhams The Arte of
English Poesie (l589). These were aimed at users of 1iterary and non-literary language, but a
distinction was frequently made between the literary and the non-literary function of rhetoric. In
George Puttenhams The Arte of English Poesie we find that there are specific regulations
regarding the correspondence between 1iterary style and subject (derived chiefly from Cicero's
distinction between the grand style, the middle style and the low, plain or simple style). The
crossing of recommended style-subject borders was regarded as bad writing, but a far more serious
offence would be committed if the most extravagant rhetorical, and by implication literary, devices
were transplanted into the serious realms of non-literary exchange. Metaphors or 'figures' are,
according Puttenham, particularly dangerous. 'For what else is your metaphor but an inversion of
sense by transport; your allegorie by a duplicitie of meaning or dissimulation under covert and
dark intendments' (l589: 158). Judges, for example, forbid such extravagances because they distort
the truth:
This no doubt is true and was by then gravely considered; but in this case, because our maker or Poet is
appointed not for a judge, but rather for a pleader, and that of pleasant and love1y causes and nothing
peri1lous, such as be for the trial of life, limme, or livelihood they [extravagant metaphors] are not in
truth to be accompted vices but for vertues in the poetical science very commendable. (ibid.: 161)

Poetry does of course involve 'perillous' matters, but what Puttenham means is that the poetic
5

function is not instrumental in activities concerned with actual 'life, limme, or livelihood'. As a
spokesman for the Renaissance consensus Puttenham shows that the Plato/Aristotle debate
regarding the dangers of rhetoric, especially in its literary manifestation, has been shelved rather
than resolved: in short, Puttenham argues that in literature it is permissible to distort reality
because literature is safely detached from the type of discourse that might have some purposive
effect upon the real conditions of its participants. What Puttenham said in 1589 remains true today:
literary and non-literary texts might share a number of stylistic features but literary texts do not
belong in the same category of functional, purposive language as the judicial ruling or the
theologied tract. This begs a question which modern stylistics, far more than rhetoric, has sought
to address. How do we judge the difference between literary and non-literary discourses? We have
not finished with rhetoric, but in order to properly consider the two issues raised by it the
relation between language and non-linguistic reality and the difference between literary and
non-1iterary texts -- we should now begin to examine its far more slippery and eclectic modern
counterpart.
Stylistics and Modern Criticism
Two groups of critics have had a major influence on the identity and direction of twentieth-century
English studies: the Russian and central European Formalists and the more disparate collection of
British and American teachers and writers whose academic careers began during the l920s and
l930s. The term New Criticism is often applied to the latter group. The objectives of the majority
of individuals in each group were the same: to define literature as a discourse and art form and to
establish its function as something that can be properly studied. Until the late 1950s the work of
these groups remained within mutually exclusive geographical and academic contexts: the New
Critics in Britain and America and the Formalists in Europe. During the 1960s New Criticism and
Formalism began to recognize similarities and overlaps in their goals and methods. Since the
1960s their academic predominance has been unsettled by a much broader network of
interdisciplinary practices: structuralism, poststructuralism, feminism and new historicism, are all
significant elements of contemporary 1iterary studies, and each draws its methodologies and
expectations from intellectual fields beyond the traditional, enc1osed realms of rhetoric and
aesthetics.
This, I concede, is a simplified history of twentieth-century criticism, but it provides us with a
framework for an understanding of how rhetoric has been variously transformed into modern
stylistics. The New Critics and the Formalists are the most obvious inheritors of the disciplines of
rhetoric, in the sense that they have maintained a belief in the empirical difference between
literature and other types of language and have attempted to specify this difference in terms of
style and effect. Structuralism at once extended and questioned these practices by concentrating on
the similarities, rather than the differences, between literature and other discourses.
Poststructuralism took this a stage further by introducing the reader into the relation between
literary and non-1iterary style, and posing the question of whether the expectations of the
perceiver can determine, rather than simply disclose, stylistic effects and meanings. Feminist
critics have examined style less as an enclosed characteristic of a particular text and more as a
reflection of the sociocu1tural hierarchies -- predominantly male -- which control stylistic habits
and methods of interpretation. Similarly, Marxists and new historicists concern themselves with
style as an element of the more important agenda of cultural and ideological change and mutation.
6

For the sake of convenience I shall divide these different approaches to stylistics into two basic
categories: textualist and contextualist. The Formalists and New Critics are mainly textualists in
that they regard the stylistic features of a particular 1iterary text as productive of an empirical
unity and completeness. They do not perceive literary style as entirely exclusive to literature -rhythm is an element of all spoken language, and narrative features in ordinary conversation -- but
When these stylistic features are combined so as to dominate the fabric of a text, that text is
regarded as literature. Contextualism involves a far more 1oose and disparate collection of
methods. Its unifying characteristic is its concentration on the relation between text and context.
Some structuralists argue that the stylistic features of poetry draw upon the same structural
frameworks that enable us to distinguish between modes of dress or such social rituals as eating.
Some feminists regard literary style as a means of securing attitudes and hierarchies that in the
broader context, maintain the difference between male and female roles.
The remainder of this Part is divided into three chapters. The first two will examine in basic terms
how modern criticism has employed stylistics to evolve theories of poetry and fiction: these
chapters will be concerned predominantly with textualist method and practice. Chapter 5 is more
concerned with contextualism and will consider the way in which the interface between text and
context can unsettle textualist assumptions.
(from Richard Bradford. (1997). Stylistics. PP. 3-14. London & New York: Routledge)

Text 2

Development in Stylistics
Paul Simpson

This unit looks at some of the important influences on stylistics that have helped to shape its
development over the years. From the classical period onwards there has been continued healthy
interest among scholars in the relationship between patterns of language in a text and the way a
text communicates. The Greek rhetoricians, for example, were particularly interested in the tropes
and devices that were used by orators for effective argument and persuasion, and there is indeed a
case for saying that some stylistic work is very much a latter-day embodiment of traditional
rhetoric. However, there is one particular field of academic inquiry, from the early twentieth
century, that has had a more direct and lasting impact on the methods of contemporary stylistics.
This field straddles two interrelated movements in linguistics, known as Russian Formalism and
Prague School Structuralism. Of the former movement, key figures include Viktor Shklovsky and
Boris Tomashevsky; of the latter, Jan Mukarovsky and Wilhelm Mathesius. One scholar, whose
work literally links both movements, is Roman Jakobson, who moved from the Moscow circle to
the Prague group in 1920. Many of the central ideas of these two schools find their reflexes in
contemporary stylistics and two of the more durable theoretical contributions are the focus of this
unit. These are the concept of foregrounding and the notion of the poetic function in language.
Foregrounding
Foregrounding refers to a form of textual patterning which is motivated specifically for
literary-aesthetic purposes. Capable of working at any level of language, foregrounding typically
involves a stylistic distortion of some sort, either through an aspect of the text which deviates from
a linguistic norm or, alternatively, where an aspect of the text is brought to the fore through
repetition or parallelism. That means that foregrounding comes in two main guises: foregrounding
as 'deviation from a norm' and foregrounding as 'more of the same'. Foregrounding is essentially a
technique for 'making strange' in language, or to extrapolate from Shklovsky's Russian term
ostranenie, a method of 'defamiliarisation' in textual composition.
Whether the foregrounded pattern deviates from a norm, or whether it replicates a pattern through
parallelism, the point of foregrounding as a stylistic strategy is that it should acquire salience in
the act of drawing attention to itself. Furthermore, this salience is motivated purely by literary
considerations and as such constitutes an important textual strategy for the development of images,
themes and characters, and for stimulating both effect and affect in a text's interpretation.
Foregrounding is not, therefore, the simple by-product of this or that writer's idiosyncratic
predilections in style. For example, Jonathan Swift, a writer with much to say about language and
style, was reputedly never very fond of words which were made up of only one syllable. Whereas
the relative scarcity of monosyllabic words in Swift's work might therefore be noticeable or salient,
it is rather more a consequence of the personal stylistic foibles of the writer than of a carefully
modulated design in literary foregrounding. In sum, if a particular textual pattern is not motivated
for artistic purposes, then it is not foregrounding.
8

The theory of foregrounding raises many issues to do with the stylistic analysis of text, the most
important of which is probably its reliance on the concept of a 'norm' in language. Given the
functional diversity of language, it is very difficult - if not impossible -- to say what exactly a
'normal' sentence in English actually' is. This constitutes a substantial challenge to foregrounding
theory because the theory presupposes that there exists a notional linguistic yardstick against
which a particular feature of style can be measured. A related issue concerns what happens when a
one deviant pattern becomes established in a text - Does it stay foregrounded for the entire
duration of the text? Or does it gradually and unobtrusively slip into the background?
One way of addressing these important questions is through a short illustration. Unit C8 of this
book develops a workshop in practical stylistics which is based on a passage from Ernest
Hemingway's novella The Old Man and the Sea (l952). That passage arguably typifies
Hemingway's written style, a style which literary critics have described with epithets like 'flat',
'dry, 'restrained', 'journalistic' or even 'tough guy' (see C8). These observations are largely based
on a perceived scarcity of adjectives in the writer's work, which is correlated with the 'machismo'
feel of much of his narrative style. It is indeed true that in the first few lines of the passage
analyzed in C8, almost all of the nouns receive no adjectival modification at all: 'the tuna', 'the
stern', 'the gaff', 'the line' and 'the fish'. Let us accept for the moment, then, that this marked
non-adjectival pattern is foregrounded because it deviates from our expectations about the 'normal'
style of twentieth-century prose fiction.
Such an interpretation immediately raises two interconnected problems. The first, as noted above,
concerns the degree to which the 'no-adjective' pattern is able to stay foreground before it
gradually slips into the background. The second is about what would happen should a phrase that
did contain adjectives suddenly appear in the text; that is, should a structure occur whose very use
of adjectives goes against the foregrounded pattern. As it happens, there is elsewhere in the
novella a rather startling example of such a deviation. When a poisonous jellyfish approaches the
old man's boat, the narrative refers to it is 'the purple, formalized iridescent gelatinous bladder of a
Portuguese man-of-war' (Hemingway 1960 [1952]: 28). This is stylistically somewhat of a
quantum leap insofar as the simple article-plus-noun configuration gives way here to a sequence
of not one but four adjectives which are built up before the main noun ('bladder). The old
fishermans superstitious mistrust of the dangerous animal, the 'whore of the sea' as he puts it, is
captured in a stylistic flourish and with a type of hyperbole that would not be out of place in a D.
H. Lawrence novel. The upshot of this is that foregrounding can be seen to work on two levels,
both across and within texts. Whereas Hemingways so-called 'flat' noun phrases may be
foregrounded against the notional external stylistic backdrop of the twentieth century novella, their
repetition in the text develops a norm which is itself susceptible to violation. This type of
secondary foregrounding, known as internal foregrounding, works inside the text as a kind of
deviation within a deviation. Moreover, it is clear that foregrounding does not stand still for long
and that a writer's craft involves the constant monitoring and (re)appraisal of the stylistic effects
created by patterns in both the foreground and in the background. The concept of foregrounding
will be further explored and illustrated in B2.
Jakobson's 'Poetic Function'
In a famous paper, that still reverberates in much of todays stylistic scholarship, the structuralist

poetician Roman Jakobson proposes a model of language which comprises six key functions
(Jakobson 1960). These are the conative, phatic, referential, emotive, poetic and metalingual
functions of language. Alongside the referential function (the content carrying component of a
message) and the emotive function (the expression of attitude through a message), there is one
function that stands out in respect of its particular appeal to stylisticians. This is the poetic
function, which Jakobson defines thus: 'the poetic function projects the principle of equivalence
from the axis of selection into the axis of combination' (Jakobson l960: 358). This rather terse
formula is not the most transparent definition you are likely to come across in this book, so some
unpacking is in order. As a short demonstration of the formula at work, consider first of all the
following example which is the opening line of W. H. Auden's elegiac poem 'In Memory of W. B.
Yeats' (l939). A key verb in the line bas been removed and you might wish to consider what (sort
of word would make an appropriate entry.
He _________ in the dead of winter
There are clearly many items that might go into the slot vacated in the line. That said, some words,
if semantically compatible, can be ruled out on the grounds of inappropriateness to the context.
However, the missing verb is actually disappeared, reinstated here for clarity:
He disappeared in the dead of winter
The technique of blanking out a word in a line, a cloze in stylistics parlance, is to force us to think
about the pool of possible lexical entries from which a choice is ultimately made. This pool of
available words is what Jakobson means by his term axis of selection. What is significant about
Auden's selection, one word taken from many possibilities, is that it engenders a series of
resonances across the line as a whole. Notice for example, how the three syllable word
disappeared creates associative phonetic links with other words in the line. Most obviously, its
initial and final consonant /d/ alliterates with those in the same position in the word dead later in
the line. Possibly more subtly, its third, stressed syllable (disappeared) contains a diphthong, the
first element of which is the vowel /i:/, the same as the vowel in 'He'. This type of vowel harmony,
known as assonance further consolidates points of equivalence across the poetic line.
However, there are also semantic as well as phonetic transferences in the line. Notice how it is the
season, winter, which takes over the semantic quality of death and, when positioned together in the
same grammatical environment, the words dead and 'disappeared' enable new types of
signification to emerge. More specifically, the parallel drawn between the words opens a
conceptual metaphor (All) where the concept of death is represented in terms of a journey. In fact,
we commonly invoke this DEATH IS A JOURNEY metaphor in everyday interaction when we
talk of the dearly departed, or of someone passing away or going to a final resting place. The
point about Audens technique, though, is that this is a novel metaphor (B11), suggesting the sense
of being lost or of straying from a journey, and this is brought about subtly by the implied
connection between the process or disappearing, and the references to death and the seasons
elsewhere in the same 1ine. In sum, the way Auden uses language is a good illustration of
Jakobson's poetic function at work: the particular language patterns he develops work to establish
connections (a principle of equivalences) between the words he chooses from the pool of possible
words (the axis of selection) and the words that are combined across the poetic line (the axis of
combination).
10

Summary
It is important to footnote the foregoing discussion with a rider. Whereas many of the precepts of
both the Formalist and Prague School movements have had a significant bearing on the way
stylistics developed, this is not for a moment to say that stylisticians have embraced these ideas
unequivocally, unanimously or without debate. We have already touched upon some of the
theoretical problems associated the theory of foregrounding, and in this context, stylisticians like
van Peer (l986) and Cook (l994) have made advances in solidifying the foundation of this
generally useful concept. Amongst other things, their work has incorporated cognitive and
psychological models of analysis to explain how text-processors perceive foregrounding in texts
(see further B10).
Application of the concept of the poetic function in language also brings with it an important
caveat. Although not articu1ated especially clearly by Jakobson, it is essential to view the poetic
function not as an exclusive property of literature but rather as a more generally creative use of
language that can pop up, as it were in a range of discourse contexts. One consequence of seeing
that poetic function as it tends to separate off interactive from other uses of language, to separate
off literature from other uses of language, and this is not a desired outcome in stylistic analysis.
This latter issue will come more to the fore in the next unit along this strand, C3, while the unit
below provides an opportunity, through the analysis of a short poem, to investigate and illustrate
further the concept of foregrounding.
(from Paul Simpson. 2004. Stylistics A Resource Book for Students. PP. 50-53. London & New
York: Routledge)

11

Text 3

Foregrounding and Interpretation

Geoffrey Leech

Poetrys unnatural', said Mr. Weller;' No one ever talked poetry 'cept a beadle on boxin' day. In
concentrating on the abnormalities of poetic language in Chapter 3, we saw that there is truth, in a
sense, in at least the first part of Mr. Weller's remark. But what we have to consider in this chapter
is something beyond Mr. We1ler's matter-of-fact wisdom: how the apparently unnatural, aberrant,
even nonsensical, is justified by significance at some deeper level of interpretation. This question
has been raised informally in earlier chapters, especially in connection with the Examples for
Discussion, for to have tried to separate deviance altogether from significance would have been a
very artificial exercise. But we need to give the subject more careful attention.
l. FOREGROUNDING
First, however, I wish to place linguistic deviation in a wider aesthetic context, by connecting it
with the general principle of FOREGROUNDING.
1.1 Foregrounding in Art and Elsewhere
It is a very general principle of artistic communication that a work of art in some way deviates
from norms which we, as members of society, have leant to expect in the medium used. A painting
that is representational does not simply reproduce the visual stimuli an observer would receive if
he were looking at the scene it depicts: what is artistically interesting is how it deviates from
photographic accuracy, from simply being a 'copy of nature '. An abstract painting, on the other
hand, is interesting according to how it deviates from mass-produced regularities of pattern, from
absolute symmetry, etc. Just as painting acts against a background of norms, so in music there are
expected patterns -- of melody, rhythm, harmonic progression, abstract form, etc., and a
composer's skill lies not in mechanically reproducing these, but in introducing unexpected
departures from them. As a general rule, anyone who wishes to investigate the significance and
value of a work of art must concentrate on the element of interest and surprise, rather than on the
automatic pattern. Such deviations from linguistic or other socially accepted norms have been
given the special name of foregrounding', which invokes the analogy of a figure seen against a
background. The artistic deviation 'sticks out' from its background, the automatic system, like a
figure in the foreground of a visual f1eld.
The application of this concept to poetry is obvious. The foregrounded figure is the linguistic
deviation, and the background is the language the system taken for granted in any talk of
deviation. Just as the eye picks out the figure as the important and meaningful element in its field
of vision, so the reader of poetry picks out the linguistic deviation in such a phrase as 'a grief ago'
as the most arresting and significant part of the message, and interprets it by measuring it against
the background of the expected pattern (see 2.4). It should be added, however, that the rules of the
English language as a unity are not the only standard of normality: as we saw in Chapter I, the
English of poetry has its own set of norms, so that 'routine licences' which are odd in the context
12

of English as a who1e are not foregrounded, but rather expected, when they occur in a poem. The
unique creative innovations of poetry, not the routine deviations, are what we must chiefly have in
mind in this discussion of foregrounding.
Deliberate linguistic foregrounding is not confined to poetry, but is found, for example, in joking
speech and in children's games. Literature is distinguished, as the Czech scholar Mukarovsky says,
by the 'consistency and systematic character of foregrounding', but even so, in some non-literary
writing, such as comic 'nonsense prose', foregrounding may be Just as pervasive and as violent (if
not more so) as it is in most poetry:
Henry was his father's son and it were time for him to go into his father's business of
Brummer Striving. It wert a farst dying trade which was fast dying.
Even in this short passage from John Lennon in his own Write, there are several instances of
orthographic, grammatical, and semantic deviation. If a longer passage were considered, it would
be seen that the linguistic foregrounding is far from being spasmodic or random -- it follows a
certain rationale of its own. It is difficult to analyze what is meant by foregrounding being'
systematic', but the notion is intuitively clear in the feeling we have that there is some method in a
poet's (and even in John Lennon's) ' madness'.
l.2 An Example
A convincing illustration of the power of foregrounding to suggest latent significance is furnished
by those modern poets (especially Pound and Eliot) who make use of the stylistic device of
transposing pieces of ordinary, non-poetic language into a poetic context. A famous example of
this kind of register-borrowing is the bar-parlour monologue in 'A Game of Chess' [The Waste
Land, III]:
When Lil's husband got demobbed, I said I didn't mince my words, I said to her myself...
Now Albert's coming back, make yourse1f a bit smart.
He'll want to know what you done with that money he gave you
To get yourself some teeth. He did, I was there...
The very fact that this passage occurs in a poem, incongruously rubbing shoulders with other,
more respectably literary types of English, causes us to pay it the compliment of unusual scrutiny.
Here it is foregrounded, whereas if it had been overheard in a pub or on a bus, it would not have
been. We find ourselves not paying heed to its meaning qua casual gossip, but rather asking what
is the point of its inclusion at this place in the poem? What is its relevance to its context ? What is
its artistic significance, in the light of what we have understood of the rest of the poem? This
method of composition recalls the painter's technique of collage; in particular, the gumming of
bits of newspaper, advertisements, etc., on to the surface of a painting. Because a piece of
newspaper, whatever its content, appears in the unwonted setting of a painting, we look at it with
more attention, and with a different kind of attention, from that of the careless eye we would cast
upon it in a customary situation. The same applies to Eliot's literary collage.

13

2 INTERPRETATION
Poetic foregrounding presupposes some motivation on the part of the writer and some explanation
on the part of the reader. A question-mark accompanies each foregrounded feature; consciously or
unconsciously, we ask: 'What is the point? Of course, there may be no point at all; but the
appreciative reader, by act of faith, assumes that there is one, or at least tends to give the poet the
benefit of the doubt. On the other hand, we must not forget the Mr. Wellers of this world, who
shrug their shoulders at each question mark, and take poetry to be a kind of outlandish nonsense.
The problem we now have to consider is the problem which stands astride the gap between
linguistic analysis and literary appreciation: When is a linguistic deviation (artistically)
significant?
2.1 The Subjectivity of Interpretation
To the foregoing question I wish to consider three answers.
ANSWER I: When it (i.e. the deviation) communicates something. According to this definition of
significance, practically all deviation is significant. Consider the following three cases:
[a] My aunt suffers from terrible authoritis.
[b] Like you plays?
[c] The Houwe [sic] of Commons.
The linguistic abnormalities in these examples are most likely to be taken as errors, as trivial
hindrances to communication. But unintentionally, they may convey quite a bit of information.
The first, if we take it to be an
example of malapropism (authoritis for arthritis), at least tells us something about the education,
character, etc., of its perpetrator. In the second examp1e, the ungrammaticality probably suggests
that its author is a foreigner with an imperfect command of English. The third example, occurring
in a printed text, informs us that the printer has made a mistake, that the author is a careless
proof-reader, etc. Such mistakes may, of course, be deliberately imitated for artistic or comic
effect, as in the case of Mrs. Malaprop herself:
An aspersion upon my parts of speech ! Was ever such a brute ! Sure, if I reprehend
anything in this world, it is the use of my oracular tongue and a nice derangement of
epitaphs. [Sheridan, The Rivals, III.iii]
However, it is clear that even the most trivial and unmotivated deviation may communicate
information of a kind.
ANSWER 2: When it communicates what was intended by its author. This definition of
'significant' narrows the first one to exclude solecisms, malapropisms, and other sorts of linguistic
blunder. It insists that a deviation is significant only when deliberate. But the one main difficulty
about this answer is that the intention of the author is in practice inaccessible. If he is dead, his
intention must remain for ever unknown, unless he happens to have recorded it; and even a living
poet is usua1ly shy of explaining 'what he meant' when he wrote a given poem. There is, moreover,
a widely held view that what a poem signifies lies within itself and cannot be added to by

14

extraneous commentary. In any case, must a poet's own explanation be treated as oracu1ar? An
interesting case of conflicting interpretation is reported in Tindall's A Reader's Guide to Dylan
Thomas. In Thomass A Grief Ago there occurs a puzzling compound country-handed:
The country-handed grave boxed into love.
Edith Sitwell discerned in the compound a' rural picture of a farmer growing flowers and corn',
whereas Thomas himself said that this was quite contrary to his intention, and that he had
envisaged the grave in the likeness of a boxer with fists as big as countries. Should we accept
Thomas's 'correction' as the last word on the subject? Or should we not rather accept Edith
Sitwell's interpretation as being valid and artistically significant in its own right?
ANSWER 3: When it is judged or felt by the reader to be significant. This answer, anticipated
above, is on the face of it the most unsatisfactory of all: it merely says that the significance of a
poem lies ultimately in the lid of the reader, just as beauty is said to lie in the eye of the beho1der.
Yet I think we are forced back on this definition by the failure of the other two to circumscribe
what people in practice take to be significant in a poem. We may go further, and say that not only
whether a deviation has a sensible interpretation, but what interpretation it is to be given, is a
subjective matter. Not that I am advocating the critical anarchy of every man's opinion being as
good as his neighbour's: there is such a thing as a consensus of interpretative judgment, in which
certain experts (critics) have a bigger voice than laymen, and in which the voice of the poet, if
heard, is probably the most authoritative of all.
This conclusion, however much of an anticlimax it may seem, is salutary if it teaches us the
difference between the objectivity (at least in spirit) of linguistic analysis, and the subjectivity (in
the last resort) of critical interpretation. It should also teach us that linguistics and literary criticism,
in so far as they both deal with poetic language, are complementary not competing activities.
Where the two meet is above all in the study of foregrounding.
2.2 The 'Warranty' for a Deviation
Assuming that a deviation can be given a sensible and constructive interpretation, let us now
examine more precisely how a particular interpretation is arrived at. In detail, this is a matter of
critical theory rather than stylistics, and I can do no more here than sketch, in a general way, the
processes involved.
A linguistic deviation is a disruption of the normal processes of communication: it leaves a gap, as
it were, in one's comprehension of the text. The gap can be filled, and the deviation rendered
significant, but only if by an effort of his imagination the reader perceives some deeper connection
which compensates for the superf1cial oddity. In the case of metaphor, this compensation is in the
form of analogy. Donne's line (from The Apparition)
Then thy sick taper will begin to wink
contains two violations of literal meaningfulness: the idea of a taper being sick', and the idea of a
taper being capable of winking. The warranty for these deviations lies in a figurative interpretation
of 'sick' and 'wink', whereby we appreciate analogies between someone who is ill and a candle
which is burning out, and between the flickering of a candle and the batting of an eyelid. The
search for a warranty can go further than this. We can ask how these comparisons contribute to the
15

total effectiveness of the poem; but for the moment we shall only investigate what can be called
the immediate warranty for a deviation.
Another kind of deviation is illustrated in the bizarre word-blends and neologisms of Joyce's
Finnegan's Wake, e.g. museyroom, wholeborrow, Gracehoper. In these cases the immediate
warranty can be divided into two parts. The first is the apprehension of a linguistic connection -actually a phonological resemblance -- between the invented word and one or more
well-established items of vocabulary: museum, wheelbarrow, grassbhopper. The second is the
attempt to match this linguistic connection with some connection outside language, perhaps some
referential connection between the invented words and the 'proper' words we map on to them.
Thus museyroom suggests, appropriately enough, that a museum is a room in which one muses,
just as in [a] of 2.1, authoritis might suggest a writing-bug which afflicts my aunt as cripplingly as
arthritis.
(from Geoffrey N. Leech. (1969/2001). A Linguistic Guide to English Poetry. PP. 56-61. Beijing:
Longman Group Limited & Foreign Languages Teaching and Research Press)

16

Text 4

How to analyze English Fictional Prose

-- A Method of Analysis and Some Examples
Geoffrey Leech and Mick Short

This chapter has the practical purpose of showing how the apparatus of linguistic description can
be used in analyzing the style of a prose text. We take for granted the principles that have been
argued in Chapters I and 2. We also take for granted a set of linguistic categories which will be
more or less common knowledge to those who have a basic familiarity with the workings of the
English language, whether in literary or non-literary contexts. One particular area in which
technical terms are likely to cause some problems is that of grammar: and here we shall follow the
terminology and general view of grammar presented in Quirk and Greenbaum's University
Grammar of English. Another area is that of foregrounding, where we draw on the terminology of
traditional poetics ('metaphor', 'metonymy', 'onomatopoeia', etc). Although many of these terms
are widely current in literary scholarship, we shall presuppose a linguistic account of these
phenomena, and for this purpose, it is convenient to refer the reader to Leech, A Linguistic Guide
to English Poetry. Following the list of categories in 3.1, explanations of selected points will be
added in 3.2.
Every analysis of style, in our terms, is an attempt to find the artistic principles underlying a
writer's choice of language. All writers, and for that matter all texts have their individual qualities.
Therefore the features which recommend themselves to the attention in one text will not
necessarily be important in another text by the same or a different author. There is no infallible
technique for selecting what is significant. We have to make ourselves newly aware, for each text,
of the artistic effect of the whole, and the way linguistic details fit into this whole.
Nevertheless it is useful to have a checklist of features which may or may not be significant in a
given text. For this reason, the following list of questions has been prepared. The answers to these
questions will give a range of data which may be examined in relation to the literary effect of each
passage. We stress that the list serves a heuristic purpose: it enables us to collect data on a fairly
systematic basis. It is not exhaustive, of course, but is rather a list of 'good bets': categories which
in our experience, are likely to yield stylistically relevant information. The stylistic values
associated with the linguistic data must be largely taken on trust at present; in subsequent chapters,
we shall endeavour to show how these values, too, can be studied systematically.
3.l A CHECKLIST OF LINGUISTIC AND STYLISTIC CATEGORIES
The categories are placed under four general headings; lexical categories, grammatical categories,
figures of speech, and cohesion and context. Semantic categories are not listed separately, since, as
suggested in 2.9, it is easier to arrive at these through other categories, we shall, for example, use
our lexical categories to find out how choice of words involves various types of meaning. Since
the purpose of the list is heuristic, there is no harm in 'mixing categories' in this way. It is also in

17

the nature of things that categories will overlap, so that the same feature may well be noted under
different headings.
A: Lexical categories
[For notes (i-xiv) on the categories see pp 80-2]
1 GENERAL. Is the vocabulary simple or complex? formal or colloquial? descriptive or
evaluative? general or specific? How far does the writer make use of the emotive and other
associations of words, as opposed to their referential meaning? Does the text contain idiomatic
phrases, and if so, with what kind of dialect or register are these idioms associated? Is there any
use of rare or specialized vocabulary? Are any particular morphological categories noteworthy (eg
compound words, words with particular suffixes)? To what semantic fields do words belong?
2 NOUNS. Are the nouns abstract or concrete? What kinds of abstract nouns occur (eg nouns
referring to events, perceptions, processes, moral qualities, social qualities)? What use is made of
proper names? collective nouns?
3 ADJECTIVES. Are the adjectives frequent? To what kinds of attribute do adjectives refer?
physical? psychological? visual? auditory? colour? referential? emotive? evaluative? etc. Are
adjectives restrictive or non-restrictive? gradable or non-gradable? attributive or predicative?
4 VERBS. Do the verbs carry an important part of the meaning? Are they stative (referring to
states) or dynamic (referring to actions, events, etc)? Do they 'refer' to movements, physical acts,
speech acts, psychological states or activities, perceptions, etc? Are they transitive, intransitive,
linking (intensive), etc? Are they factive or non-factive?
5 ADVERBS. Are adverbs frequent? What semantic functions do they perform (manner, place,
direction, time, degree, etc)? Is there any significant use of sentence adverbs (conjuncts such as so,
therefore, however; disjuncts such as certainly, obviously, frankly)?
B: Grammatical categories
l SENTENCE TYPES Does the author use only statements (declarative sentences), or does he
also use questions, commands, exclamations, or minor sentence types (such as sentences with no
verb)? lf these other types are used, what is their function?
2 SENTENCE COMPLEXlTY. Do sentences on the whole have a simple or a complex structure?
What is the average sentence length (in number of words)? What is the ratio of dependent to
independent clauses? Does complexity vary strikingly from one sentence to another? Is
complexity mainly due to (i) coordination, (ii) subordination, (iii) parataxis (juxtaposition of
clauses or other equivalent structures)? In what parts of a sentence does complexity tend to occur?
For instance, is there any notable occurrence of anticipatory structure (eg of complex subjects
preceding the verbs, of dependent clauses preceding the subject of a main clause)?
3 CLAUSE TYPES. What types of dependent clause are favoured: relative clauses, adverbial
clauses, different types of nominal clauses (that clauses, wh-clauses, etc)? Are reduced or
non-finite clauses commonly used, and if so, of what type are they (infinitive clauses, -ing clauses,
--ed clauses, verbless clauses)?

18

4 CLAUSE STRUCTURE. Is there anything significant about clause elements (eg frequency of
objects, complements, adverbials; of transitive or intransitive verb constructions)? Are there any
unusual orderings (initial adverbials, fronting of object or complement, etc)? Do special kinds of
clause construction occur (such as those with preparatory it or there)?
5 NOUN PHRASES. Are they relatively simple or complex? Where does the complexity lie (in
premodification by adjectives, nouns, etc, or in postmodification by prepositional phrases, relative
clauses, etc)? Note occurrence of listings (eg sequences of adjectives), coordination, or apposition.
6 VERB PHRASES. Are there any significant departures from the use of the simple past tense?
For example, notice occurrences and functions of the present tense; of the progressive aspect (eg
was lying), of the perfective aspect (eg has/had appeared, of modal auxiliaries (eg can, must,
would, etc).
7 OTHER PHRASE TYPES. Is there anything to be said about other phrase types: prepositional
phrases, adverb phrases, adjective phrases?
8 WORD CLASSES. Having already considered major or lexical word classes, we may here
consider minor word classes ('function words'): prepositions, conjunctions, pronouns, determiners,
auxiliaries, interjections. Are particular words of these types used for particular effect (eg the
definite or indefinite article; first person pronouns I, we, etc; demonstratives such as this and that;
negative words such as not, nothing, no)?
9 GENERAL. Note here whether any general types of grammatical construction are used to
special effect; eg comparative or superlative constructions; coordinative or listing constructions;
parenthetical constructions; appended or interpolated structures such as occur in casual speech. Do
lists and coordinations (eg lists of nouns) tend to occur with two, three or more than three
members?
C: Figures of Speech, etc
Here we consider the incidence of features which are foregrounded (1.4) by virtue of departing in
some way from general norms of communication by means of the language code; for example,
exploitation of regularities of formal patterning, or of deviations from the linguistic code. For
identifying such features, the traditional figures of speech (schemes and tropes) are often useful
categories.
l GRAMMATlCAL AND LEXICAL SCHEMES. Are there any cases of formal and structural
repetition (anaphora, parallelism, etc) or of mirror-image patterns (chiasmus)? Is the rhetorical
effect of these one of antithesis, reinforcement, climax, anticlimax, etc?
2 PHONOLOGlCAL SCHEMES. Are there any phonological patterns of rhyme, alliteration,
assonance, etc? Are there any salient rhythmical patterns? Do vowel and consonant sounds pattern
or cluster in particular ways? How do these phonological features interact with meaning?
3 TROPES. Are there any obvious violations of , or departures from the linguistic code? For
example, are there any neologisms (such as Americanly)? deviant lexical collocations (such as
portentous infants)? semantic, syntactic, phonological, or graphological deviations? Such
deviations will often be the clue to special interpretations associated with traditional figures of

19

speech such as metaphor, metonymy, synecdoche, paradox, irony. If such tropes occur, what kind
of special interpretation is involved (eg metaphor can be classified as personifying, animizing,
concretizing, synaesthetic, etc)? Because of its close connection with metaphor, simile may also be
considered here. Does the text contain any similes, or similar constructions (eg 'as
ifconstructions)? What dissimilar semantic fields are related through simile?
D: Context and cohesion
Finally, we take a preliminary look at features which will be more fully dealt with in Chapters 7 to
10. Under COHESION ways in which one part of a text is linked to another are considered: for
example, the ways in which sentences are connected. This is the internal organization of the text.
Under CONTEXT (see the discussion of discourse situation, 8. I) we consider the external
relations of a text or a part of a text, seeing it as a discourse presupposing a social relation between
its participants (author and reader; character and character, etc), and a sharing by participants of
knowledge and assumptions.
l COHESION. Does the text contain logical or other links between sentences (eg coordinating
conjunctions, or linking adverbials)? Or does it tend to rely on implicit connections of meaning?
What sort of use is made of cross-reference by pronouns (she, it, they, etc)? by substitute forms
(do, so, etc), or ellipsis? Alternatively, is any use made of elegant variation -- the avoidance of
repetition by the substitution of a descriptive phrase (as, for example, 'the old lawyer' or 'her uncle'
may substitute for the repetition of an earlier 'Mr Jones')? Are meaning connections reinforced by
repetition of words and phrases, or by repeatedly using words from the same semantic field?
2 CONTEXT. Does the writer address the reader directly, or through the words or thoughts of
some fictional character? What linguistic clues (eg first--person pronouns I, me, my, mine) are
there of the addresser--addressee relationship? What attitude does the author imply towards his
subject? If a character's words or thoughts are represented, is this done by direct quotation (direct
speech), or by some other method (eg indirect speech, free indirect speech)? Are there significant
changes of style according to who is supposedly speaking or thinking the words on the page?
(from Geoffrey N. Leech and Michael H. Short. (1981/2001). Style in Fiction: A Linguistic
Introduction to English Fictional Prose. PP. 74-80. Beijing: Longman Group Limited & Foreign
Languages Teaching and Research Press)

20

Text 5

Foreword to Cognitive Stylistics

Elena Semino and Jonathan Culpeper

This collection aims to represent the state of the art in cognitive stylistics a rapidly expanding
field at the interface between linguistics, literary studies and cognitive science. Cognitive stylistics
combines the kind of explicit, rigorous and detailed linguistic analysis of literary texts that is
typical of the stylistics tradition with a systematic and theoretically informed consideration of the
cognitive structures and processes that underlie the production and reception of language.
Cognitive stylistics, as we have just defined it, is both old and new. It is old in the sense that, in
focusing on the relationship between linguistic choices and effects, stylistics has always been
concerned with both texts and readers' interpretations of texts. It is important to remember, for
example, that Foregrounding theory (Mukarovski l970), which played a major role in the
development of modern Anglo-American stylistics, is concerned with the cognitive effects of
particular linguistic choices and patterns (and this in spite of the fact that it stemmed from a school
known as Formalism). It is therefore no coincidence that van Peer's seminal book on the empirical
investigation of Foregrounding theory was entitled Stylistics and Psychology: Investigations of'
Foregrounding (van Peer l986). Similarly work in stylistics has been influenced by theoretical and
methodological advances in Reader Response Criticism (e.g. Fish l973) and the Empirical Study
of Literature (e.g. Short and van Peer l988).
Traditional stylistic analysis, however, tends to make use of linguistic theories or frameworks in
order to explain or predict interpretation. What is new about cognitive stylistics is the way in
which linguistic analysis is systematically based on theories that relate linguistic choices to
cognitive structures and processes. This provides more systematic and explicit accounts of the
relationship between texts on the one hand and responses and interpretations on the other.
What is shared by all of the twelve chapters included in the volume, therefore, is (a) a concern for
specific texts or textual phenomena, (b) the adoption of analytical approaches that explicitly relate
linguistic choices to cognitive phenomena, and (c) the claim that a satisfactory account of the text
or phenomenon in question can only be arrived at by means of a cognitive stylistic approach. More
specifically, a recurrent goal in most of the chapters is that of explaining how interpretations are
arrived at, rather than proposing new interpretations of texts. In several chapters the analysis is
partly aimed to account systematically for how the same (stretch of) text can give rise to different
interpretations (notably Freeman, Hamilton, Popova, Stockwell, Tsur).
While the commonalities between the twelve chapters justify their inclusion under the umbrella of
what we call cognitive stylistics, our choice of contributors also aims to represent the variety of
work that can be subsumed within this field. A relatively minor aspect of variation lies in how
contributors prefer to label the enterprise they are involved in. Some use "cognitive stylistics,
others "cognitive poetics", yet others explicitly present the two as synonymous (see also Wales
2001: 64). We also regard the two labels as largely overlapping, but have adopted "cognitive
stylistics" in the title of this volume in order to emphasize a concern for close attention to the
language of texts. The term "cognitive stylistics" was also used as the title of a section of Weber's
21

The Stylistics Reader from Roman Jakobson to the Present (Weber l996). This section included a
chapter by Donald Freeman, who is the author of the "Afterword" of the present collection.
A more significant aspect of variation within this book is to do with the particular cognitive
approach adopted in each chapter. Not surprisingly most chapters are influenced by cognitive
linguistics as associated with the work of Langacker, Lakoff and others. Indeed, it could be argued
that the rise of what we call cognitive stylistics at this particular point in history is partly due to the
increasing influence of cognitive linguistics and, more specifically, of cognitive metaphor theory
in the Lakoff tradition. However, those contributors who draw from this particular paradigm differ
in how exactly they position themselves in relation to it. Freeman, Hamilton, Popova and Steen
see cognitive stylistics as part of the cognitive linguistic paradigm. Others treat cognitive
linguistics as one of the cognitive theories or paradigms that can feed into cognitive stylistics -- a
position which we also share. The chapters by Semino, Stockwell and van Peer & Graf draw both
from cognitive linguistics and from other theories from psychology and cognitive science
generally. Emmott, and also Steen, assess the potential of cognitive linguistics in accounting for
the phenomena they discuss, and point out some of its weaknesses. Emmott emphasizes the lack of
proper consideration for discoursal and narrato1ogical phenomena, while Steen claims that
cognitive linguistics needs to take better account of the psycho1Ogy of text processing. The
remaining chapters are largely independent of the cognitive linguistics paradigm. Shen develops
and tests out his own cognitive theory of figurative language use in poetry but spells out its
relationship (and compatibility) with cognitive metaphor theory. Attardo and Culpeper propose
their own frameworks for the study of humour and characterization respectively, drawing from the
tradition of text processing research. Finally, Tsur presents his theory of "cognitive poetics" as "a
far cry from" or, in some respects, "even diametrically opposed" to what goes nowadays under
the labe1 'cognitive linguistics."' The ordering of chapters in this volume roughly follows this
account of their respective theoretica1 positions.
Finally' the twelve chapters included in this collection cover a wider range of literary texts, literary
periods, and phenomena, including poetry, fictional and non-fictional narratives, and plays. Some
of the chapters provide innovative approaches to phenomena that have a long tradition in literary
and linguistic studies (e.g. humour, characterization, figurative language, metre), others for on
phenomena that have not yet received adequate attention (e.g. split-selves phenomena, mind style,
spatial language). There is also variation in the particular analytical methodology used, and in the
way in which the linguistic analysis is related to possible interpretations of the relevant (stretches
of) texts. While some contributors hypothesize about likely or possible responses on the basis of
their own personal readings, several refer to the reactions of literary critics (notably Popova and,
to a lesser extent Freeman, Hamilton, Semino, Stockwell), and some make use of informant
testing in order to test out specific hypotheses (Shen, Steen, Tsur).
Outline of chapters
In the first three chapters of the book, some of the central concepts and insights from cognitive
linguistics are applied to the analysis of specific texts. In Chapter 1, Craig Hamilton shows how
"conceptual integration" or "blending" theory can explain the use and interpretation of metaphor,
analogy and allegory in Christine de Pizan's l5th century text The Book of the City of Ladies (Le
Livre de la cit des dames). His analysis accounts systematically for readers' intuitive
understanding of de Pizan's book, and for both similarities and differences in the way in which it
22

has been interpreted across different historical periods and cultural groups. Hamilton claims that
"many scholars today see that literary criticism needs to make the cognitive turn lest it become an
entirely bogus and meaningless enterprise", and exposes what he sees as the lack of proper
scientific foundation for fashionable notions such as "the female reader" and "gendered memory'.
In Chapter 2 Margaret Freeman argues that the cognitive linguistic emphasis on the embodied
nature of mind and language leads to a reevaluation of the physical shape of texts, and poems in
particular. Freeman analyses the original hand-written versions of two poems by Emily Dickinson,
which were significantly altered in existing printed versions and which have generally been
regarded as obscure by critics. Freeman shows how an adequate reading of both texts can only be
arrived at by taking into account the cognitive import of all their original visual characteristics, as
well as the particular knowledge domains and conceptual metaphors that make up Dickinson's
conceptual universe. Freeman emphasizes that "cognitive poetics has explanatory power," in that it
is able to explain how meanings are created in the production and reception of texts. In Chapter 3
Yanna Popova discusses Henry James's The Figure in the Carpet-- a story which has often been
claimed to be highly ambiguous, and which has given rise to a range of different interpretations on
the part of literary critics. Popova's aim is not to resolve the ambiguity but to use insights from
cognitive linguistics to explain why the ambiguity cannot be resolved, i.e. why the text is
inherently ambiguous. Her analysis traces the cause of the ambiguity to the presence in the texts of
two main, and partly incompatible, metaphorical conceptualization of the "secret" which is at the
centre of the plot. Popova also considers around thirty published discussions of James's tale, and
show how her analysis can account for the two main interpretative lines in the story's criticism.
Popova claims that a cognitive linguistic approach can distinguish between a basic, "archetypal"
level of interpretation and other less basic levels, and also account for interpretative variability on
the one hand and, on the other, the fact that not all interpretations are equally acceptable. This, she
argues, "constitutes a much needed middle ground between pluralism and objectivism in theories
of interpretation." Chapters 4 to 8 draw eclectically from cognitive linguistics and other cognitive
theories and approaches, While Chapters 7 and 8 also point out some of the limitations of the
cognitive linguistics paradigm. In Chapter 4 Peter Stockwell analyses four sonnets by Milton in
terms of their "texture" a combination of formal and psychological features that contribute to
"how we feel our way through the reading of a text". Stockwell takes a cognitive approach to
linguistic phenomena such as deixis, syntax and negation, and draw on a range of cognitive
notions in order to explain their possible effects, including conceptual metaphors, the concepts of
"attraction" and distraction", and the contrast between "figure" and "ground". In his analyses,
Stockwell accounts for different possible on-line readings of the four texts and show how their
different types of texture can only be adequately captured in terms of the relationship between
textual features and cognitive processes. In Chapter 5 Elena Semino argues that the phenomenon
known as "mind style" can only be properly accounted for by relating linguistic features to
cognitive structures and phenomena via relevant cognitive theories. Semino differentiates the
notion of "mind style" from that of "ideological point of view", and claims that the former "is to
do with how language reflects the particular conceptual structure and cognitive habits that
characterize an individual's world view." She draws from schema theory cognitive metaphor
theory and blending theory in order to show how two particular mind styles are linguistically
created, namely those of a minor character in Louis de Bernieres's Captain Corelli's Mandolin,
and of Frederick Clegg in John Fowless The Collector. In Chapter 6, Willie van Peer and Eva

23

Graf also start from the cognitive stylistic assumption that linguistic patterns in texts reflect
cognitive processes, and investigate this assumption by analyzing the linguistic realizations of
spatial concepts in the language of children and adults in Stephen King's novel IT. The analysis
shows that the use of spatial language in the speech of the main characters as adults is more
complex, both linguistically and conceptually, than that of the same characters as children. This,
the authors show is consistent with what is known about real-life cognitive development in the
conceptualization of space and of other areas of experience which are related to it via conceptual
metaphors. The discussion of the textual samples draw from the cognitive linguistic account of
spatial metaphors, but also from the psychological literature on the understanding of space and
cognitive development. On the basis of their analysis, Van Peer & Graf evaluate King's
achievement in realistically creating a contrast between children's and adults' language and
cognition.
In Chapter 7 Catherine Emmott analyses a small corpus of "split selves" narratives, including both
fictional and non-fictional examples (e.g. Doris Lessing's Children of Violence novels and the
autobiographies of two stroke victims). Emmott highlights the wide range of split selves
phenomena contained in her data, and proposes a preliminary typology. She shows how some of
these phenomena have been recognized and accounted for by cognitive linguists (using notions
such as the CONTAINER metaphor, etc.), while others require the contribution of narratological
and discoursal approaches. These can better account for plots, narrative voices and the way in
which readers keep track of characters and characters' selves while reading. In Chapter 8 Gerard
Steen argues that the notion of genre as a type of mental representation needs to play a central role
in a cognitive account of text production and interpretation, as well as of the linguistic make-up of
texts themselves. He focuses specifically on metaphorical language and its relationship with genre,
and analyzes in detail the use of metaphor in the lyrics of Bob Dylans song "Hurricane". On the
basis of his analysis, Steen proposes eight variables that could affect metaphor recognition on the
part of readers, and discusses the results of an informant-based study aimed at investigating the
role of these variables. The results show that the some of the variables that were shown to have
significant effects can be related to the structure of the teat as a song lyric. Steen draws from the
cognitive linguistic approach to metaphor, but, like Emmott, he argues that cognitive linguistics
needs to be complemented by insights from other disciplines, notably work in the psychology of
reading and text processing. Steen also proposes his view of "empirical research, including both
text analysis and informant testing, and highlights the need for cognitive stylisticians to be aware
of relevant methodologies in the empirical sciences.
Chapter 9, l0 and 11 propose independent cognitive theories aimed at explaining particular
phenomena. In Chapter 9 Yeshayahu Shen introduces and demonstrates his Cognitive Constraints
Theory (CCT) in order to provide a cognitive account of specific structural patterns in the use of
figurative language in poetry. The CCT states that the structural regularities shown by figures of
speech in poetry reflect a "compromise" between the aesthetic goals of novelty and originality on
the one hand and the communicative goal of comprehensibility on the other. Shen presents the
results of a linguistic analysis of zeugma, synaesthesia and oxymoron in a range of poetic corpora,
spanning different languages, historical periods and literary movements. In each case he identifies
two main options in the realization of each figure of speech, and show how one option is
significantly more frequent than the other in his data. He then reports the results of informant tests

24

which show how the most frequent structural option is also the one that is cognitively most "basic",
i.e. easier to understand, recall and so on. These findings can be explained by Shen's CCT and are
compatible, as Shen shows, with relevant aspects of cognitive metaphor theory. Shen emphasizes
how his work differs from other cognitive approaches to literary 1anguage in that it demonstrates
the existence of cognitive constraints on creativity in verbal art, rather than arguing that the latter
goes against or disrupts "normal" cognitive processes. In Chapter l0 Salvatore Attardo focuses on
humour -- a phenomenon which is relevant to a wide range of discourse contexts and text types.
The notion of humour captures the effects of some communicative stimulus on interpreters, and
therefore requires an approach that includes both the stimulus (language in our case) and the
cognitive structures and strategies used by the interpreter. Attardo introduces his General Theory
of Verbal Humour (GTVH), a framework whose components include both linguistic choices and
patterns on the one hand, and cognitive structures and processes on the other. Attardo shows how
the GTVH is able to account for a range of humorous effects and to distinguish between different
types of humorous texts. A detailed analysis of Oscar Wi1de's short story "Lord Arthur Savi1e's
Crime" shows how the GTVH can identify the particular type of humour that characterizes a text,
highlight different patterns of humour in texts, and help address questions of interpretation and
aesthetic value. In Chapter 11 Jonathan Culpeper deals with another phenomenon that is central to
a wide variety of text types, namely characterization: the incremental construction of mental
representations of characters in text processing. Culpeper describes and demonstrates a model of
characterization that explains how the words in a text create a particular impression of a character
in the reader's mind and how that impression may change in the course of processing a text.
Culpeper locates his work in the realms of text comprehension, particularly as represented by van
Dik and Kintsch's (l983) influential work. In addition, he integrates into his model aspects from
social cognition, specifically work on social schemata (or cognitive stereotypes). He demonstrates
how his model can explain a wide variety of characters in a variety of text-types, and lead to
testable hypotheses about reader's interpretations of texts. Culpeper argues that his approach is a
corrective to those critical lines that have treated characters as either purely linguistic or purely
cognitive phenomena, and a response to those structuralist critics who have acknowledged that
both language and cognition must be taken into account in the study of characterization.
Finally, in Chapter 12 Reuven Tsur presents some central aspects of his own theory of cognitive
poetics, which, he argues, "offers cognitive theories that systematically account for the
relationship between the structure of literary texts and their perceived effects" and also
"discriminates which reported effects may be legitimately related to the structures in question, and
which may not." Tsur uses his approach to explain how poems can convey emotional qualities, or,
in other words, how poets create verba1 equivalents of the structure of emotions. He focuses
particularly on poems that convey what he calls "altered states of consciousness," and explains
how different responses to the same stretches of text may arise as a consequence of different
reading styles and different degrees of tolerance for ambiguity and disorientation on the part of
readers. Tsur ends by providing a cognitive account of the perception of poetic rhythm, focusing
particulal1y on how metrical regularity is perceived by readers in case where a stretch of verse
contains very few metrically regu1ar lines. Throughout, Tsur refers informally to readers's
reactions to the examples he discusses, and, in his discussion of poetic rhythm, he tests his
hypotheses by analyzing different oral performances of a particular poem. As we mentioned earlier,
Tsur contrasts his own approach to cognitive poetics with that proposed within the cognitive
25

linguistic paradigm. His main contention is that, while cognitive linguists emphasize
conventionality, his approach emphasizes creativity and focuses on what is individual and unique
to each poetic expression. In this respect, he argues, his theory provides a systematic cognitive
account of phenomena that were dealt with by New Criticism, Structuralism and Formalism
"sometimes quite brilliantly in a pre-theoretical manner."
Finally, in the "Afterword" Donald Freeman provides an overview of all twelve chapters, and
reflects on the potential contribution of cognitive stylistics to the future of literary studies.
(from Elena Semino and Jonathan Culpeper. (2002). (eds) Cognitive Stylistics Language and
Cognition in Text Analysis. PP. IV-XVI. Amsterdam/Philadelphia: John Benjamins Publishing
Company)

26