Language and Literature: Stylistics

Peter Stockwell 742 31 Language and Literature: Stylistics PETER STOCKWELL 1 Introduction It might seem obvious to the non-specialist that literature, the most culturally valued and aesthetically prestigious form of language practice, is best studied using the resources developed in the field of linguistics. However, this truism has not always been obvious to a wide range of disciplines, all of which claim a different stake in the study of the literary. Much of this contentiousness has arisen out of the historical baggage accumulated by institutionalized disciplines, out of territorial self-interest, and (it must be said) out of intellectual laziness, as well as the legitimate arguments around the validity and scope of linguistics. Stylistics is the discipline that has bridged these areas, and stylisticians have found themselves engaged in arguments not only with literary critics, cultural theorists, philosophers, poets, novelists and dramatists, but also with practitioners of linguistics. On the one hand it is argued that the artistic endeavour of literature cannot be amenable to the sort of rigorous analytical procedures offered by linguistic analysis; on the other hand it is argued that descriptive linguistics cannot be applied to artificial texts and readerly interpretations. For one group, stylistics simply and reductively dissects its object; for the other, the object simply cannot be described in a scientifically replicable and transparent manner. The multivalent position of stylistics has its roots in the histories of language study and literary criticism, and the institutional make-up of modern universities and department divisions which fossilize particular disciplinary boundaries and configurations. Stylistics has therefore come to be regarded as an essentially interdisciplinary field, drawing on the different sub-disciplines within linguistics to varying degrees, as well as on fields recognizable to literary critics, such as philosophy, cultural theory, sociology, history and psychology. However, by the end of this article, I would like to argue that stylistics is in fact a single coherent discipline: in fact, is naturally the central discipline of literary study, against which all other current approaches are partial or THOC31 742 1/10/06, 1:58 PM Language and Literature: Stylistics 743 interdisciplinary. In order to arrive at that position, we must consider the history of stylistics, the status of stylistic analysis, some examples of stylistic practice, and a review of the latest paradigms and principles in stylistics research. 2 A Brief History of Stylistics Broadly viewed as the analysis of linguistic form and its social effects, stylistics can be seen as a direct descendant of rhetoric, which constituted a major part of the training of educated men for most of the past two and a half millennia. Specifically, stylistics overlaps considerably with ‘elocutio,’ the selection of style for an appropriate effect. (The other four divisions of rhetorical skill were: invention, the organization of ideas, memory, and delivery). It is important to note the dual aspect in the discipline: rhetoric was concerned not only with linguistic form but also inextricably with the notion of the appropriacy of the form in context. The context was typically and primarily for spoken discourse, though rhetorical discussion was also applied to written texts. In the course of the twentieth century, stylistics developed with an almost exclusive focus on written literature, while at the same time the link between formalism and readerly effects became weakened. According to Fowler (1981), there were three direct influences which produced stylistics: Anglo-American literary criticism; the emerging field of linguistics; and European, especially French, structuralism. Early twentieth century literary criticism tended either to be historical and based in author-intention, or more focused on the texture of the language of literary works. The latter, though also encompassing textual editing and manuscript scholarship, mainly focused on the ‘practical criticism’ of short poems or extracts from longer prose texts. Such ‘close reading’ was largely informed by a few descriptive terms from the traditional school-taught grammar of parts of speech. This British practical criticism developed in the US into the ‘New Criticism.’ Where the former placed readerly interpretation first with the close reading to support it, the New Critics focused on ‘the words themselves.’ Famous essays by Wimsatt and Beardsley (1954a, 1954b) and others argued for the exclusion of any considerations of authorial intention or the historical conditions of contemporary production of literary works, and also against any psychologizing of the literary reading experience. Despite the rather uncompromising stance taken by New Criticism, the belief that a literary work was sufficient unto itself did not amount to a purely descriptive account of literary texts. Interpretative decisions and resolutions simply remained implicit in terms of the social conditions and ideologies that informed them, while being dressed up in an apparent descriptive objectivity. A more rigorous descriptive account was being developed in the field of linguistics. As Fowler (1981) points out, Bloomfieldian structural linguistics evolving between the 1920s and 1950s offered a precise terminology and THOC31 743 1/10/06, 1:58 PM 744 Peter Stockwell framework for detailed analyses of metrical structure in poetry. Chomskyan transformational-generative grammar from 1957 onwards provided a means of exploring poetic syntactic structure with far more sensitivity to detail than had ever been possible in literary criticism. And Hallidayan functionalism added a socio-cultural dimension that began to explain stylistic choices in literary texts. The third area which influenced stylistics was European structuralism, arising out of Saussurean semiology and Russian Formalism through the work of Jakobson, Barthes, Todorov, Levi-Strauss, and Culler, among others. Branded ‘formalists’ by their detractors, many of the main concerns of modern poetics were in fact developed by the Moscow Linguistic Circle, the St Petersburg group Opayaz, and later the Prague School linguists. These concerns included studies of metaphor, the foregrounding and dominance of theme, trope and other linguistic variables, narrative morphology, the effects of literary defamiliarization, and the use of theme and rheme to delineate perspective in sentences. The Formalists called themselves ‘literary linguists,’ in recognition of their belief that linguistics was the necessary ground for literary study. Stylistics began as a distinct approach to literary texts in the hands of Spitzer (1948), Wellek and Warren (1949), and Ullmann (1964), for example, but it really emerged from the 1960s onwards as the different influences mentioned above came to be integrated into a set of conventions for analysis. From Formalism and practical criticism came the focus of interest on literature and the literary, and from linguistics came the rigour of descriptive analysis and the scientific concern for transparency in that description. Though stylistic analysis could be practiced on any sort of text, much discussion involved the specification of ‘literariness’ and the search to define a ‘literary language’ – this preoccupation dominated to such an extent that stylistics has come to be identified very strongly with the discussion of literature, with non-literary investigations delineating themselves separately as ‘critical linguistics’ or ‘critical discourse analysis’ or ‘text linguistics,’ and so on. Of course, the notion of literariness makes no sense within a formalist or structuralist paradigm, since a large part of what is literary depends on the social and ideological conditions of production and interpretation. Nevertheless, stylistic analyses flourished in the 1970s, especially explorations of the metrics and grammar of poetry, and explanations of deviant or striking forms of expression in prose. Concerns with literariness, the investigation of artificial rather than natural language, and the spectre of capricious interpretation all served to make theoretical and applied linguists in other areas of linguistic study rather suspicious of stylistics. At a time when the other branches of linguistics were claiming prestige and institutional funding as social sciences, those who were interested in literary analysis tended to be regarded as operating at the ‘soft’ end of the discipline. Equally and contrarily, literary critics and philosophers tended to regard the practices of stylisticians as being mechanistic and reductive. Since stylisticians often worked in literature departments, the most heated debates occurred with literary critics: traditional liberal humanist critics attacked a THOC31 744 1/10/06, 1:58 PM Language and Literature: Stylistics 745 perceived irreverence for literary genius and its ineffable product; critics excited by the rise of literary theory as a discipline attacked stylistics for claiming to be merely a method without an ideological or theoretical underpinning. Notorious examples of the antagonism include the debate between the stylistician Roger Fowler and the literary critic F. W. Bateson (see Fowler, 1971 for an account), centring on the question of rigorous descriptiveness against literary sensibility; or the attack by Stanley Fish and defence by Michael Toolan (see Fish 1980; Toolan 1990), circling around the status of interpretation in literary reading. Although vigorous defences of stylistics continued to be raised in the 1970s, the field largely sidestepped the theoretical quagmire by taking an explicitly practical approach in the form of ‘pedagogical stylistics.’ This was a natural consequence of teaching (English) language using literary texts: foreign language learners took most readily to a linguistic approach to literature without importing any undue concern for theoretical niceties nor any misplaced reverence for the literary artefact. Teaching language through literature mirrored stylistics very clearly: texts tended to be those of contemporary literature; stylistically deviant texts were popular because they were fun and made it easy for the teacher to illustrate a specific point of usage; grammar and lexical choice were discussed as a motivating means of accessing the literature, rather than studied rather dryly for their own sake. Stylistics thus took itself out of literature departments and found adherents in education and modern language study around the world, enthusiastically supported by the international cultural promotion agency of the UK government, the British Council. (See Widdowson 1975, 1992; Brumfit 1983; and McCarthy and Carter 1994.) At the same time, advances in pragmatics, sociolinguistics and discourse analysis in the 1970s allowed stylistics to move beyond the analysis of short texts and sentence-level phenomena. Studies involving speech act theory, norms of spoken interaction, politeness, appropriacy of register choice, dialectal variation, cohesion and coherence, deictic projection, turn-taking and floorholding all allowed stylistics the opportunity of exploring text-level features and the interpersonal dimension of literature, especially in prose fiction and dramatic texts. New labels for a host of sub-disciplines of stylistics blossomed: literary pragmatics, discourse stylistics, literary semantics,’ ‘stylometrics,’ ‘critical linguistics,’ ‘schema poetics,’ and so on. Stylistics came to identify itself as virtuously interdisciplinary, though it should perhaps properly be seen in this period as ‘inter-sub-disciplinary.’ By the early 1980s, stylistics had established itself as a coherent set of practices largely based in Europe, mainly in Britain and Ireland, with strong centers in the Germanic and Scandinavian countries, representation in Spain as a major EFL destination for British teachers, with a separate tradition of stylistique operating in France, Italy, Greece and Turkey. Stylistics also developed where teaching links to Britain were strongest: in Australasia, India, Japan, and parts of Africa in the Commonwealth. The term ‘stylistics’ was nowhere near as widely used in North America, where generative grammar maintained its paradigmatic hold on linguistics, and post-structuralist theory enthralled those THOC31 745 1/10/06, 1:58 PM 746 Peter Stockwell literature departments that aspired to more than character-study and a simple historicism. 3 The Status of Stylistic Analysis One reason for the historical debates around stylistics has been the difficulty of defining ‘style.’ Even in its most simple sense of variation in language use, many questions instantly arise: variation from what? varied by whom? for what purpose? in what context of use? The different sub-disciplines that have been drawn on in stylistics have also brought along different senses of the term. Variationist sociolinguists treat style as a social variable correlated with gender, or class, for example, and have developed a cline of formality on this dimension. Anthropologists and ethnomethodologists have identified style with the contextual ‘domain’ in which the language variety is used, so that style has developed a wider sense close to that of ‘register.’ Style as an interpersonal feature involves psychological and socially motivated choices, so style can be seen as the characteristic pattern of choices associated with a writer’s or projected character’s ‘mind-style,’ or the pattern associated with particular periods, genres or literary movements. Most broadly, since every dimension of linguistic expression represents a choice – whether idiosyncratic or socially determined – the limits of ‘style’ can be seen to be the limits of language itself, which is not very helpful. One central tenet in modern stylistics has been to reject the artificial analytical distinction between form and content. Contrary to the practice of traditional rhetoric, style cannot be merely an ornamentation of the sense of an utterance, when it is motivated by personal and socio-cultural factors at every level and is correspondingly evaluated along these ideological dimensions by readers and audiences. Style is not merely free variation. Even utterances which are produced randomly (as can be seen in surrealist and nonsense works) are treated conventionally against the language system in operation. Moreover, there can be no synonymy in utterances, since the connotations even of close variations are always potentially significant. Taking this argument to its logical end, even the same sentence uttered twice is ‘stylistically’ non-synonymous since the context of the second occasion of utterance is different from that of the first. Clearly, the sense of ‘stylistic’ being used here has moved on a great deal from the earlier formalist sense of ‘the words themselves.’ The sorts of things stylisticians have been doing over the last twenty to thirty years have added more and more dimensions to the strictly ‘linguistic’ level, encompassing more of what language is while not losing sight of the necessity to ground descriptions in tangible evidence. Socio-cultural and psychological factors have become part of stylistic considerations. Since the early 1980s, stylistics has continued in an expansive phase. Criticized for constantly focusing on deviant or odd texts, stylisticians shifted to THOC31 746 1/10/06, 1:58 PM Language and Literature: Stylistics 747 the analysis of less stylistically striking writing, and presented variation in terms of norms and patterns that were internally marked in the literary work. The search for a linguistic definition of literariness was largely abandoned, with the literary being located in contexts of production and interpretation. The emphasis turned to examining the continuities between literary creativity and everyday creativity, and to how literary reading is continuous with the reception of language in general. Sociolinguistic findings informed literary analysis. Cognitive psychological aspects fed into stylistic exploration. Developments in pragmatics and discourse analysis continued to offer new tools and areas of investigation for stylistics. Insights into language use provided by corpus linguistics were drawn on, and computational techniques applied to literary works. Through the 1990s, stylistics in its most broad sense became one of the most dynamic and interdisciplinary fields within applied linguistics. In response to its invigorated position within literary studies, stylistic practice has recently attracted a new series of methodological attacks, as well as debates between stylisticians themselves around theoretical issues and ideologies. However, the key arguments and issues being discussed can still be seen as rehearsals of concerns that have been of interest throughout the history of poetics. For example, there have been several variations on the theme of the position of stylistics as a science or as part of a more artistic endeavour. Most stylistics adheres to the scientific practices of presenting rigorous and systematic method and being explicit about its assumptions. Studies mainly conform to a Popperian approach to scientific method: they are transparent, explicit in their hypotheses and expectations, and are therefore falsifiable in the sense that other readers can compare their own readings and see how they differ from the stylistician. Only the principle of the replicability of the study is problematic in stylistics, since the reading experience is unrepeatable. For integrationalists (such as Harris 1980, 1981, 2000; and Toolan 1996), this is a serious problem: in rejecting the Jakobsonian ‘code’ model of language as involving what they scornfully term ‘telementation,’ in effect they remove any possibility of stable or comparable analyses. Mere formalism is not an analysis of language as communication, they argue, but then the move of stylistics towards encompassing more context ultimately renders the products of analysis merely as idiosyncratic readings, little different from the intuitive expressions of traditional literary criticism. The key issue here is the question of interpretation, and the importance of noticing a difference between the textual object, reading, and interpretation. As I have argued elsewhere in response to the integrationalist critique (see Stockwell 2002b), stylistics can be regarded theoretically as a form of hermeneutics. Texts exist as autonomous objects, but the ‘literary work’ is an actualization of that object produced only by an observing consciousness (in the terms used by Ingarden 1973a, 1973b). The object of stylistic analysis (the literary work as opposed to the material literary text) comes into existence only when read. Since readers come with existing memories, beliefs and both personal and social objectives, the context of the literary work is already THOC31 747 1/10/06, 1:58 PM 748 Peter Stockwell conditioned by interpretation, even before reading begins (see Gadamer 1989). This means that reading is the process of becoming consciously aware of the effects of the text in the process of actualization: reading is inherently an analytical process, in this sense. Stylistics is simply the formal and systematic means of recording the same process and making it available for comparison. As Toolan (1990: 42–6) points out, stylistics can be used for a variety of purposes, including the teaching of language and of literature. It can also be used as a means of demystifying literary responses, understanding how varied readings are produced from the same text; and it can be used to assist in seeing features that might not otherwise have been noticed. It can shed light on the crafted texture of the literary text, as well as offering a productive form of assistance in completing interpretations, making them more complex and richer. Stylistics can thus be used both as a descriptive tool and as a catalyst for interpretation. These two possible functions of stylistics have been debated as if they were mutually exclusive: is stylistics a type of descriptive linguistics or is it a type of critical theory? The sense of exclusivity arises only if it is assumed that description is non-ideological. There are some stylisticians who argue that stylistics is simply a tool which can then be used in the service of a range of critical and interpretative positions. For example, it is an objective fact that a certain poem has a certain set of noun phrases from a particular semantic domain. Or it is a fact that the viewpoint in a certain novel is consistently a first person focalization. However, I would argue against this position, firstly on the theoretical dimension set out above that interpretation at least partly precedes analysis, and secondly on the practical dimension: since stylistics as a tool can only be manifest by being used, the fact that it is a descriptive tool in an ideal state is true but irrelevant in practice. As soon as stylistic analysis is undertaken, it partakes of ideological motivations, from the nature of the reading to the selection of the particular work and particular linguistic model for analysis. Examining noun phrases in the poem, rather than verb phrases, or describing them as a semantic domain, or choosing to explore focalization are all matters of ideological selection. So we might as well admit the fact and accept the ideological foundations on which we are operating. Such debates within stylistics indicate that the field is far from settled at the theoretical level. It is a strange fact that the emphasis on practical application has meant that stylistics has a generally accepted method and approach while theoretical disagreements about the status of the discipline have continued around a relatively consistent analytical practice. Any differences in stylistic approach tend to arrange themselves along a cline from ‘linguistic stylistics’ to ‘literary stylistics’ (see Carter 1997), reflecting the motivations of the researcher rather than any programmatic political attachment. Linguistic stylisticians tend to be interested in exploring language using literature; literary stylisticians tend to be interested in exploring literature through analysis of its language. The former are more likely to be language teachers and the literary text is the equivalent of the data in applied linguistics. The latter are more likely to be THOC31 748 1/10/06, 1:58 PM Language and Literature: Stylistics 749 cognizant of critical theoretical issues. However, the best stylisticians, in my view, are those who perceive an animating value in both positions. 4 Some Examples of Stylistic Practice A consequence of the expansion of stylistics into matters of socio-cultural and readerly context is that stylistics has also come to be interlinked with related fields such as narratology, social semiotics, critical discourse analysis, cognitive poetics, and other approaches concerned with literary and culturally important texts. To attempt to represent this diversity, even for illustrative purposes only, in a short article such as this would be impossible. In this section, then, I will simply indicate the sort of practical work that has been undertaken under the umbrella of stylistics. For convenience here, examples will be arranged roughly along the linguistic rank scale, and according to the areas of linguistics set out in this Handbook, though it is important to remember that few modern stylistic studies are so exclusively focused. Work in, for example, the point of view of fictional characters might involve an analysis of how lexical choices, modal expressions, the directionality of verbs and other deictic features combine to produce the overall effect and characterization. Early stylistic studies (as mentioned above) were often in the area of poetic metrics, and there has been a recent resurgence of interest in matters of prosody and phonology in poetry. Traditional descriptions of ‘feet’ and metre were enriched by ‘generative metrics,’ which sought to establish the transformational rules by which well-formed stress patterns in poetry were related to an abstract metrical pattern (see Chomsky and Halle 1968; Chatman 1964; and more recently Attridge 1982, 1995; and Fabb 1997). Stylistics shifted attention from metrics as a descriptive labelling to a consideration of the foregrounding patterns in verse: this involved an explanation of how certain features were made more salient than the background pattern, often through repetition, parallelism, positioning or co-occurrence with thematically significant syllables, words or phrases. This allowed phono-aesthetic effects to be explored, without resorting to simplistic equations of sound and sense. The point here was to demonstrate the poet’s craft in organizing the texture as a literary feature in support of the developing meaning of the work. To illustrate with a very simple example, here is the first part of Thomas Hardy’s ‘The Darkling Thrush’: I leant upon a coppice gate When Frost was spectre-gray, And Winter’s dregs made desolate The weakening eye of day. The tangled bine-stems scored the sky Like strings of broken lyres, And all mankind that haunted nigh Had sought their household fires. THOC31 749 1/10/06, 1:58 PM 750 Peter Stockwell Written on New Year’s Eve, 1900, the poem continues to describe the apparent death and starkness of the landscape, explicitly symbolic of the old century’s end. A stylistic analysis would note the uniformly regular rhythm in the prosody here, supported by the repetitions of /p/ in the first line, /s/ in the second, /d/ in the third, and so on. These produce an unremittingly strong emphasis throughout, with heavy pauses at the end of each line in spite of the syntax which runs across the line-endings. In particular there are repetitive consonant clusters /sk/, /st/, /zd/ which often coincide with stressed syllables. Almost every word is mono- or disyllabic, leaving the heaviest emphasis to fall on key content words: ‘Frost,’ ‘spectre,’ ‘Winter,’ ‘dregs,’ ‘tangled,’ ‘scored,’ ‘haunted,’ and so on. The two exceptions are ‘weakening,’ which is itself prosodically weakened in context towards a disyllabic pronunciation as ‘weak’ning,’ and the only other key polysyllabic word which is thus prominent in this stanza: ‘desolate.’ In everyday speech, this word would take heavy stress on the first syllable, and contrastive lighter stress on the second and third syllables. The metre and end-line rhyme position in the poem forces attention on the word, making it difficult to read – especially for inexperienced readers – without a temptation to emphasize the final syllable as ‘late.’ After a second stanza which largely hammers home the same effect as the first, the third stanza begins: At once a voice arose among The bleak twigs overhead In a full-hearted evensong Of joy illimited; An aged thrush, frail, gaunt, and small, In blast-beruffled plume, Had chosen thus to fling his soul Upon the growing gloom. The contrast here is striking, and a stylistic analysis would again draw attention to the differences apparent here against the phonological norms set up by the poem so far, such as the obvious multiple repetition of the vowel in the first line. Notice, too, how lexis is being selected to maintain the patterns already established: ‘illimited,’ not ‘unlimited’; ‘plume,’ not ‘plumage.’ Where the repetitions of consonants and consonant clusters in the first stanza were largely embedded within word boundaries, here they are more properly alliterative as word-initial elements (‘blast-beruffled,’ ‘growing gloom’). The third line of this stanza breaks the monotonous rhythm at the same time as the lexical choices begin to shift from the semantic field of superstition (‘spectre,’ ‘haunted’) to that of religion (‘evensong,’ ‘soul,’ and in the next stanza, ‘carollings’ and ‘blessed Hope’). Again, a stylistic discussion would notice the correspondences between metrics and thematics here, in order to support a particular interpretative line and demonstrate the reading. Though such phonological exploration of poetry most typically remains focused at the micro-level, stylistics has also investigated suprasegmental THOC31 750 1/10/06, 1:58 PM Language and Literature: Stylistics 751 and sociolinguistic aspects of the phonological dimension in, for example, the representation of accents in prose fiction. Given that novelists tend not to write in phonetic notation, the graphological creativity involved in representing Scots (Hugh MacDiarmid, Irvine Welsh, James Kelman), a Dorset accent (Thomas Hardy), Mississippi (Mark Twain) or West African (Ken Saro-Wiwa) forms of pronunciation are all of interest to stylisticians. Again, such studies would not treat the literary representation as dialectological data but as a symbolic representation in which language establishes identity, develops characterization, conveys realist texture, and asserts a political ideology. Notice, of course, how even my simple illustrations here inescapably spill out of the purely phonological level, drawing in semantics, graphology, and syntax however briefly. In a similar way, stylistic analyses which focus on lexical choices are also likely in reality to draw in aspects of syntax and grammar. My own studies of science fictional neologisms, for example, necessarily paid attention to the syntactic positioning, the word-class and the derivations and inflections in context that increase the sense of plausibility and verisimilitude in those science fictional worlds (see Stockwell 2000). Often, the interaction between different linguistic levels serves to signal some literary complexity. For example, surrealist poetry might have a highly normative syntactic form but a highly unusual set of lexical collocations: Philip O’Connor’s ‘Blue bugs in liquid silk/talk with correlation particularly like/two women in white bandages’ is syntactically well-formed and is even suggestive of an explanatory register, except that the semantic sequence is extremely odd. Several W. H. Auden poems set up a serious topic (cancer, death, state repression) in lexical choices from coherent and consistent semantic fields, only to undermine them by setting the poem to a nursery-rhyme style of prosody, in order to signal irony, satire or bathos. These few examples illustrate that both deviant texts and relatively normative texts are amenable to stylistic study, even narrowly at the level of lexis. The lexical choices made in a poem or ascribed to a character in fiction serve as clear markers of the imagined speaker’s perspective, opinions and identity. Naming and pronominal choices, expressions of modality, the selection from among synonyms, and idiomatic forms are often deployed to be consistent with lexical collocates, and with grammatical organization. Stylistic analysis can reveal very subtle differences between characters’ styles of speaking and thinking; when those styles are highly deviant from typical everyday discourse, a stylistic analysis can illuminate the micro-craft of the literary work, suggesting connections between parts of the text that might otherwise have been only subconsciously realized. For example, Steve Aylett’s (1999) novel, The inflatable volunteer, presents a first person narrative that is either set in a rich fantasy world or represents the hallucinatory imagination of the focalizer, Eddie. Eddie’s narrative style generates a sense of discomfort and unease in most readers of the novel; my students describe it impressionistically as very weird, but not weird enough to be completely mad: THOC31 751 1/10/06, 1:58 PM 752 Peter Stockwell Bone midnight Eddie – the little red lizard curled up in a rose. Yeah there’s nightmares and nightmares – you know what I’m saying. I’ve taken part in some where the curtains have caught fire off the devil’s roll-up and the clueless bastard ghosts have barged in late and we were all of us shuffling apologies to the poor sod on whom we were meant to be slamming the frighteners. Torment’s not what it was. Subjective bargaining and the bellyflop of the old smarts flung a spanner in the works an age ago Eddie. That and lack of imagination. Nothing like a spider in the mouth to get you thinking. (Aylett 1999: 5) Here Eddie is talking to himself, and his lexical choices include phrases which are genuine casual idiomatic expressions (‘you know what I’m saying,’ ‘spanner in the works’). However, these are also mixed up throughout the novel with lexical clusters that sound almost idiomatic (‘slamming the frighteners’ ᐅ ‘putting the frighteners on’?), as well as a great number of phrases that have the flavor of idioms but seem to be newly invented (‘the devil’s roll-up,’ ‘Nothing like a spider in the mouth’). The effect of the entire novel is a disconcerting defamiliarization of the world, accompanied by the disorientating effect of the prose style. These effects can be locally identified and explained at the lexical level, where the style of the novel plays a major part in its success. Again, though an analysis focusing on the lexical and phrasal levels would be the most interesting here, a stylistic account of representative passages from the novel would also need to explore the more global features of narratological style and the various shifts in point of view. Even a micro-analysis of selected passages would probably draw in matters of lexico-grammar more broadly, including the syntactic organization and matters of transitivity, for example. Indeed, Hallidayan functional grammar has been a very productive approach in the field of stylistics over the past thirty years. One of the earliest and still most famous such studies was that presented by Halliday (1971), in which he investigated the unusual patterning of transitivity in William Golding’s 1955 novel, The inheritors. Large parts of the novel are written from the point of view of Lok, a neanderthal man living in a community which encounters a more technologically advanced group he calls ‘the new people.’ Halliday shows firstly how Lok’s limited world view is represented by his inability to name new technology: bows and arrows, for him, are unlexicalized, and he has to explain the effect of a stick becoming shorter and longer and a tree next to him acquiring, with a click, a new branch. Halliday develops these observations at the lexical level into an analysis of the transitivity relations in the clauses used by Lok. His focalization is dominated by material action processes and intransitives, in order to represent a simple world view with a limited sense of abstraction, generalization, and cause and effect. As Halliday (1971: 360) points out, ‘In The inheritors, the syntax is part of the story.’ Clearly, in setting out to explore the texture of novels, any stylistic analysis of readable length cannot possibly be exhaustive, and I have mentioned that a process of selection and excerpting of key passages is necessary. This unavoidable selection is part of what makes stylistics an interpretative enterprise rather THOC31 752 1/10/06, 1:58 PM Language and Literature: Stylistics 753 than a mechanistic or purely descriptive approach. Scenes or passages that appear intuitively to be key parts of the text, or which create oddities in readerly sensation, are often good places to begin a more systematic stylistic analysis. It could even be said that the mark of a good stylistician is someone who selects a particular analytical tool best suited to the passage in hand. Sometimes this selection is very obvious: it makes sense to investigate the murder scenes in crime novels in order to discern elements of blame, justification, motive, disguise of the identity of the murderer, and other narratological factors crucial to the novel’s suspense or psychological tension. Carter (1997) for example, explores the transitivity relations in the murder scene of Joseph Conrad’s 1907 novel The secret agent, showing how the agency is deflected from the victim’s murderer and it is inanimate objects and disembodied limbs which appear to act. The murder is thus depersonalized and blame is shifted away from the murderer. For illustration, here is another murder scene: Just after 8.15 p.m. that same evening a man was taking the lid off the highlypolished bronze coal-scuttle when he heard the knock, and he got slowly to his feet and opened the door. ‘Well, well! Come on in. I shan’t be a minute. Take a seat.’ He knelt down again by the fire and extracted a lump of shiny black coal with the tongs. In his own head it sounded as if he had taken an enormous bite from a large, crisp apple. His jaws seemed to clamp together, and for a weird and terrifying second he sought frantically to rediscover some remembrance of himself along the empty, echoing corridors of his brain. His right hand still held the tongs, and his whole body willed itself to pull the coal towards the bright fire. For some inexplicable reason he found himself thinking of the lava from Mount Vesuvius pouring in an all-engulfing flood towards the streets of old Pompeii; and even as his left hand began slowly and instinctively to raise itself towards the shattered skull, he knew that life was ended. The light snapped suddenly out, as if someone had switched on the darkness. He was dead. (Dexter 1991: 517) The reader of this terrifying passage in the crime novel, The silent world of Nicolas Quinn, knows the identities neither of the victim nor his murderer. However, there are several clues in the style of the passage that might pass into the reader’s awareness and can be illuminated through a stylistic analysis. For example, the identity of the victim is kept secret by the careful selection of referential style in the cohesive chain of noun phrases. He is first unspecified as ‘a man,’ then co-referred to simply using the pronouns ‘he’ and ‘his,’ suggesting this is the reader’s first encounter with him (in fact, this is a red herring, since we later find out the victim is Mr Ogleby, a character we have met previously). However, certain definite noun phrases then signal a point of view shift into the man’s head: the proximal deictics of ‘that same evening,’ ‘the lid,’ ‘the . . . scuttle,’ ‘the knock,’ ‘the door’ and ‘the tongs’ all suggest his THOC31 753 1/10/06, 1:58 PM 754 Peter Stockwell familiarity with the contents of the room. In particular, ‘the knock’ (rather than ‘a knock’ here) suggests that the visitor (and his/her knock at the door) was expected and also known to the victim. The reader might even begin to gain a sense of characterization in the style of the passage. Someone who, rather redundantly, specifies ‘Just after 8.15 p.m. that same evening’ appears to be someone who likes precision and is rather fastidious – note also how his coal-scuttle is highly-polished. As the psychological viewpoint in the narrative, he also likes the specification offered by multiple adjectival modifiers: ‘highly-polished bronze,’ ‘lump of shiny black,’ ‘enormous,’ ‘large, crisp,’ ‘weird and terrifying,’ ‘empty, echoing,’ and so on. These aspects of his life contrast sharply with the stark unmodified statement, ‘He was dead.’ The phatic greeting (‘Well, well!’) suggests not only familiarity with the visitor, and a certain warmth (‘Come on in’ rather than, say, ‘Come in,’ or ‘Do come in’), but also offers inferences to be made about the speaker’s age and social class: ‘shan’t’ also supports my sense of an upper middle-class middle- aged educated and rather pedantic man. This sense of his level of education is also perhaps confirmed by the erudite reference to the ancient destruction of Pompeii. In fact, as we later discover, the victim, Mr Ogleby, is an Oxford academic responsible for the examinations system. We discover his murder at the same time as the detective, Inspector Morse, and so even though we have been given access to this striking passage, we share some of Morse’s shock at the news. The subtlety of stylistic clues in the passage also reflects, of course, the piecing together of clues by the detective in the crime novel. In the long third paragraph, noun phrases denoting the parts of his body are used metonymically to stand for him: he is already being stylistically disembodied at the moment of his murder. Furthermore, a quick analysis of the predicate processes in this paragraph reveals that the disembodied limbs are the active participants in material processes (‘right hand held,’ ‘left hand began to raise’). Mostly, though ‘he’ is distanced from the action by being placed as a participant in relational and mental processes: ‘seemed,’ ‘sought,’ ‘willed,’ ‘knew.’ The main actions take place in conditional or subordinate level clauses, relativized by ‘as if.’ The outcome of all of this textual organization is that the victim’s conscious mind is immobilized in his dying body, and his desires for action are rendered unproductive. The first five sentences of the paragraph are extremely hypotactically complex; the final sentence consists of a single clause expressing, ironically, an existential process: ‘He was dead.’ The choice of past tense for the verb here generates particular horror: the sentence plays out for the reader the realization in the mind of Ogleby that in fact he has been dead for the duration of the paragraph. This brief illustrative stylistic analysis, focusing on lexico-grammar, connects the selected passage with matters of characterization, suspense and point of view. In the process, I have drawn briefly on pragmatics, discourse analysis, sociolinguistics, narratology and the cognitive effects of cohesion. At these macro-linguistic levels, it is easy to see the possible linkages to be made with THOC31 754 1/10/06, 1:58 PM Language and Literature: Stylistics 755 more purely literary concerns such as characterization, narrative structure, tone and atmosphere, genre, texture, realism and viewpoint, for example. From the standpoint of more well-established branches of linguistics, this practice might look hopelessly eclectic. However, for stylistics to account fully for the organizational patterns and readerly effects of literary works, such a wideranging approach is essential, since the object of study itself is various, protean and complex. As a result of its interdisciplinary contact with critical and cultural theory, modern stylistics is currently addressing itself to providing a principled account of the textural complexity of literature. 5 Emerging Work in Stylistics There is a growing body of work in stylistics which marries up detailed analysis at the micro-linguistic level with a broader view of the communicative context. Indeed it is this integrative direction that seems to me to characterize the various emerging concerns of the discipline. Of the numerous different developments that I outline below, all have in common the basic stylistic tenets of being rigorous, systematic, transparent and open to falsifiability. All set out to draw the principled connections between textual organization and interpretative effects. In short, they present themselves as aspects of a social science of literature, rather than a merely poetic encounter with the literary. Modern stylistics continues the century-old tradition of denying any separation of interpreted content from textual form, and it is interesting to note books and courses appearing which exchange the term ‘stylistics’ for the term ‘literary linguistics,’ reappropriating the Russian Formalists’ term for themselves. In this respect, stylistics necessarily involves the simultaneous practice of linguistic analysis and awareness of the interpretative and social dimension. The act of application is what makes stylistics a fundamentally singular discipline of applied linguistics, arguing that formal description without ideological understanding is partial or pointless. If there is a paradigm in stylistics, it is this, and it seems to me to make stylistics a unified discipline at heart, with spin-offs into history, social study, philosophy and literary archaeology, as practiced in literature departments around the world. The discipline of stylistics is currently drawing much of this work to itself. For example, studies of the sociolinguistics of writing have led to a renewed emphasis on the various literatures of the world in different international Englishes. The ways that writers use different vernaculars to represent a greater richness of cultural voices are being explored stylistically. These studies include explorations of particular authors and communities around the world, as well as more theoretical work on how ‘voice’ is represented in literature. The holistic sense of ‘voice’ involves many of the historical concerns of stylistics: mind-style, character viewpoint, deixis, modality, and so on. In some respects, THOC31 755 1/10/06, 1:58 PM 756 Peter Stockwell the current interest in voice represents a re-evaluation of these textual patterns renewed through the readerly construction of the psychology of the speaker. Also along the readerly dimension, a major evolution in stylistics has been the development of ‘cognitive poetics’ (also called ‘cognitive stylistics’). Applying the growing field of cognitive science to the experience of literary reading has been generating many interesting new insights into literature. These range from the almost purely psychological to the almost purely textual, but the vast majority of cognitive poetic studies combines our understanding of readerly cognitive processes with textual reality in the stylistic tradition (see Stockwell 2002a; Gavins and Steen 2003; and Semino and Culpeper 2002). Cognitive poetics adds new facilities to stylistics, enabling the field to address key current issues such as a principled account of ‘texture,’ an understanding of how the thematics of reading a literary text works, or how a piece of literature can generate and sustain emotion. These developments simply extrapolate the continuing evolution of stylistics towards encompassing matters that were traditionally the ground of literary critics alone. Underlying much of this principled interest in social and psychological context is a renewed sense of ethics in stylistic research. Non-literary stylistic analysis has developed through critical linguistics and critical discourse analysis alongside stylistics: the interaction between the two fields has been constant and close and consequently very productive (see Fairclough 1995; and Mills 1995, for example). Along with the ethical awareness that the literatures of the world ought to be studied sociolinguistically, fields such as feminist linguistics have worked to remind stylisticians (and all applied linguists) of our ethical responsibilities and the impossibility of an ideologically neutral linguistic theory. Stylistics has also continued to draw on methodological innovations in linguistics. In particular, corpus linguistics and the use of computerized concordances and other empirical analytical tools have revolutionized the systematic study of literary texts. (See Thomas and Short 1996.) The continuities between literary creativity and the creativity apparent in everyday discourse have been revealed in all their complexity largely out of the fruitful interaction of stylistics and corpus linguistics. New methods such as these can be used to explore levels of language from lexical collocations right up to narrative organization. At the same time, the pedagogical element in stylistics has also developed strongly. Stylistic methods are now the paradigmatic approach in the foreign language classroom, and the applied study of creativity is becoming standard in native-speaker language teaching too. Stylistics, as a discipline, is therefore very much in its heyday. It is a progressive approach in the sense that stylisticians strive constantly to improve their knowledge of how language works, while at the same time being aware of the useful insights of its own tradition. Its challenges arise from an apparently boundless appetite for drawing in the different disciplines and levels of language study, and the desire of its practitioners to be at once rigorously disciplined and also engaged and passionate about verbal art. THOC31 756 1/10/06, 1:58 PM Language and Literature: Stylistics 757 FURTHER READING Bex, T., Burke, M., and Stockwell, P. (eds.) (2000) Contextualised stylistics. Amsterdam: Rodopi. Carter, R. (ed.) (1982) Language and literature: an introductory reader in stylistics. London: Allen & Unwin. Carter, R. and Simpson, P. (eds.) (1989) Language, discourse and literature: an introductory reader in discourse stylistics. London: Unwin Hyman. Culpeper, J., Short, M., and Verdonk, P. (eds.) (1998) Exploring the language of drama: from text to context. London: Routledge. Fowler, R. (1996) Linguistic criticism, 2nd edn. Oxford: Oxford University Press. Pope, R. (1995) Textual intervention: critical and creative strategies for literary studies. London: Routledge. Short, M. (1996) Exploring the language of poems, plays and prose. Harlow: Longman. Simpson, P. (2004) Stylistics: a resource book for students. London: Routledge. Verdonk, P. (ed.) (1993) Twentieth century poetry: from text to context. London: Routledge. Verdonk, P. and Weber, J.-J. (eds.) (1995) Twentieth century fiction: from text to context. London: Routledge. Wales, K. (ed.) (2001) A dictionary of stylistics, 2nd edn. Harlow: Longman. Weber, J.-J. (ed.) (1996) The stylistics reader. London: Arnold. Journals Genre Journal of Literary Semantics Language and Literature Language and Style Poetics Poetics Today Social Semiotics Style REFERENCES Attridge, D. (1982) The rhythms of English poetry. Harlow: Longman. Attridge, D. (1995) Poetic rhythm: an introduction. Cambridge: Cambridge University Press. Aylett, S. (1999) The inflatable volunteer. London: Phoenix House. Brumfit, C. J. (ed.) (1983) Teaching literature overseas: language-based approaches. Oxford: Pergamon. Carter, R. (1997) Investigating English discourse. London: Routledge. Chatman, S. (1964) A theory of metre. The Hague: Mouton. Chomsky, N. and Halle, M. (1968) The sound pattern of English. New York: Harper & Row. THOC31 757 Dexter, C. (1991) The silent world of Nicolas Quinn. London: Macmillan. Fabb, N. (1997) Linguistics and literature. Oxford: Blackwell. Fairclough, N. (1995) Critical discourse analysis. Harlow: Longman. Fish, S. (1980) What is stylistics and why are they saying such terrible things about it? In S. Fish (ed.). Is there a text in this class? The authority of interpretative communities (pp. 68–96). Cambridge, MA: Harvard University Press. Fowler, R. (1971) The languages of literature. London: Routledge & Kegan Paul. 1/10/06, 1:58 PM 758 Peter Stockwell Fowler, R. (1981) Literature as social discourse. London: Batsford. Gadamer, H. G. (1989) [1965]. Truth and method. Trans. J. Weinsheimer and D. G. Marshall. New York: Crossroad Press. Gavins, J. and Steen, G. (2003) Cognitive poetics in practice. London: Routledge. Halliday, M. A. K. (1971) Linguistic function and literary style: an inquiry into the language of William Golding’s The inheritors. In S. Chatman (ed.). Literary style: a symposium. London: Oxford University Press, 330–65. Harris, R. (1980) The language myth. London: Duckworth. Harris, R. (1981) The foundations of linguistic theory. London: Routledge. Harris, R. (2000) When will stylistics ever grow up? Paper presented at the 20th PALA conference, Goldsmith’s College, London, July 2000. Ingarden, R. (1973a) [1931]. The literary work of art: an investigation on the borderlines of ontology, logic, and theory of literature. Trans. G. Grabowicz. Evanston, IL: Northwestern University Press. Ingarden, R. (1937; 1973b) The cognition of the literary work of art. Trans. R. A. Crowley and K. Olson. Evanston, IL: Northwestern University Press. McCarthy, M. and Carter, R. (1994) Language as discourse: perspectives for language teaching. Harlow: Longman. Mills, S. (1995) Feminist stylistics. London: Routledge. Semino, E. and Culpeper, J. (2002) Cognitive stylistics. Amsterdam: Benjamins. THOC31 758 Spitzer, L. (1948) Linguistics and literary history: essays in stylistics. Princeton, NJ: Princeton University Press. Stockwell, P. (2000) The poetics of science fiction. Harlow: Longman. Stockwell, P. (2002a) Cognitive poetics: an introduction. London: Routledge. Stockwell, P. (2002b) A stylistics manifesto. In S. Csábi and J. Zerkowitz (eds.), Textual secrets: the message of the medium. Budapest: Eötvös Loránd University, 65–75. Thomas, J. and Short, M. (eds.) (1996) Using corpora for language research. Harlow: Longman. Toolan, M. (1990) The stylistics of fiction: a literary linguistic approach. London: Routledge. Toolan, M. (1996) Total speech: an integrational linguistic approach to language. Durham, NC: Duke University Press. Ullmann, S. (1964) Language and style. Oxford: Basil Blackwell. Wellek, R. and Warren, A. (1949) Theory of literature. Harmondsworth: Penguin. Widdowson, H. (1975) Stylistics and the teaching of literature. Harlow: Longman. Widdowson, H. (1992) Practical stylistics. London: Oxford University Press. Wimsatt, W. K. and Beardsley, M. C. (1954a) The affective fallacy. In W. K. Wimsatt (ed.), The verbal icon. Lexington, KY: University of Kentucky Press, 21–39. Wimsatt, W. K. and Beardsley, M. C. (1954b) The intentional fallacy. In W. K. Wimsatt (ed.). The verbal icon. Lexington, KY: University of Kentucky Press, 3–18. 1/10/06, 1:58 PM