Stylistics

Cognitive Stylistics Cognitive Poetics

Despite cognitive stylistics, also known as cognitive poetics, having only relatively recently been embraced by mainstream stylistics, it has rapidly become an ever expanding, entrepreneurial and extremely productive branch. At its most basic, a definition can comprise a single sentence: "Cognitive poetics is all about reading literature" (Stockwell, 2002, 1). On a more complex level, Stockwell expands it to:

That sentence looks simple to the point of seeming trivial. It could even be seen simply as a close repetition, since cognition is to do with the mental processes involved in reading, and poetics concerns the craft of literature. (Stockwell, 2002, 1)

Viewed from this perspective, cognitive stylistics/poetics highlights the aspects of reading that literature consumers operate when they process literary texts. Cognitive stylistics, essentially, has emanated from the application to literature of models originally used in disciplines such as cognitive linguistics, cognitive psychology and artificial intelligence. Of special relevance are the multiple frameworks in which this branch has managed to capture issues such as "what do people do when they read" and "what happens to people when they read" (Burke, 2006a, 218).

Because of the data that cognitive stylistics is concerned with, i.e., literature, this branch is intricately linked to literary stylistics, alternatively known as literary linguistics. In fact, cognitive stylistics is said to have derived directly from it (Burke, 2006a, 218). By prioritizing the textual components of literature, literary stylistics embodies the most traditional ways of stylistic analysis based on the interface between form, function, effect and interpretation whereas cognitive stylisticians argue that the mental component of the meaning creation process should be included. Influences from disciplines such as psychology, cognitive psychology and cognitive linguistics are responsible for shifting the emphasis to take into consideration the mental aspects of reading too.

For instance, schema theory is one of such disciplines that, although originating from Gestalt psychology, has been extremely influential in bringing stylistics to the cognitive camp. Schema theorists claim that meaning is not only contained in the text; meaning needs to be built up by the reader using the

Key Terms in Stylistics

text in negotiation with their own background knowledge. These two essential facets of understanding, which are complementary and dependable on one another, are known as bottom-up or stimulus-driven processes and top-down or conceptually-driven processes (Rumelhart and Ortony, 1977, 128). The former prompt the reader to construct a particular mental world thanks to the linguistic characteristics of the text, whereas the latter mobilize the background knowledge that the reader is already in possession of and that becomes activated when prompted by the specific linguistic props. Most sub-branches within cognitive stylistics accept that this negotiation is essential if we are to provide an accurate account of how understanding actually takes place.

Although the terms cognitive stylistics and cognitive poetics can and are used interchangeably, certain practitioners point out some slight differences in meaning. The first edited collection on cognitive stylistics research, published by Semino and Culpeper (2002), defends the former term:

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. (Semino and Culpeper, 2002, ix)

By maintaining the term "stylistics" in the label, the rigor and replicability of stylistic methods is also safeguarded because it is underscored. These differences, however, appear unnecessarily fastidious to some scholars that consider both labels totally interchangeable. The variety of frameworks and models of analysis found under the umbrella term cognitive stylistics/poetics also evidences the healthy status of the discipline. Some of these frameworks include blending theory (Dancygier, 2005, 2006), conceptual/cognitive metaphor theory (Steen, 1994), contextual frame theory (Emmott, 1997), schema theory (Cook, 1994; Semino, 1997) and text world theory (Gavins, 2007; Werth, 1999). Although these frameworks differ as to how they explain the relationship between bottom-up and top-down processes (contextual frame theory uses "frames,"schema theory uses "schemata" and text world theory uses "text worlds," for instance), all models rely on the notion of mental constructs.

Corpus Stylistics

Corpus stylistics has only recently entered the landscape of stylistics, but it has all the more forcefully begun to exploit the potential of combining corpus linguistics and stylistics. Defining corpus stylistics as the cooperation

11

Key Terms in Stylistics

By developing a particular style, a producer of a spoken or written text privileges certain readings, certain ways of seeing things, while suppressing or downplaying others. [...] The purpose, in other words, is to probe under the surface of language, to decode the stylistic choices which shape a text's meaning. (Simpson, 1993,8)

For this probing. Simpson explores linguistic phenomena such as, for instance, the manifestation of attitude through language (modality), the linguistic construction of experience (transitivity) as well as pragmatic aspects of meaning-making. Finally, Jeffries in Critical Stylistics (2010) fuses critical discourse analysis and stylistics with a strong emphasis given to-and tools provided for the analysis of the actual linguistic manifestation of social meanings.

Emotion: Stylistic Approaches

Affective and emotional approaches to stylistics are recent additions to the eclectic range of formulations that stylistics feeds on. Such up-to-datedness, however, does not imply a totally novel way of looking at (mainly although not exclusively) literary texts as proved by Aristotle's concerns with the emotional aspects of reading (ie., his theory of catharsis or "purging").. This historical perspective notwithstanding, stylistic approaches to emotion have recently received a new impetus as more and more scholars have started to incorporate affective components into their analyses. The first hurdle to negotiate, though, is a disambiguation of the label itself. "Emotion," "affect," "feelings" and "mood" are terms amply discussed and defined in psychological and cognitive circles where they are not treated as synonymous. "Affect" is generally considered to be the most general term, "used to include emotions, moods and preferences" (Oatley et al., 2006, 412). "Emotion" refers to a more complex set of affairs, "typically a multi-component response to a challenge or an opportunity that is important to an individual's goals" (Oatley et al., 2006, 415). Some of these multi-faceted components include a conscious mind, bodily changes, face expressions, gestures or a marked tone of voice and finally readiness for action. Despite these two terms invoking different meanings in disciplines such as cognitive psychology and psychology, stylistics has generally conflated the two and treated them synonymously: "Basically affective refers to feelings, hence it means emotional'" (Wales, 2001, 10). Besides, stylistically-informed analyses have succeeded in looking into the emotional components of literary discourse as a whole, whether these affect the production level (author-induced emotion), the textual level (linguistic means) or the reception level (reader response).

Initially, the views propounded by traditional affective criticism gave rise in the 1940s and 1950s to what the New Critics called the "affective fallacy"

(Wimsatt and Beardsley, 1954). On the one hand, the traditional affective views following the Aristotelian tradition investigated the potential of a given text to raise some kind of emotional reaction in the reader, including a physical one. so the focus for the inducing of emotions was identified as emanating from the text The New Critics dismissed such stance as unnecessarily subjective and called it the "affective fallacy" as an evaluation of literature solely based on the emotional impact it would have on readers was considered methodologically and formally inappropriate. This position does not seem tenable nowadays, though; as Burke states, this "anti-subjective view is completely implausible because of what is now known about the crucial role that top-down processing also plays in reading procedures" (Burke, 2006b, 127). The belief in the potential of texts for raising emotions was carried through and picked up by the formalist scholars of the 1960s as exemplified in Roman Jakobson's formulations on language (1960). Despite a reticence to deal with emotional issues openly because of his formalist background, Jakobson famously identified six different functions of language among which the "emotive" is to be found alongside the "conative," "metalinguistic," "poetic," "phatic" and "referential." It could be argued, therefore, that even those movements that claim to define themselves as devoid of "subjective" content seem to have incorporated some e notional component in their formulations. It is not until the 1980s that the affective aspects of the reception level of discourse (that is, that of the reader) are highlighted, particularly in reader response criticism. The label "affective stylistics" was famously coined by Stanley Fish (1980), who states:

I have argued the case for a method of analysis which focuses on the reader rather than on the artefact [...]. The chief objection, of course, is that affective criticism leads one away from the "thing itself" in all its solidity to the inchoate impressions of a variable and various reader [...]. In the category of response I include not only 'tears, prickles, and other psychological symptoms' but all the precise mental operations involved in reading, including the formulation of complete thoughts, the performing (and regretting) of acts of judgement, the following and making of logical sequences. (Fish, 1980, 42—3)

Fish's innovative interpretation not only shifts the focus of attention onto the reader, but also underscores the importance of including an account of the psychological processes involved in reading. Nevertheless, his initial defense of an investigation into readers' emotional concerns became weakened and never developed into a fully-fledged theory of emotional responses to literature. Moreover, interest in emotional aspects waned after Fish's initial surge despite the fact that cognate disciplines such as Systemic Functional Linguistics and discourse analysis embraced the challenge fully (Burke, 2006b, 127). The

Key Terms in Stylistics

branch of stylistics known as cognitive stylistics or cognitive poetics has more recently become the catalyst for a rekindling of the interest in the role of affect. Semino's analysis of schema theory (1997), on the one hand, and van Peer's appraisal of a poeücs of emotion (1997), on the other, are a case in point:

Emotions, then, are intimately related to cognition. Thus, in assessing the ernotional potential of literature, we shall have to take this relation into account [...]. Reading literature is [...] one such form in which our emotional involvement has clear cognitive overtones. (van Peer, 1997,227)

Van Peer signals here the intricate connections existing between cognitive components of human comprehension and their emotional counterparts, something that has not always been accepted as a given in the field of psychology. Van Peer's adamant emphasis on the existence of such connections seems to explain the newly accepted take on emotional concerns emanating from cognitive analyses of literature. Finally, it is worth noting that a great deal of recent work on affective responses to literature has highlighted emotional aspects at the reception level of discourse, that is, the reader's emotional reaction rather than the encoding of emotion in language.

Empirical Study of Literature

Empirical perspectives on the study of literature advocate the implementation of a rigorous, primarily, although not exclusively, quantitative study of literary texts by adopting an observational and analytical perspective. The empirical study of literature (ESL) feeds directly from methodologies employed in the social sciences such as anthropology, psychology, cognitive psychology and psychologists and argues for the applicability of their most characteristic tools of these methodologies to literary analysis. This interest in rigor highlights that analyses of literature and the arts in general can actually be conducted in as scientific a way as those carried out in the social sciences. This way of looking into literary concerns is a newly taken enterprise, especially adopted by the members of the International Society for the Empirical Study of Literature and Media (IGEL) founded in 1987.

Despite this defense of the use of testable methods for the analysis of literature, literary empiricists maintain that their work falls well within the boundaries of literary analysis:

We shall argue in the next pages that the empirical study of literature embodies a shift of perspective and emphasis in comparison with more traditional literary scholarship, but that it remains within the bounds of literary studies proper. (Schram and Steen, 2001, 2)

As such, these empirical studies deal with issues of literature as a cultural and social artefact, comparable to other artistic and media forms. Of special importance in this approach is the emphasis placed on the role of the reader as a cultural and social participant, which has subsequently resulted in a much tighter and methodologically sound treatment of studies of reader response. It is not surprising, thus, that empirical studies tie in especially well with general stylistic interests as the principles of applicability, testability, falsifiability and retrievability that characterize stylistics are particularly enhanced by an empirical perspective. As with the former, stylistics, in general, maintains that any methodology that encourages a rigorous perspective to look at form in combination with function, effect and interpretation will be of benefit.

The fruitful nature of empirical approaches is also reflected in the interdisciplinarity that characterizes this take on literary texts. Scholars whose work falls well within the remit of stylistics have collaborated closely with colleagues in psychology and cognitive psychology departments. One clear case of this collaboration is the STACS project (Stylistics, Text Analysis and Cognitive Science: Interdisciplinary Perspectives on the Nature of Reading) (Emmott et al., 2007, 213) at the University of Glasgow, which has been successful in bringing together tools traditionally used in stylistics such as foregrounding, and psychological testing methods, such as "anomaly testing" and "text change detection method" (Emmott et al., 2007, 204-5). As Hakemulder states, "the empirical study of literature (ESL) concerns all aspects of literary communication" (2006,274); so, empirical testing functioning at the levels of production and reception is especially conducive to the integration of all these diverse disciplines and methodologies. As a result, quite a lot of empirical work has focused on aspects of linguistic foregrounding in relation to its effect on the reader. Emmott et al. (2007), for instance, analyze how traditional stylistic takes on fiction tend to focus on the way salient features can underscore "key plot and thematic moments [...] guiding the interpretation of readers at these points" (2007, 205). They argue that looking at those salient features tends to be done by enticing readers to produce accounts of those features they consider to be especially striking. The main problem arising out of assessing foregrounding effects in this way is the inevitable subjectivity colouring respondents' accounts. Moreover, as Emmott et al. point out (2007, 207), such technique might provide significant results as far as the overall impression of the text is concerned, but not in relation to the actual way in which psychological processing takes place. When the effect on the reader is measured by employing the rather more accurate and strict methods generally used in the social sciences results can be not simply different but sometimes even contradictory to those based on subjective evaluation of text saliency.

Key Terms in Stylistics

As is common practice in academia, though, some discordant voices claim that empirical approaches to literature also have their disadvantages. Schreier (2001, 35), for instance, points out that the perspective of over 25 years that empirical studies now enjoy has allowed some much needed objectivity on the ironically!) objective claims made by ESL practitioners:

There have been voices of scepticism or even discontent from within ESL, voices focusing on the question of fit between the subject matter of ESL and the methods used to study it. Brewer (1995), for instance, has criticized that the methods used in ESL for data collection have focused on the cognitive-instrumental aspects of reception to the exclusion of, for instance, affective or aesthetic ones. Groeben (1994) has stressed the necessity of adapting methods so as to suit current changes and extensions of the subject matter of ESL, such as the inclusion of other media; and Ibsch has repeatedly expressed concem over the neglect of complex issues, such as the reception of longer texts or text reception by experts

(1994, 1998). (Schreier, 2001,35) To the above, Schreier also adds the problem of how to analyze longer stretches of text (indeed, novels), how to control the variables respondents are being assessed on (questionnaires and multiple choice forms to be filled in on computer screens) and how to collect "think aloud protocols" (Schreier, 2001, 35). In sum, there have been suggestions that vouch for an inclusion of the socalled qualitative methods as they have been developed in the social sciences (Andringa, 1998; Groeben, 1994; Ibsch, 1998)" (Schreier, 2001, 35).

Feminist Stylistics

Feminist stylistics aims at utilizing stylistic tools for the investigation of those concerns and preoccupations traditionally identified in feminist 19 approaches to the study of language. Like feminist studies in general, a

feminist stylistic perspective is keen to flag up gender issues although the focus crucially shifts to the linguistic (and also multimodal) manifestations of these concerns. As Mills puts it: "Feminist stylistics is concerned with the analysis of the way that questions of gender impact on the production and interpretation of texts" (Mills, 2006, 221). She goes on to describe the way this branch has developed from its incipient applications to the present moment and highlights its main focus as being the following:

Rather than assuming that notions of gender are simply a question of discriminatory messages about sex difference embedded in texts, feminist stylistics is concerned with unraveling the complex messages which may be deduced from

texts and also with analyzing the way that readers piece together or resist these messages. (Mills, 2006, 221)

Contrary to what might have been the case in the past, recent feminist views on the crucial role of language to project social and political standpoints is not circumscribed to perpetually alleging the existence of discriminatory values. Instead, feminist stylistic views are more interested in spelling out those values that do exist in texts, whether these may be prototypically patriarchal or not. In addition, recent feminist stylistic positions also acknowledge that binary considerations of gender as simply male or female are deeply reductive as neither males nor females form a homogeneous or discrete group.

If a feminist perspective is to continue being successful, Mills claims (2006, 221), it is necessary that scholars are capable of moving on from an exclusive textual analysis performed at the micro-level of language that is, the use of the generic "he," or generic nouns to encode sexism), to a more comprehensive discourse level which will ensure, for instance, the investigation of linguistic structures such as direct or indirect speech, and the way these are exploited with reference to male and female characters, or the study of lexical collocations in relation to the prototypical language patterns associated with male and female textual entities (Mills, 2006, 221).

Feminist stylistics scholars have been particularly prolific at producing accounts of the way the micro-level of language encodes ideologically-loaded messages, especially those in which female characters are presented in a disadvantageous social position. The now classic study by Burton (1982), for instance, illustrates this point clearly. Her analysis of Sylvia Plath's semi-autobiographical novel The Bell Jar (1963) highlights the main protagonist's lack of control when, due to her precarious mental state, she is taken to a mental institution and is subjected to electric shock treatment. Burton uses a transitivity analysis to illustrate the protagonist's powerlessness as she is never presented as an actor in the text. Instead, most of the many material action processes used in the extract discussed by Burton identify the female protagonist as the goal of those actions in a way that underscores her incapacitated state. As Simpson confirms, "Burton argues provocatively for a political dimension in textual interpretation and suggests that links between literary analysis and political standpoint can be articulated clearly through systematic and principled methods of analysis" (Simpson, 2004, 185–6), for which endeavor feminist stylistics seems especially suited.

This focus on the marriage of political and social dimensions to literary linguistic analysis has been taken on successfully elsewhere. In a collection of essays on female writing edited by Wales (1994), female scholars

investigate not only the advantages but also the necessity of looking at feminism from the perspective afforded by linguistic means:

They present an original and close analysis of a "literary" text, or range of texts, by applying the methodology or framework of linguistic (grammatical, lexical, pragmatic, discourse) theories, in order to address directly questions and ideas that have been raised in feminist literary theory, criticism and linguistics about gender and style. (Wales, 1994, vii)

The tradition established by Burton can be felt in this volume as the contributors discuss the pervasiveness of gender concerns in different types of discourse as well as across various periods in time, all the time striving to account for such concerns via linguistic frameworks. Thus, Jeffries (1994) analyses the issue of "apposition" in contemporary female poetry, Wareing (1994) looks at the submissive, passive roles generally associated with popular fiction protagonists and Calvo (1994) concentrates on the discourse tactics used by the character Celia in Shakespeare's As You Like It (1599), especially as far as her positive politeness strategies are concerned. This focus on textual practice typically seen in the 1990s has not waned as can be seen in Jeffries's (2007) recent work on the textual manifestations of the female body. Faithful to a primarily linguistic focus, this monograph looks at the way women's bodies are characterized, discussed and textually constructed in women's magazines. Jeffries argues that such construction both mirrors and projects twenty-first century preoccupations such as appearance, looks or weight concerns. Mills (1995) has previously done some work on the construction of the female self, but has successfully also included multimodal analyses of printed adverts in which textual features are combined with various other graphological and photographic components to similarly embed prototypically patriarchal values in advertising for women.

Finally, because of their widespread impact, feminist positions have also been appropriated and exploited in narratological approaches to fiction:

In its broadest sense, feminist narratology embraces the study of narrative (including its formal features, interpretation, and function) with particular attention to the ways in which these might inform or be informed by aspects of feminist theory. (Page, 2006, 482)

Further to the above, Page (2007b) advocates elsewhere for the complementarity of feminist narratological and feminist stylistic perspectives on fiction. She states that whereas feminist narratology has traditionally been saturated with preoccupations on aspects such as plot, focalization or voice, feminist stylistics has opted to concentrate on what Mills defines above as the

microlevel of language, that is, the use of pronouns, nouns and phrases, among other linguistic structures. Other stylistic scholars prefer to merge the two branches or simply to place them in a continuum rather than seeing them as discrete categories.

Film Stylistics/The Stylistics of Film

The application of traditionally textual tools of analysis to the study of film and moving images has resulted in a new approach within stylistics known as film stylistics or the stylistics of film. Some stylistics practitioners have claimed that many of the frameworks used in textual analysis can in fact be used to explicate formats other than the printed text. As is the case with textual stylistics, film stylistics aims for a more retrievable way of analyzing cinematic forms based on frameworks which have already proven successful in the study of textual forms. McIntyre (2008), for instance, points out that this interest should hopefully result in a better understanding of the general construction of meaning in a variety of formats:

It is also the case that stylisticians should consequently find themselves better able to describe and explain how particular textual effects are realized, how readers' interpretations are constructed and how these can be supported through analysis. Some stylisticians have already begun to engage with such issues, providing analyses of texts which incorporate significant multimodal elements (see, for example, Boeriis and Nørgaard, 2008 [sic]). Nonetheless, there remains a substantial amount of work to be done in this area. (McIntyre, 2008,310)

It is also hoped that by incorporating new forms into the broad spectrum of textual material already used by stylisticians the discipline as a whole will benefit from the new challenges that these new formats will no doubt bring along. Nevertheless, such a recent interest on the part of stylisticians to look at cinematic varieties does not mean that filmic forms have not been previously investigated by other scholars. On the contrary, "film studies" is a well-established discipline that has been concerned with the structural and functional features of film for quite some time. Some of its main proponents have, rather successfully, applied a variety of approaches to the study of films such as considerations of aspects of the theory of film (Carroll, 1996a, 1996b; Tredell, 2002), evaluations of movies as cultural constructs (Kellner, 1999; O'Regan, 1999), or assessments of filmic forms from a psychoanalytic perspective (Allen, 1999, Hochberg and Brooks, 1996), among others. A stylistic slant on the analysis of cinema cannot and should not do away with all

instead, it should incorporate all the findings from film studies as the basis on which to build further perspectives.

A particularly rich area of study for film analysts has been that of cinematic adaptations of (mainly, although not exclusively) fictional novels. One of the topics that adaptation scholars tend to raise is the way in which the original textual versions are transposed into a new medium, which also generally gives rise to the issue of fidelity of the latter towards the former. For instance, McFarlane (1996, 2000), Thomas (2000) and Whelehan (1999) all consider the relationship between the two forms but, as McFarlane (2000) points out, this is a question on which even the general public feels entitled to pass comments:

It is [...] quite common to come out of a cinema after viewing an adaptation or to engage in casual conversation about it afterwards and to hear such comments as 'Why did they change the ending?" or 'she was blonde in the book' or, almost inevitably, I think I liked the book better.' (McFarlane, 2000, 165)

Besides this concern with the faithfulness with which textual forms are translated into cinematic formats, film scholars interested in incorporating stylistic approaches into their analyses have focused on the way existing stylistic frameworks can aid to explain fiction's potential for trans-medial manifestations. For instance, Forceville (2002a) suggests that the already-existing models for narrative analysis should be re-evaluated in light of the increasing multimodal manifestations of fictional forms:

Since stories increasingly take on pictorial and mixed-medial forms, narratology needs to investigate to what extent narrative devices exceed the boundaries of a specific medium. One way to examine this issue is to focus on film adaptations of narratologically complex novels or stories. (Forceville, 2002a, 119)

Forceville succeeds in achieving such an aim by looking into the way nonverbal means are exploited in the filmic version (Schrader, 1990) of Ian McEwan's The Comfort of Strangers (1981) to convey the same sense of confusion created via free indirect speech and thought in the printed form. Other traditionally stylistic phenomena whose film manifestations have been analyzed are mind style (Montoro, 2010a, 2010b), character dialogue (McIntyre, 2008), the functioning of speech acts in film dialogue (Short, 2007b) and the possible pedagogical-stylistic applications of film (Montoro, 2006b).

Finally, stylisticians are also beginning to pay attention to the rather enriching research output emerging from scholars working within a systemic-functional tradition and multimodal texts. Of special interest are the new attempts at combining computerized methods of analysis with the study of the moving

image, broadly understood as those formats which employ dynamic pictures, be it in the form of films, adverts, documentaries, computer-generated imagery, etc. For instance, Baldry and Thibault (2006) have devised a rather exhaustive way of looking at multimodal texts based on multimodal concordancers which are capable of monitoring several items (sound, movement, colour or verbal input) in the various semiotic modes which can simultaneously be at work in the formats encompassed by the notion of the moving image, that is the sonic, the musical, the pictorial or the linguistic modes. Their rather thorough annotated descriptions of multimodal texts allow for a more comprehensive analysis of film, adverts or documentaries as multi-layered pieces of discourse.

Formalist Stylistics

Formalist stylistics refers to the type of stylistic work done from the 1910s to the 1930s by a diverse group of theoreticians known as the Russian Formalists and later taken up by stylisticians, especially in Britain and the United States, in the 1960s and early 1970s. The Russian Formalists were a fairly heterogeneous group of people consisting primarily of members of the Moscow Linguistic Circle (founded 1915) and the Petrograd Society for the Study of Poetic Language (founded 1916). Common to these were an interest in poetic language and a wish to make literary inquiry more "scientific" by modeling it on linguistics and thereby anchoring it solidly in observations about the formal features of the texts in question. The overriding interest of the formalist approach was in poetic form, or "literariness" in Jakobson's terminology (1960), which led to a focus on elements of the literary text which made it "literary" and set it apart from other types of text.

According to Jakobson's model of communication (1960), the poetic function of language is dominant in texts which "focus on the message for its own sake," i.e., in texts where lexical, grammatical or phonological choices, for instance, draw attention to themselves and hence to the poetic nature of the text. Formal features such as parallelism and deviation from the linguistic norm are seen as stylistic features which would mark the text as literary, or poetic. While the poetic function is seen as the dominant function of poetry, it is not exclusive to that genre, but may also occur in other types of text as in Jakobson's own example of the political slogan for Dwight D. Eisenhower: "I like Ike." A similar approach to the poetic function of language is expressed by Victor Sklovskij, who introduced the concept of "defamiliarization" (ostranenie,"making strange") as a central aspect of the technique of art:

The technique of art is to make objects "unfamiliar," to make forms difficult, to increase the difficulty and length of perception because the process of perception is an aesthetic end in itself and must be

prolonged. Art is a way of experiencing the artfulness of an object, the object is not important. (Sklovskij, [1917) 1988, 20; Šklovskij's italics)

In šklovskij's terms, the function of art is thus to defamiliarize the familiar to make us re-perceive what we have stopped noticing because of its familiarity and to make us recognize the artfulness of the expression itself. In line with formalist thinking, the Russian folklorist, Vladimir Propp, broadened the scope of formalist enquiry by setting out to identify the basic plot components and structures of folk narrative, resulting in his Morphology of the Folktale (1928).

Without doubt, the work done by the Russian Formalists was seminal to the growth of stylistics in the 1960s and early 1970s, but the formalist approach to stylistic analysis has also had its critics. In particular, formalist stylistics has been criticized for its overriding interest in linguistic form at the expense of considerations about the function and effects of the formal features put up for examination. Another point of criticism concerns the tendency in formalist stylistics to investigate literature in isolation from contextual factors such as the social and historical contexts of the text. In Weber's words,

the problem with these formalist stylistic analyses is that they strike one as mechanical, lifeless, sterile exercises, and largely irrelevant to the interpretation of the literary work that they are describing. And if the critics try to ascribe some function or meaning to the formal patterns that they have uncovered, then a huge leap of faith is

required to move from description to interpretation. (Weber, 1996,2)

Others, as, for instance, Stanley Fish (1973), have criticized formalist stylistics for its claim to scientific objectivism and for ignoring the role of the reader in the identification of stylistic effects.

Later examples of formalist stylistics were those inspired by Chomsky's generative grammar. Here the focus was likewise placed on form, but the analysis now pivoted on the rules which lie behind the generation of grammatical sentences. Altogether, the Chomskyan approach never grew to have many practitioners in stylistics except from a US-based branch investigating the application of generative ideas to the study of metre.

Functionalist Stylistics

After a period of stylistic concern with poetic form relatively detached from considerations about the contexts, functions and interpretational significance of the formal phenomena under investigation, stylistics took a functional turn in the late 1970s. According to Leech:

Functionalism (in the study of language) is an approach which tries to explain language not only internally, in terms of its formal

properties, but also externally, in terms of what language contributes to larger systems of which it is a part or subsystem. Whether we call these larger systems 'cultures,' 'social systems,' 'belief systems,' etc. does not concem me. What is significant is that functionalist explanations look for relations between language and what is not language, whereas formalist explanations look for relations between the elements of the linguistic text itself. (Leech, 1987, 76)

The stylistic shift in focus towards functionalism was largely due to the emergence in linguistics of different functional approaches to language, and, in particular, to the development and general popularity of Halliday's functional model of language, now known as Systemic Functional Linguistics (Halliday. 1994). At the crux of Hallidayan linguistics is an interest in language in use and a recognition of the fact that all language use takes place in context situational as well as cultural. Every linguistic choice is seen as functional and meaningful and the grammatical labelling employed for linguistic analysis is intended to reflect semantic function rather than form. With the functionalist approach also came an interest in longer stretches of text which enabled stylisticians to tum their attention more easily to longer texts such as narrative fiction and play texts.

The functionalist approach to language has had an impact in many comers of stylistics. Due to its focus on meaning-making in context, various contextually and/or ideologically oriented branches of stylistics such as feminist stylistics and critical stylistics are indebted to the functionalist approach, as is much of the work done in pragmatic stylistics, which, among other things, subscribes to the functionalist concern with language in use.

Historical Stylistics

Historical stylistics is the application of stylistic approaches, tools and methods in order to investigate diachronically changing or stable styles of particular linguistic phenomena in historical (literary) texts, a particular situation, or a particular genre, for example. It also refers to the synchronic investigation of a particular historical (literary) text from a stylistic perspective (Adamson, 1995, 2001; Busse, 2006, 2006, 2007). The stylistic framework may include any of the approaches to which stylistics has branched out. It also embraces the description of the interplay between language usage and contexts as well as its theorizing, and a focus on how a historical text means what it does. As such, historical stylistics can be seen as an "interdiscipline" (Leech, 2008, 1) between linguistic description and literary) interpretation. Historical stylistics approaches have shown that the language literature divide is a myth.

The dominating influence of the new technologies has also had an impact on historical stylistic approaches. Due to digitization, more historical

texts are electronically available now and there are new ways of engaging with because the procedures we can take on to search, browse or link texts have enormously simplified. For some, academic interest in the linguistics of older stages of English has in fact been revived, kept alive and eventually increased through the availability of corpora and through computerized texts analysis. However, number-crunching for its own sake does not constitute a complex historical stylistic framework, but the investigation of corpora may reveal phenomena which would otherwise have gone unnoticed. Yet, the potential of historical corpora for an explicit historical stylistic investigation has only been explored rather tentatively. This is despite the fact that literary texts constitute an important part of historical corpora and not simply due to the lack of spoken sources for historical periods.

Busse (2010a) introduces the term "new historical stylistics" and argues that it is time to take stock and to describe the methodological, theoretical as well as practical challenges involved in this new enterprise. She also stresses that new historical stylistics can and should consolidate the potentials for stylistic investigation of historical texts with more traditional approaches. Furthermore, by explicitly pointing to the stylistic aspects of new historical stylistics and emphasizing the stylistic notion of how a historical text means, she stresses that "new historical stylistics" contributes to issues at stake in modern historical linguistics alike.

The challenging task for a historical stylistician is that of the historical linguist in general: how do we make our interpretations valid (Fitzmaurice and Taavitsainen, 2007Taavitsainen and Fitzmaurice, 2007). The historical stylistician also has to ask in what ways it is possible for us, as modern historical Stylisticians, to reconstruct the past and establish the styles of a particular genre, a linguistic phenomenon, etc. Fitzmaurice and Taavitsainen (2007) draw our attention to the methodological flaws when investigating older stages of the English language through corpora or by means of form-to-function approaches. The focus of stylistics on an informed, systematic, retrievable and contextual analysis, which aims at describing and explaining how we come from the words on the page to its meaning, meets the need for securing the validation of data.

A new historical stylistic analysis of texts dating from older stages of the English language presupposes a comprehensive knowledge of the period, context and the language in which the text was produced. It also assumes knowledge of genre conventions, existent editions, copy texts and spelling variation, and the role of the editor as a mediator (Taavitsainen and Fitzmaurice, 2007, 21). In addition, a good starting point for a new historical Stylistic analysis of a particular linguistic phenomenon, thus, would be to use

findings of the phenomenon (or related features) in contemporary English, for example, and then investigate whether these findings can also be transferred to historical data. For example, for an investigation of speech, writing and thought presentation in a corpus of nineteenth-century narrative fiction (Busse, 2010a) it is necessary to use Modern English findings and investigations, but, at the same time, it is also necessary to put these results for Present-day English into a historical perspective of nineteenth-century discourse presentation and to account for nineteenth-century particularities. This includes a variety of contextual information which guides our reading: generic knowledge and encyclopedic background knowledge as well as knowledge of historical schemas and scripts, belief systems, which may all lead to what Toolan (2009, 7) calls a "coloring by the reader"- in the historical stylistic sense, the said modern reader reconstructing the past. But our modern assumptions about a historical text can also easily lead us astray. Language change always happens. It is visible in such grammaticalized forms as "actually," which used to be an adverbial of time before it changed into a pragmatic epistemic stance marker indicating actuality and reality (Biber et al., 1999, 870; Busse, 2010b).

In the process of interpretation of historical (literary) data the creative interaction between quantitative and qualitative investigations is crucial to the new historical stylistic approach (Busse, 2010a). This interplay must avoid circularity, a research question which only matches the data, and numbercrunching for its own sake (Taavitsainen and Fitzmaurice 2007). However, it cannot rely only on individual and purely subjective readings alone. Cooperation between intuition and corpus linguistic methods (Semino and Short, 2004; McIntyre, 2007) is paramount for a new historical stylistic approach. Corpus-based investigations aim at identifying forms and repetitive patterns, but we need to situate them within their contexts, otherwise we only establish their frequency (Taavitsainen and Fitzmaurice, 2007, 18). It takes a human analyst to make sense and interpret the discourses at stake (Toolan, 2009.16). Yet, the investigation of large amounts of data also provides us with a framework and a norm against which the results of a new historical stylistic intestination are measured to establish the discourses of a particular genre, the stylistic realization of a particular linguistic feature or creative language in general.

Establishing (historical) linguistic and stylistic norms and irregularities relates to another highly crucial concept within stylistics thote indispensable for new historical stylistics, namely the theory of foregrounding. The interplay between the analysis of a linguistic phenomenon inchi text and the various contexts in which it occurs is highly complex. It allows the analyst to establish what changes, what remains constant, or, in other words, it

allows analysts to establish the relationship between the conventional and the innovative (Taavitsainen and Fitzmaurice, 2007, 27). In order to measure and describe the levels of foregrounding (including deviation and parallelism) on a historical dimension, it is impossible to argue on a one-to-one basis that what occurs in a big historical reference corpus constitutes the ordinary or the norm. A more delicate and contextually based analysis is necessary, which deals with context(s) and envisages the notion of "emergent grammar" or emergent styles (Taavitsainen and Fitzmaurice, 2007). Styles in historical texts are not always stable in terms of form-to-function or function-to-form, but may be constantly modulated within a historical framework. What constitutes a writer's motivated choice may, in the course of time, be a norm and become highly frequent. In addition, Leech's (2008) distinction between deviation on a primary, secondary and tertiary level needs to be seen as interdependent when exploring the effects of linguistic processes in historical contexts. Language norms, discourse specific norms as well as text-internal norms play a role when evaluating stylistic change and stability within a historical dimension

in historical pragmatics (Jacobs and Jucker, 1995) and historical sociolinguistics much has been said and done to illustrate the usefulness of functionally situated approaches for broadening their scope and to illustrate what we know about the usage of English in former times. Written data has been accepted as a valuable and necessary source, and the time-span for the investigation of "spoken" language is no longer restricted to Present-day English as can be seen, for example, in the extensive body of studies on the historical development of speech acts in English (for example, Jucker and Taavitsainen, 2000, 2008; Kohnen, 2000). An investigation of the relationship between language form and meaning within a new historical stylistics framework (Busse, 2010a) uses and enhances methods and terminology from historical pragmatics and historical sociolinguistics in order to discover both change and stability. Both a pragmaphilological and a diachronic approach (Jacobs and Jucker, 1995), including a form-to-function and a function-to-form mapping, are seen in new historical stylistics as interdependent and too complex to be set apart from one another.

Multimodal stylistics

Multimodal stylistics is a fairly new branch of stylistics which aims to broaden the modes and media to which stylistic analyses can be applied. Thus, the (extended multimodal) stylistic toolkit, in addition to being useful for the analysis of the printed word, can illuminate how other semiotic modes such as typography, color, layout, visual images, etc. do also construct meaning. From this stylistic perspective, all communication and all texts are considered multimodaleven waventional literary narratives without special visual effects, since written

verbal language automatically and without exception involves both wording and typography (or graphology) as well as realization in space in terms of layout. Multimodal stylisticians furthermore broaden out the concept of, for instance, the novel to include not only the narrative of the wording and possible visual images, typography and layout but also the book cover, the paper quality and other aspects of the book's material realization. With its focus on meaningmaking as a multi-semiotic phenomenon, multimodal stylistics thus also allows for more comprehensive stylistics analyses of drama and film.

The aim of multimodal stylistics is to develop as systematic descriptive "grammars" of all semiotic modes as those already developed for the mode of wording (i.e., the lexical and grammatical aspects of verbal language). Much of the work in multimodal stylistics is based on research done more generally on multimodality by scholars like Kress and van Leeuwen (1996, 2001), O Toole (1994), O'Halloran (2004), Baldry and Thibault (2006) and Bateman (2008), who base their theoretical and methodological framework on Halliday's "social semiotics", as well as by scholars like Forceville (1996) and Currie (2004), whose work on filmic multimodality is informed by cognitive theory. Multimodal stylisticians draw on the descriptive apparatus that these pioneers of multimodal thinking have developed for modes other than the verbal. In addition to wording, the semiotic modes most prominently involved in literary meaning-making are those of typography, layout and visual images.

The analysis of typography, for instance, focuses on the meaningmaking potential of the visual side of verbal language. In this connection, the meanings created by various typographic features such as the use of italics, boldface and majuscules (i.e., capital letters) as well as of different typefaces and of lettering in different colors are considered and systematized. To this end, multimodal scholars (e.g., van Leeuwen, 2006b) have developed a system of characteristic typographic features to enable a detailed description of different typefaces and the characteristics that set them apart from other typefaces. To reflect the methodological affiliation to linguistics, these characteristic features are sometimes referred to as distinctive features (van Leeuwen, 2006, 147– 52). Another way of systematizing typographic meaning-making involves the categorization of typography according to the semiotic principles behind the meanings created by the visual side of verbal language (van Leeuwen, 2005b; developed further by Nørgaard, 2009). One example of this is the use of majuscules to create the meaning of "shouting. Here the typographic salience of majuscules may be seen as iconic in that it visually imitates the sonic salience of someone shouting. Another example is the use of a typical typewriter font such as Courier to create the meaning of "typewritten." Here the

relation between the typographic signifier and the signified may be characterized as indexical, since the typeface may be seen as a (fictive) indexical marker that the text has been produced by a typewriter. Other semiotic principles involved in typographic meaning-making are those of symbol and discursive import. Arguably, plain black typography in literary texts may be regarded as symbolic, since the relation between the visual side of the typographic signifier and that which is signified can be seen as unmotivated and arbitrary. Discursive import, on the other hand, occurs when typographic signs and their associated meanings are "imported" into a context where they did not previously belong (van Leeuwen, 2005b, 139). A good example of this is provided by Owens and Reinfurt (2005), who, albeit in different terms, discuss the discursive import of the typeface "Data 70" from the electronic processing of cheques into entirely different contexts. Based on the typography of "E13B" (known from cheques), which was designed as a typeface that could be read by machines, Data 70 is a fully fledged alphabet, which, with its associations of automated systems and computerization has been imported discursively into the context of book covers, music albums, film titles, etc.

The analysis of layout aims to systematize the meanings created by the arrangement of text and images in the spatial layout of the page. One of the things cxamined in relation to layout is information value (Kress and van Leeuwen, 1996, 186-211), that is, the question of whether special meaning is ascribed to the top and bottom, left and right, centre and margin of the page. A second issue concerns linking (van Leeuwen, 2005a, 212-47), which refers to the ways in which different layout units are linked, whether they be verbal or visual, as well as the status the different modes hold in the communication: Do visual images, for instance, illustrate the verbal text, or do they create meanings that are not expressed verbally, and, if so, to what extent? Further. compositional resources are those of framing and salience. Framing is a resource for connecting or disconnecting elements in a visual layout and is typically realized by lines, color and/or blank space. The most frequent type of framing in literary texts is probably that constituted by margins and blank space. These layout features may, for instance, play a part in the construction of the meaning of "different text," such as that of a letter, by demarcating the separation, and hence special status, of such text from the rest of the narrative. salience, on the other hand, concerns elements which stand out, for instance, in e layout of the page. In written prose, such as that of Foer's novel, Extremely

loud and Incredibly Close (2005), photographic images and other visual acts are hence salient against the background of the rest of the text, and against the background of the genre conventions of the novel altogether. At the same time, however, elements in these photographic images may be salient in

themselves such as the figure of a falling man in one of the photographs (Foer 2005, 205), who is conspicuously distinctive against a black and white background, which is respectively suggestive of one of the sides of the Twin Towers of the World Trade Center and the sky. The salience of the falling man in this image is arguably strengthened by him being the only human figure in the image and by the (for a human, unusual and fateful) process of falling or Tloating-mid-air itself, of which he is represented as the participant.

For the analysis of visual images, whether they be drawings or photographic images, Kress and van Leeuwen's Reading Images. The Grammar of Visual Design (1996) provides a fairly comprehensive grammar delineating how the visual, like the verbal, may be seen to express ideational, interpersonal and compositional (Halliday's "textual") meaning. According to Kress and van Leeuwen, visual images construct ideational meaning through the representation of participants, processes and circumstances (1996, 43–118). Interpersonal meaning is created by the positioning of the viewer and is analyzed in terms of what Kress and van Leeuwen call gaze, size of frame/social distance, perspective and visual modality (1996, 119-80). Compositional meaning, in turn, is realized through information structure, linking, framing and salience (1996, 181-229). By combin ng aspects of Kress and van Leeuwen's visual grammar with a more traditional stylistic analysis of the dialogue of McKellen's (1995) film version of Richard T. McIntyre (2008) demonstrates how the visual may be described as systematically as the verbal, how the verbal and the visual interact and, ultimately, how a multimodal approach allows the analyst to make a more comprehensive analysis of a filmed play than that captured by a stylistic analysis of the verbal text only.

From a cognitive perspective, Forceville (1999, 2002a), for instance, works on the interface of novels and their cinematic adaptations. He illustrates how multimodal formats need to be viewed as manifestations of the same mental constructs responsible for the realization of linguistic forms, the particular ones he looks into being linguistic metaphors (1999) and representations of free indirect thought (2002a). Forceville (1996) also analyses advertising discourse as it appears in print and in billboards from a cognitive multimodal perspective and concludes that verbo-pictorial metaphors are clear examples of how conceptual metaphors are indeed mental constructs capable of manifesting themselves in more than one semiotic mode.

Altogether, the multimodal take on stylistics would seem a promising approach for analysts who acknowledge that all texts, including literary ones, are multimodal, and who wish to employ and further develop tools for the

description of multimodal meaning-making which are as delicate and systematic as those traditionally employed in stylistics for analysis of verbal forms.

Narratology

Narratology, far from being a subdomain of stylistics, is a fullyfledged discipline in its own right. Narratological approaches to the study of texts have been included in the work of humanities scholars since Russian Formalism took hold of literary studies, although the beginnings of narratology are generally acknowledged to have primarily been informed by the structuralist views of the 1960s. Thus, although not to be classified as a sub-branch of stylistics, this discipline has traditionally offered plenty of working tools to stylisticians, especially to those concerned specifically with narrative fiction. Herman (2007b) broadly defines the discipline as follows:

An approach to narrative inquiry developed during the heyday of structuralism in France. Instead of working to develop interpretations of individual narratives, narratologists focused on how to describe narrative viewed as a semiotic system—that is, as a system by virtue of which people are able to produce and understand stories. (Herman, 2007b, 280)

We would be wrong, however, to assume that narratology can be conceived of as a univocal body of research. Instead, the multifarious interpretations as to how to best describe the boundaries of this scholarly enterprise are sometimes dictated by a definition of the object of study this discipline is interested in, that is, the notion of narrative itself, but such endeavor is not an easy task either: "since narratology is the science of narrative (or a theory of narrative), its very scope depends on the definition of the latter" (Prince, 2003b, 1). As is customary in the humanities and the arts, it seems more profitable to avoid a stern definition of narrative and narratology so that the various trends, subbranches and developments can be accommodated. There seems to be some consensus, though, as far as the various phases that narratological studies have lived through. These delimitations are, once more, made very broadly and with lots of scope for further fine-tuning, but most scholars (for instance, Cornils and Schernus, 2003; Darby, 2001; Kindt and Müller, 2003) appear to agree on the existence of three major stages in the development of narratology, sometimes re-distributed and amalgamated into two. Herman (2007a) opts for the latter option and distinguishes two main periods which he terms the "classical" and "postclassical" (13) approaches to the study of narrative:

I use the term classical approaches to refer to the tradition of research that, rooted in Russian Formalist literary theory, was extended by structuralist narratologists starting in the mid 1960s, and

refined and systematized up through the early 1980s by scholars such as Mieke Bal, Seymour Chatman, Wallace Martin, Gerald Prince, and others. I also include under the rubric of classical approaches work in the Anglo-American tradition of scholarship on fictional narrative; some of these scholars were influenced by and in turn influenced the Formalist-structuralist tradition. (Herman, 2007a, 12)

In this camp, we find the work of authors such as Tomashevsky, Sklovskij and Propp; Sklovskij's distinction ([1925]1990) between fabula and sjuzhet, for instance, or Propp's Morphology of the Folktale ([1928]1968) became some of the referents on which further perspectives on narrative were subsequently built. Kindt and Müller (2003) are among those scholars that prefer to divide the classical stage into two separate phases:

The first phase, beginning in the mid-nineteenth century in Europe and the USA [...] took its material from three main sources: the remnants of normative rhetoric and poetics, the practical knowledge of novelists and the observations of literary critics [...] It was only in its second phase that 'narratology' became a distinct subdiscipline of textual studies, after the term first used in 1969 by Tzvetan Todorov in his Grammaire du Décaméron found wide international acceptance. Todorov's account of the airns and themes of narratology was heavily influenced by Russian and Czech Formalism and structural linguistics [...]. Subsequently, however, the 'high structuralism' of these generative grammarians achieved far less international currency than the 'low structuralism of Gérard Genette. (Kindt and Müller, 2003, v-vi)

Whether the first stage is divided into two sub-groups or considered as a unified whole does not detract from the fact that the scope of the so-called classical narratology extends well into the 1980s, when its focus on traditional structuralist methodologies and concerns starts to wane under the influence of new trends emerging from other humanities and social science disciplines such as anthropology, psychoanalysis, cognitive linguistics and cognitive psychology. Irrespective of whether the initial stages are amalgamated or not, most scholars claim that the revival of the discipline took place in the 1990s when the third period of narratology is said to have started:

Narratology, it is argued, is now more alive than ever before, having undergone something of a renaissance since the 1990s after a period of stagnation and crisis during which its demise was repeatedly proclaimed. The 1990s produced such a proliferation of heterogeneous approaches that narratologists such as David

Herman find it more appropriate to speak of 'narratologies' in the plural. (Cornils and Schernus, 2003, 138)

Whether scholars refer to this new era as the third stage or whether they adopt the "postclassical" term coined by Herman, most of them convene in acknowledging that the new phase is fraught with influences and interdisciplinary links with cognate areas. This has permitted new forms of inquiry into the nature of narrative forms to be incorporated into the already existing frameworks. Of special interest to stylisticians, in particular to those exploring the cognitive dimensions of literary processing, is the new application of cognitive theories to the study of narratives:

Study of the cognitive dimensions of stories and storytelling has become an important subdomain within the field of narrative analysis. Concerned both with how people understand narratives and with narrative itself as a mode of understanding, cognitive approaches have been brought to bear on stories in a variety of media [...]. Equally various are the disciplinary traditions from which cognitive approaches borrow descriptive and explanatory tools. Source disciplines include cognitive linguistics; pragmatics; discourse analysis; narratology; communication theory, anthropology; stylistics, cognitive, evolutionary, and social psychology; rhetoric; computer science; literary theory, and philosophy. (Herman, 2006a, 452)

This acceptance of influences from such a wide spectrum of disciplines is positively conducive to cross-fertilization and general enriching of both narratology and stylistics as the latest work on narrative carried out by stylisticians, as much as the publications by narratologists on stylistic issues, proves (Gibbons, 2010; Herman, 2005b, 20060; McHale, 2007; Page, 2007b).

Pedagogical Stylistics

Pedagogical stylistics has at least two facets. One embraces the pedagogical usefulness and potential of stylistics for teaching (the language of) literature. The other includes the role of stylistics in Ll and L2 pedagogies, that is, the teaching of (English) language through literature.

The complex developments within literary criticism over the last 20 years have affected the interplay between literature, language and language education in the language classroom. Reflections on the author's intentions, textuality , measurements of responses of readers as well as discussions about ay texts to be included in the canon formation influence approaches to the literature is taught and the view that in learning a foreign language its

literature should be read. In turn, because of the role of English in the world and the status of World Englishes, there has been a growing awareness of the need for highly profiled environments in which English is taught as a Foreign Language (EFL), for example. Research in second and foreign language studies and in pedagogical stylistics has increased immensely. The pedagogical aim of stylistics in teaching (literary) language and how this language functions is based on how we as readers native and non-native--come from the word on the page to its meanings. Stylistics as a method can help explain how a particular use of language works within a text for both the native and the nonnative speaker and how texts are interpreted and understood by the reader.

According to Carter and Stockwell (2008, 249), pedagogical stylistics developed in the 1970s and became very practical as one way of evading the attacks leveled at stylistics. To a large extent, language was then taught also using literary texts which represented a rather attractive option for some L2 learners as contemporary literature was often chosen. Some language teachers were of the opinion that works that prominently made a foregrounded use of linguistic tools were especially suited to show language in action. For others the former would only be appropriate for advanced learners. But also lately a more concerned view has been voiced which stresses that stylistics has contributed to methodologies in the teaching of literature and that, therefore, Ll and L2 methods are embedded in stylistics.

Textual transformations, which highlight the tylistic notion of meaning as choice and the theory of foregrounding, are crucial. For example, cloze procedures, where students are asked to insert a missing word in a verse line to later evaluate the author's choice, are commonly used. Alternatively, comparative analysis and re-writing texts from different perspectives are also widespread in L2 environments and in stylistics. These demand a different set of linguistic choices and knowledge of the linguistic features of a variety of genres and text types in order to transform, for example, a poem into a blog. Readers are guided through an active process of reading and learners become explorers of linguistic, textual and cultural transformations in which the text is seen as a corrective of complex linguistic structures. These observed linguistic choices construe meaning and they make the interpretation by the reader as reader or the reader as learner retrievable.

The notions of creativity and literariness are explained against the background of various norms and conventions. Furthermore, Carter (2004) stresses the fact that everyday language is also creative and that this has an impact on the understanding of literature: "the teaching of literature in a variety of cultural contexts may be better informed by understandings of the

pervasively creative character of everyday language [...]. The idea that creativity exists in a remote world of literary genius can be demotivating to the apprentice student of literature, especially in contexts where a second or foreign language literature is taught" (2004, 213). Stylistics has contributed to the methodologies of teaching literature and this in turn has affected developments in L1 and L2 pedagogy.

As Carter (2007a) explains, the new developments within stylistics present not only new challenges but also new exciting opportunities for the teaching of stylistics:

The intellectual excitement of cognitive poetics (with its re-tuning of the significance of language processing), the awesome power of corpus linguistics, the depth and richness of studies in narrative analysis, the ever new angles on the literariness of language all now take place routinely alongside developing opportunities for web-based teaching and learning, the pedagogic possibilities afforded by hypertexts, more refined rhetorical-analytical tools, enhanced paradigms for greater empirical investigation, more and more successful integration of quantitative and qualitative methodologies. (Carter, 2007a, viii)

The new branches within and influences on stylistics certainly need to be addressed from a pedagogical perspective. Issues that are now investigated as a result of so many new perspectives are, for exaniple, how learners appropriate themselves of other discourses, how stylistics helps show the contextual nature of all texts and a need for close investigation and how a stylistic analysis of a literary text can be used to promote identity and feeling (Watson and Zynger, 2007). Also, Mick Short's (2007) web-based learning environment (http://www.lancs.ac.uk/fass/projects/stylistics/), which introduces students to stylistics, illustrates not only the versatile didactic and pedagogical potential of stylistics, but also discusses the challenges of introducing students to stylistics in a web-based format (Short, Busse, Plummer, 2006).

Pragmatic Stylstics

Pragmatic stylistic approaches combine approaches from pragmatics and stylistics to answer questions about how (literary) language is used in context and how it contributes to the characterization of the protagonists in a literary piece of art or how power structures are created and so on. Pragmastylistic investigations have influenced general pragmatic approaches, methods and theories on both a synchronic and a diachronic dimension, too. Especially in historical pragmatic investigations, which include a pragma-philological and a diachronic pragmatic analysis, literary texts have been a source frequently

drawn on, because there is no spoken data available for historical periods, and play texts constitute an important source to explore "the spoken," although admittedly, this is the constructed" spoken language. Other points of intersection between pragmatics and stylistics include the focus on context and on the effects of the interactional strategies used in context. Furthermore, pragmatic stylistics has stressed a comprehensive holistic approach to conversational interaction and includes the complex interplay between norms and deviations as well as forms and meanings. On the assumption that norms and conventions from natural language usage are built upon in literary conversation, those pragma-stylistic findings have something to say about linguistic realizations of politeness strategies in general. The same holds true for the realization of speech acts or discourse markers in literary texts. Pragmatic stylistic approaches and multimodal stylistics have also drawn attention to the need for including other semiotic modes in order to account for the interplay between language and the visual, etc. in films, for example (Busse, 2006b; McIntyre, 2008). More recent approaches combine pragma-stylistic investigations with corpus stylistic approaches and relate the identification of linguistic patterns to interactive features. In addition, within a broad and comprehensive framework that resulted also from the pragmatic analysis of historical texts, the pragmastylistic focus on language as exchange and the contextual features of language also embraces the analysis of fictional narrative passages, e.g., the relationship between narrative passages and discourse presentation or a combination of pragma-stylistic and cognitive stylistic considerations.

Underlying a pragma-stylistic investigation of dialogue are some central questions of stylistic analysis: Why and how docs a play text/dialogue mean what it does? What is the specific style of a conversational exchange? How can it be analyzed? What are the effects of the linguistic choices made? What do these choices say about the characters/speakers' interpersonal relations and their inherent power structures? How is humor created? Why do we perceive interactional exchange as, for example, impolite? The systematic, rigorous application of pragma-linguistic concepts and tools helps answer those questions. Short (1989, 1996, 2007b) illustrates the systematic investigation of play texts, their specific style(s) and the relationship of play texts with performance. Short (1989) develops a stylistic tool kit for the exploration of conversational exchange and draws on areas of pragmatics and discourse analysis (areas which do not play a major role in the analysis of poetry, for example).

Major foci of the pragma-stylistic tool kit are on contextual features of language use and on seeing conversation as exchange. The notion of context mav of course include various aspects: for example, what Schiffrin (1987) has described as the physical, personal and cognitive context, or what we would

generally understand as social, cultural, linguistic, authorial or editorial contexts of production and reception.

Norms and conventions of authentic everyday communication can be seen as a base for the interpretation of fictional characters' use of speech. Otherwise, it would not be possible to detect foregrounded use of politeness markers, irony, over-decorous greetings or comedy. Dialogue in drama or in passages of speech presentation in narrative fiction is clearly "constructed" (or purpose-built) dialogue because the author has been in control and the mediator of it. Nevertheless, pragmatic findings can be applied to its analysis because the principles of social interaction are exploited. "Dramatic action, as Herman (1995, 6 points out, becomes meaningful in relation to the 'authenticating conventions' that are drawn from the wider world of affairs in which the dramatic activity is embedded" (cited in Simpson, 1998, 41). This process of identification is of course rendered more complex in the case of historical stylistic investigation of e.g., play texts from older stages of the English language.

The question of how a reader perceives conversational exchange and interprets it also leads us to an illustration of the discourse architecture of prototypical dramatic texts. According to Short (1996), these comprise two discourse levels. One relates to the discourse level between writer and reader (or audience in the case of a performance), the other is embedded and relates to the discourses exchanged between character and character.

Addresser 1 Message Addressee 1

(Playwright) (Reader/Audience)

Addresser 2 Message Addressee 2

(Character) (Character)

An example from Shakespeare's Hamlet will exemplify this:

King. But now, my cousin Hamlet, and my son,

Ham. [Aside] A little more than kin and less than kind

King. How is it that the clouds still hang on you?

Ham. Not so, my lord, I am too much in the sun. (Hamlet 1.2.64–7)

The inherent nature of conversation that of exchange can be seen here, because we find four turns, one of which, as the stage direction from the edition taken the Riverside Shakespeare (Evans, 1997-informs us, is explicitly uttered with the writer-audience level in mind, as it is an "aside, the other three are relevant to the character-character interaction, but are also addressed to the reader or the audience during a performance. The interactive features of these utterances are also expressed by means of the address forms

exchanged. However, the lexical choices are different: King Claudius chooses a personal name, "Hamlet," and two kinship terms, "cousin" and "my son": Hamlet only resorts to the conventional title "my lord" (Busse, 2006b). The choice of address terms construes, at least superficially, their power structures, Even if we did not know the social relations between Hamlet and Claudius. their choice of address forms would allow us to infer that, conventionally speaking, Claudius's position is higher up the social ladder than that of Hamlet. because King Claudius is allowed to address Hamlet by his personal name, Yet, the frequency by which Claudius uses these forms of address also illustrates how much he tries to gain favor with Hamlet. The King alludes to both of Hamlet's roles within his nuclear family: that of being his nephew and his son, a kinship relation which results from King Claudius's marriage to Gertrude, Hamlet's mother. Hamlet's reply is cautious and reserved. He uses a very conventional and frequently used form of address, "my lord," which denies any intimate and personal kinship relations. With his use of positive politeness strategies, on the one hand, and his severe moral criticism of their marriage, on the other, Hamlet also tries to redress his sadness about his father's death/murder. Hamlet's homophonous pun on "son" and "sun," which stresses that he is too much in the light of the present King, with "sun" being the royal emblem, is an indicator of his realization of evil. This last point draws our attention to the need for including in the interpretation process contextual information of Early Modern England as weli as of the linguistic context. At the same time, it would have been ideal for the historically situated interpretation of those lines to include editorial considerations. But for ease of understanding, a modern spelling edition is drawn on.

This example also shows another fundamental question to be addressed in the pragma-stylistic analysis of play texts: that is the relationship between the play as a text and as performance. A pragma-stylistic point of view stresses that a sensitive analysis of drama can be achieved through the analysis of the text and that this stylistic analysis provides the analyst with a framework as to how a text should be performed. Productions of plays are then seen as variations of the same interpretation of a play, but not as new interpretations (each performance of a production is then a set of different instantiations) (Short, 1998, 8). In the example above, the stage directions inform us that the utterance "A little more than kin and less than kind" is addressed to the audience, who is in the know about Hamlet's real state of mind. Claudius also criticizes Hamlet for still mourning his father's death, which gives an additional clue as to Hamlet's visual expressions during the performance.

Next to the classic stylistic tool kit of investigating graphological information, sound structure, grammatical structure or lexical patterning,

pragmatic models like speech act theory, Grice's (1975) "Cooperative Principle," politeness, implicatures or turn-taking management are among the pragma-stylistic approaches frequently applied to the language used in play texts. The individual or multiple application of these discourse areas addresses such questions as how conversation functions as exchange and how it reveals (power, social or interpersonal relations among participants. Background knowledge of the world, which is also frequently arranged in schemata, plays as important a role for the pragma-stylistic analysis of play texts as does knowledge of sociolinguistic conventions or of the various contexts--social, cultural, political, production, linguistic, editorial in which the play (text) is set.

Due to space restrictions, it is impossible to elaborate on all pragmatic concepts that lend themselves fruitfully to a pragma-stylistic investigation. Therefore, only some of the major (and most frequently applied concepts) will be explained, although this does, by no means, entail that what linguists call the establishment of "phatic communion" or "adjacency pairs" is less important for the identification of fictional dialogue as exchange.

The "turn" is one of the central concepts of interaction and conversation analysts of the ethnomethodological school have illustrated that, unlike our expectations, in ordinary conversation, turn-management is systematic (Sacks et al., 1978) and rule-governed. The order of speech in drama suggests itself to an analysis of the turn-taking management. Important questions to be asked are then, for example, who speaks to whom or who interrupts and who has the longest/shortest turns (Short, 2007c). In the example mentioned above, King Claudius switches topics and attempts to initiate a conversational exchange with Hamlet and to gain favor with him.

Another important concept to be drawn on for a pragma-stylistic analysis of a play text is that of Grice's (1975) cooperative principle, which also explains why, as readers, we are able to draw inferences. Conversation is assumed to be goal-directed. These goals are usually fulfilled because we cooperate in conversation and follow the maxims of quantity, quality, manner and relation. Failing to observe one of those maxims (violating or flouting it, for example) is due to the fact that speakers often say something indirectly. In order to understand the irony of Hamlet's reply "Im too much in the sun" it is useful to explain how the maxim of quality is flouted on the discourse level of writer and audience and violated on the character-character level, when seen from Claudius's perspective. Claudius must perceive Hamlet's reply "I'm too much in the sun" as uncooperative (violating the maxims of quality and quantity as well as relevance), because Hamlet does not initially confirm that he is mourning his Taher's death. On the contrary, his reference to the "sun" superficially refers to

him as being allegedly in good spirits. But, as mentioned, it also ambiguously criticizes Claudius's usurpation of the throne and of fatherhood, which is indicated by the noun "sun." The audience, however, knows that the Ghost of Hamlet's father has shown Hamlet Claudius's murder of Hamlet's father. Hence, the audience will understand the pun on "sun" and "son."

The stylistic application of Grice's (1975) cooperative principle frequently goes hand in hand with the identification of politeness strategies. Among the frequently used models are Brown and Levinson's (1987) work on politeness, Claudius's choice of vocative forms could be seen as address strategies that are directed both at Hamlet's positive and negative face, while Hamlet's choice of "my lord" is rather unmarked because this address form is one of the most frequently used forms in the Early Modern English period (Busse, 2006b), which also undergoes a process of semantic generalization. As such, it is not addressed at Claudius's social position or aims at establishing a bond between them.

The identification of politeness strategies also involves the ability to relate linguistic realizations of utterances to their illocutionary forces. In other words, it involves knowledge of speech acts (speech act theory). The pragmastylistic analysis of speech acts and determining their illocutionary force through the way they are linguistically realized is not an easy task, especially if noncontemporary drama is observed. Politeness phenomena in historical texts cannot be compared on a one-to-one basis to Present-day realizations of speech acts because, for example, in Early Modern English requests there was less hedging and less indirectness (Culpeper and Archer, 2008). Furthermore, felicity conditions for the realization of a speech act need to be carefully considered. In the example from Hamlet, it is essential to realize the homophony between "sun" and "son," and at the same time, to know about the meaning of "sun" as an indicator of royal status. Otherwise, Hamlet's utterance cannot be identified as an insult on the writer-audience discourse level, and it cannot be understood why, on the character-character level, initially at least. Hamlet's utterance is ambiguous between a compliment and an insult.

Due to the attention paid mainly to poetry by early stylisticians, stylistic investigation of fictional literary text types containing dialogue and (inter-) action first emerges as late as the 1980s. Pragmatic approaches were certainly around at the end of the 1960s, but the tardiness in applying them within a stylistic framework may be a result of the character of spoken conversation, which has for a long time been seen as debased when compared with written expository or literary texts. Also, stylistics had to get used to investigating pieces of texts larger than poems. To date, the stylistic investigation of literary play texts or (constructed, fictional) dialogue has not as frequently been pursued

as that of narrative fiction or poetry. Yet, especially since the 1990s, there has been a more extensive pragma-stylistic investigation of play texts and an extensive broadening of the pragma-stylistic tool kit, which also interplays with corpus, cognitive and multimodal stylistic approaches.

Reader Response Criticism

Reader response criticism is a label that encompasses a series of critical approaches to literary studies popular since the 1960s and 1970s. Reader response criticism does not, in fact, refer to a unified body of work but embodies a series of theoretical strands all linked by their concern with the role of the reader in literary interpretation. As Wales summarizes:

There was the affective stylistics of Fish (1970f) in the United States and of Riffaterre (1959f) in France; the structuralist poetics of Culler (1975) with his literary competence; and the semics of Eaton (1966f). [...] Reader response criticism, like post-structuralism, tried to move away from the text as critical focus, and even more so from the intentions of) the author. (Wales, 2001,331)

Most of these movements strove to separate themselves from the previous textcentred concerns of New Criticism and formalism. As a result, they all highlight, albeit in different ways, the central role played by the reader viewed as active participant rather than as simple "recipient" of the product that is the literary text. For instance, Fish's famous "affective stylistics" (1980) primarily considers the emotional response of readers, but also bears in mind the psychological processes involved in text processing. The strand known as Reception Theory, on the other hand, is associated with the work of the Constance School in Germany, more precisely with Jauss, Iser and certain other German scholars (Schneider, 2005c, 492). Whereas the former's preoccupations lie with issues of literary history, the latter's work boosted a new conceptualization of readers as active participants. Iser (1978), in turn, not only looks at meaning construction by assessing readers' mental imput but also avoids neglecting the way the text itself predetermines how such meaning is actually constructed. Other perspectives normally considered under the label reader response are those proposed by Culler (1981), Eco (1979) and Holland (1968, 1975).

The various taxonomies of reader "types" suggested by the different strands are a further commonality between these branches. The similarities end here, though, as the various scholars are prompt to put forward their own labels which, nonetheless and more often than not, appear to refer to similar concepts. Thus, whereas most authors accept the concept of "ideal reader (iser, 1978) as the entity that most closely resembles what authors consider the perfect receptacle for their particular works, Riffaterre (1959) opts for the "super

reader," Fish (1970) proposes the "informed reader" and Eco (1979) suggests the "model reader." More recent narrative theory, however, appears to have abandoned a classification of "types of reader" in favor of what Schneider calls "reader constructs" (2005a, 482). Most of the theoretical considerations assessing the particular role of readers are, in fact, conceptualizing such entity as an abstract construct rather than as real individuals facing texts. The original approaches to the study of reader responses, thus, primarily remained theoretical assessments and it is only the new perspectives afforded by empirical studies of literature that have tried to redress this fact. The work done on issues of foregrounding has been particularly successful at reinterpreting what role the reader is actually endowed with:

van Peer (1986) found that the attention of readers is indeed attracted by the deviations and parallelisms contained in a text, that readers find text passages containing a high concentration of such devices more important and more worthy of discussion. In replicating this investigation, Miali & Kuiken (1994a), Hakemulder (2004), and Sopcak (2004) also found evidence for the affective value foregrounding passages have for readers: they are read significantly more slowly and enhance aesthetic appreciation, they influence readers' perceptions of the world, and they are evaluated more highly on a second reading In all these studies, moreover, readers' personalities played only a marginal role, confirming van Peer's earlier findings: foregrounding devices operate partly independently of reader

apparently characteristics. (van Peer et al., 2007,7)

Thanks to the new empirical perspective taken on by van Peer and his collaborators (van Peer, 1986; van Peer et al., 2007), a new lease of life has been iniected into studies of reader response. The determining role of textual devices described in empirical assessments of literature is clearly reminiscent of Iser's preoccupation with the way the text predetermines the reader's understanding, but this new stance is characterized by a far more structured and rigorous methodology. Crucially, this new take on the reader's role also allows us to investigate the emotional involvement created by the presence of certain foregrounding markers, thus echoing the claims put forward by Fish in relation to his affective stylistics. Finally, despite van Peer et al.'s seeming dismissal of readers' personalities as non-essential components in the reading experience, work undertaken by stylisticians in conjunction with psychologists and psychologists (Emmott et al., 2007) has actually underscored the readers' individual psychological characteristics play in the processing of texts.

CHAPTER TWO

Aliteration

Alliteration is a stylistic device consisting of the repetition of the same consonant sound in nearby words. To qualify as alliteration, in the strict sense of the word, the consonant sounds must occur in word-initial position. Alliteration has a cohesive effect, since identical sounds tend to tie words together if they occur in close vicinity. It can be employed for emphasis and mnemonic effects and is frequently used as a means of foregrounding in poetry, advertising, newspaper headlines, political slogans, etc. Examples of alliteration thus occur in the first line of Keats's poem "To Autumn" ([1820] 1983): "season of mists and mellow fruitfulness!," as well as in the title of Gurinder Chada's film Bend it Like Beckham (2002).

Alliteration is sometimes used in a broader sense to refer also to what is otherwise known respectively as assonance and consonance. Assonance is the repetition of vowel sounds in nearby words, usually in stressed syllables. Consonance, on the other hand, is used by some to refer to the repetition of consonant sounds in nearby words irrespective of their position, while others employ the term to refer to the repetition of final consonant sounds only.

Assonance

Assonance is a stylistic device consisting of the repetition of the same vowel sound in nearby words, usually in stressed syllables. Assonance has a cohesive effect, since identical sounds tend to tie words together if they occur in a sequence of nearby words. Like alliteration and consonance, assonance can be employed for emphasis and mnemonic effects and is frequently used as a means of foregrounding in e.g., poetry, advertising, newspaper headlines and political slogans, as in "I like Ike," the slogan for Dwight (lke) Eisenhower in the American presidential campaign of 1952.

Coherence

While cohesion refers to the linking of sentences into text, coherence coherence the appropriateness of a given text (and the various choices involved

is composition) in its communicative context. Some choices may thus be ore appropriate, and hence more coherent, than others in the situational and cultural contexts in which the text occurs. In a legal text, for example, expected.

choices would be linguistic markers typical of formal written language whereas the use of informal conversational wording and structures would be considered inappropriate and consequently incoherent in a legal context.

Cohesion

Cohesion refers to the way sentences combine into text by means of textinternal ties. The concept of cohesion (as well as that of coherence) is largely owed to Halliday and Hasan, whose seminal study of the phenomenon. Cohesion in English (1976), has been greatly influential in stylistics as well as in linguistics more generally. Halliday and Hasan (1976) identify five main types of cohesive tie in English-four grammatical and one lexical Grammatical ties are realized by conjunction, ellipsis, substitution and reference. Conjunction and conjunctive expressions such as "but," "then,"

"accordingly" and "nevertheless" work by explicitly signaling how different o parts of the text relate to each other. Secondly, ellipsis links different parts of a

text by means of omission as in "I could offer her a lift. But no, I won't " where a cohesive tie exists between the empty slot following "I won't" and "offer her a lift" in the previous sentence. Closely related to the cohesive resource of ellipsis is that of substitution. In an alternative version of the sentence above, a placeholder may thus occupy the empty slot, thereby linking the two sentences: "I could offer her a lift. But no, I won't do so." Reference, in turn, occurs when linguistic items "make reference to something else for their interpretation" (Halliday and Hasan, 1976, 31). Prevalent examples of reference are pronouns ("you," "she," "his"), demonstratives ("this," "that," "those") and articles ("a," "the"). The element referred to i.e., the referent) may precede the referring item (anaphora) or follow it (cataphora). Lexical cohesion falls into three subcategories: repetition, synonymy and collocation Repetition and synonymy are both examples of reiteration. Where the former occurs when a lexical item is either repeated in its identical or near identical form (Prun"/"run," "run"/"running"), in the latter case a lexical item reiterated by some kind of synonym ("car"/"automobile"), including hyponymy, that is from specific to general ("poodle"/"dog"), metonymy of part-whole ("finger"/"hand") and antonymy or opposition (''happy"/"sad"). Collocation, or expectancy relations, is cohesion created by means of words that have a tendency to co-occur. Certain words tend to come in pairs i "make" and "decision," "freezing" and "cold," "waste" and "time," but the may also be a broader sense in which words are likely to occur together thereby forming lexical sets. When we read about "computer," "screen" "keyboard," for instance, we may expect to also encounter "mouse pad" "desktop," since these lexical items tend to collocate and thereby tie the text.

Hag out." "put the flag back," "caddie" and "balls" in combination gradually reveal that the people Benjy is watching are playing golf, thereby narrowing down the meaning potential of the outdoor setting to that of a golf course. In addition to the lexical and grammatical means of cohesion listed above. elements such as alliteration assonance consonance and different kinds of rhyme likewise display the ability to establish cohesive links between different elements in a text. The cohesive effect of alliteration, assonance, consonance and rhyme is arguably a visual as well as an auditive phenomenon. When we read the text aloud, we hear the patterns of similar sounds and hence sense the cohesive ties which are established sonically, yet we may also spot the similarities as visual similarities when we read the text silently.

Collocation

Collocation embraces the Firthian (1957) notion that a word can be described by the company it keeps: "Meaning by collocation is an abstraction at the syntagmatic level and is not directly concerned with the conceptual or idea approach to the meaning of words. One of the meanings of night is its collocability with dark" (Firth, 1957, 196), similarly the adverb "ago" frequently collocates with "a week" or "a month" to form the expression "a week/month ago." In order for an expression to be a collocation, the word needs to occur together with others in more than regular frequency.

Sinclair (2004, 14) defines a collocation as "the co-occurrence of words with no more than four intervening words." Hoey (2005, 5) adds an additional focus on psychology to his definition of collocation: "A psychological association between words (rather than lemmas) up to four words and evidenced by their occurrence together in corpora more often than is explicable in terms of random distribution." Toolan (2009, 19), in turn, restresses the textual basis of collocation. His definition follows from his corpusstylistic analysis of narrative progression in twentieth-century short stories:

Collocation is the far-greater-than-chance tendency of particular words to co-occur (adjacently or within a few words of one another). These co-occurrence tendencies have text-constructional and semantic implications. Proficient native language users are equipped, by their communicational experience, with the commoner collocational tendencies and implications. (Toolan, 2009, 19)

Frequently, collocation is mentioned hand in hand with the term "colligation, which describes the ways grammatical words appear with particular lexi items "to cover relationships between grammatical categories and partic lexical words" (Butler, 2004, 154). For example, the word "ago" frequent functions grammatically as an adverb of time in initial position.

Sinclair (2004), Hoey (2005) and Toolan (2009) all relate ration(s) (and colligation) to "naturalness" and to the native speaker's competence to produce them, which makes them appear fluent and proficient. The concept of "lexical bundles" developed by Biber et al. (1999) and Biber et o 02004) also has to be mentioned here because it refers to extended collocations of three-, four-, five- or even six-word sequences that occur in a statistically marked frequency. Biber et al. (1999) used a large corpus as well as computer analysis for identification of those multi-word units. This is an area of corpus linguistic research which also focuses on so-called n-grams with "n" standing for the number of words that can be seen to occur in more than just random frequency. It also correlates to what Sinclair (1991, 2004) has called the 'idiom principle," which means that what we hear, read or use are often fabricated multi-word phrases. This principle interacts with the "open choice principle," which relates to the ability of a speaker to constantly fill a slot within a sentence, a process which is only governed by grammatical rules. From both a syntagmatic and paradigmatic point of view it should also be stressed that one principle to which collocations can be related is that of compositionality (Kortmann, 2005, 193). Often the meaning of expressions and their syntactic relations can be inferred from their parts and refers to what is named after the German philosopher and mathematician Gottlob Frege, that is, the "Fregean Principle" (Kortmann, 2005,5). However, idioms cannot be reconstructed from the parts of words they consist of.

Collocations can be identified for content words (like adjectives, adverbs and nouns) and function words (like articles, pronouns and so on) alike, even though the latter are usually more frequent than the former. Frequent words also have their own repetitive patterns. When we investigate collocational patterns of content words, there will be function words in their surrounding, which makes collocations one of the linguistic proofs of why the lexis-grammar interface cannot be separated.

Mike Scott's (2004) key words tool of the software WordSmith compares the keywords in one text with those in another, larger reference corpus. In other words, since we know that frequency is a relative concept, comparison and reference corpora and statistical procedures are needed in order to be able to establish collocation patterns. Frequencies of words have to be seen in relation to words in their context.

one tool for further analysis of collocations used in corpus linguistics is to build concordance lines based on large amounts of data, i.e., a corpus. A concordance line brings together brings together and displays instances of use of a particular word from the widely disparate contexts in which it occurs. A collocation then

Critical Discourse Analysis

Critical discourse analysis (CDA) investigates the relations between language and society. It assumes that language plays a crucial role in creating, maintaining and legitimating inequality, injustice and oppression in society. The most important aims of CDA are to raise awareness of the power of linguistic constructivism (language construing reality) and of its impact on society and to trigger change.

Although critical discourse analysis does not follow a particular School of thought (Wodak and Meyer, 2001), Halliday's (1994, 2004) systemic functional grammar is functional grammar is most frequently drawn on. This is visible in

on denotative meaning in his model of the sign, theorists after himnotably among them, the French semiotician Roland Barthes (1915-80) have argued for a more nuanced view of the signified which acknowledges that the signified concept of a given signifier comprises denotative as well as connotative meaning. A popular example of this is the signifier "home" which denotes the meaning of "a place where one lives," yet at the same time it may also connote concepts such as "privacy," "security," "comfort" and "family."

The denotation and connotation of a word or expression may vary according to the context in which it occurs. In one context, the lexical item "bank" will be employed to denote "a financial institution where people can borrow or keep money," while in another, the denotation of the same word will be "a raised area of land along the side of a river." Similarly, the word "feminist" may carry positivel connotations such as "independent," "progressive," "aware of gender issues" for some people, but negative connotations such as "aggressive," "militant" and "man-hater") for others. It is generally believed that the denotation of a given signifier is more fixed and less open to interpretation than its connotative meaning. It is important to note, however, that the connotations of a given word are not completely subjective, and that even the denotative meaning of words, as well as

of other semiotic signs, is socio-culturally specific and never completely neutral. nanlocotel

In semantics and philosophy, "reference," "referent" and "referential meaning" are employed to describe) what is termed denotational meaning above, i.e., the use of language to represent the extra-linguistic world. The referent of the lexical item book" is thus a real, physical, book, which is consequently referred to by linguistic means. In these fields, reference" contrasts with "sense, which concerns the linguistic meaning of a lexical item and its relations to other linguistic items in the system to which it belongs.

Discourse Presentation

Discourse presentation is a ubiquitously used term and seems to refer to a variety of linguistic and socio-cultural phenomena. For some, it may allude to the presentation of speech or conversation. For others, it may refer to the linguistic approach of discourse analysis or discourses in general and their contexts. And still others who use the term "discourse presentation" do not clearly differentiate between speech or thought presentation categories. The term discourse presentation is used here as a cover term to refer to the strategies reporters or narrators use to present other people's speech, thought or writing. It is also employed to describe the possibilities of labeling those modes and their functions. The term "presentation" instead of "representation" is the preferred choice because it is less ideologically loaded. Toolan remarks that:

[y]ounger readers, fully of the digital age, may be puzzled that these opening remarks about the power to record and represent speech are not focused on audiotape recording and digitised sound files, now very widespread means of representing speech. But this [...] is about an affordance of the much earlier and culture-changing technological breakthrough the development of writing, and literary authors' rich repertoires of means for presenting characters' words on the page. (Toolan, 2006, 698)

This section will use Leech and Short's (1981) and Semino and Short's (2004) model of discourse presentation. The earlier model by Leech and Short (1981) is mainly based on the analysis of literature. Yet, it should be stressed that the model and its underlying principles can be transferred to other varieties and genres (Toolan, 2006c). In fact, it has been applied to journalism, parliamentary debates and magazines.

Because the presentation of other people's speech seems to be most pervasive (and easiest to understand), it makes sense to begin with some examples from this mode of discourse presentation to illustrate in general what it entails if other people's speech is presented. It is possit le for a reporter to present the words of a speaker in a direct way, by means of the so-called direct speech (DS). This can be done with the help of a reporting clause (1) "he said," which is sometimes also called the inquit clause or "franting clause." It can also be done without it (2), which is labelled free direct spex ch (FDS):

(1) "she is a lovely person," he said.

(2) "she is a lovely person."

In example (1), the distinction is between the reporting clause, which identifies the speaker of the reported clause and the verb of communication, and the reported clause, which presents the speaker's message. The stretch of direct speech is marked by inverted commas/speech marks. The reporting clause helps the reader to understand that this stretch is reported verbatim. In addition, all deictic markers (pronouns, tense, time and place expressions (Toolan, 2006 699)) construe the speakers' position and their deictic centre. Often the reporting clause also contains paralinguistic information or reference to the speaker's mood or attitude. These additional linguistic markers function as evaluation or indexical cues (Toolan, 2006c, 699). We may also report the words of others in a more indirect way (3).

(3) He said that she was a lovely person. Although the reporting clause in (3) is still used ("he said') the inverted commas have disappeared and there is also a tense shift from "is" as in (1) to "was " The

stretch "she was a lovely person" is indirect speech (IS). The reported clause is usually introduced by a complementizer or subordinator, among which "that" is most common in Modern English (others are "if" or "whether"). While in direct speech (DS) all deictic elements mark the position of the speaker, in indirect speech (IS), the narrator controls them both with reference to space and time. Spatial and temporal markers like "here," "today," "this" and "now" will be replaced by "there," "on that same day" and so on. Here the foreign language learner of English might be reminded of the complicated rules that had to be studied to understand "reported speech." In IS, the speaker's real words used in the antecedent discourse are often not reported on a one-to-one basis. But the "content, gist or illocution of the actual or implied DS utterance" (Toolan, 2006c, 699) is given. It is especially in narrative fiction that a variation of source is assumed because a variety of possible antecedent versions of DS are possible. In "real" or "non-fictional" discourses a need for what is called faithfulness is more important, because even in IS, it is assumed that what is presented is very similar to the antecedent wording. For example, as readers of newspaper reports about political speeches or decisions made by politicians we assume that the words reported are the exact words used by the speaker.

Toolan (2006c, 699) points out that "When contemplating Direct Speech [...] at least two binary distinctions need to be considered: is the text in which it appears literature or non-literary, and is it hypothetical or actual." Despite Toolan's distinction between literary and non-literary forms, it should be stressed that DS is still considered the most faithful category of speech presentation for fiction and non-fiction alike.

All the examples presented so far illustrate the presence of a "teller" who has purposely decided to report the speech of somebody else. Direct speech and indirect speech differ in terms of their realization, propositional content and function, and such differences are carried through for the presentation of thought and writing alike. Leech and Short (1981) claim that these three varieties of reported discourse work on parallel scales although their functions in language are different. So although thought presentation is utterly different in function (especially in terms of faithfulness claims, in terms of effects, in terms of the construal of thought in linguistic form and in terms of the possibility of summarizing thought), Short (2007a) stresses that it makes sense to display the three presentation modes on parallel scales.

Semino and Short (2004) extend the discourse presentation model from that in Leech and Short (1981/2007) by not only making a distinction between speech, thought and writing presentation, but by also introducing new parallel categories on each scale which result from their analysis of a corpus of

twentieth-century narrative fiction, autobiography and newspaper reports. As can be seen in the diagram below, the new categories are the narrator's representation of voice (NV) on the speech presentation scale, narrator's representation of writing (NW) in the writing scale and internal narration (ND) in the thought presentation scale. Category N (narration) often occurs in brackets on the left-end scale of the respective categories because it is closely connected with NV and the narrator's report of a speech act (NRSA), but it is not real speech presentation. Most recently, Short (2007a) has suggested considering NI to be part of narration (like Toolan, 2001) rather than part of the thought presentation scale, and he introduces the category of NT, narrator's presentation or reference to thought.

The order of the different speech presentation scales reflects the linguistic features involved in indirect and direct forms and the faithfulness claims between anterior and posterior discourse or, in other words, the rising degree of narrator involvement moving to the left-hand side and the character involvement moving to the right-hand side of the model.

Speech presentation

As can be seen in the diagram, "speech presentation scales," the presentation of the speech of others can take on different forms depending on the relationship between posterior and anterior discourse. NV contains presentations of minimal speech or verbal activity as well as its manner or style as reported by the speaker, but no mention of the actual form or content of the

utterance is made. NV is different from NRSA to the extent that, in NRSA, the narrator provides information about the illocutionary force of the reported utterance (Semino and Short, 2004, 44); Page's (1973) idea of submerged speech and Short's (1996) Narrator's Representation of Speech are similar concepts. An example is "To her voluntary communications" (Austen, [18161 1985, 262) from Austen's Emma where the reader is informed that speech has taken place. In terms of effect it can be said that it illustrates the point of view of a particular character and that it functions as an introductory statement to speech. It has a signaling function to what follows. It may also be used as reference which summarizes what a group of people has said.

NRSA often consists of one clause where the speech report verb is followed by a noun phrase or a prepositional phrase indicating the topic. Hence, the speech act value is provided, but not the exact content. It is often used to summarize or to give background information. An example is "went on to express so much wonder at the notion of my being a gentleman" (Dickens, [1860] 1999, 109), where the speech act verb "express" is followed by a nominal group after the preposition "at": "the notion."

The following utterance from Charles Dickens's Great Expectations is an example of IS: "Once more, I stammered with difficulty that I had no objection" (Dickens, [1860] 1999, 109). The reporting clause is "Once more, I stammered with difficulty" and the reported clause is "that I had no objection." As mentioned, indirect stretches of speech presentation are often introduced by conjunctions like "that," "if" or "whether" and the reporting clause conventionally precedes the reported clause. In addition, the relationship between reporting and reported clause is that of hypotaxis (that is, of dependence), while for DS it is often claimed that reported clause and reporting clause stand in a paratactic relationship (that is, they are clauses at the same level). According to Lubock, indirect discourse forms tell a character's words whereas direct discourse forms show them. Nevertheless, in comparison to NRSA, indirect speech (IS) seems closer to the character because it contains the speech act, the topic and the propositional content, but not necessarily the precise words of the utterance. As such, it is more faithful to the original, but it also foregrounds the act of reporting.

FIS carries linguistic features from DS and from IS. Other terms used are erlebte Rede or style indirect libre. While the reporting clause from IS and the quotation marks from DS do not often occur, it may contain deictic features from both DS and IS. Therefore, its claims of faithfulness in relation to the Words and structures used in the anterior discourse are not always clear and FIS and FIT are not always easy to distinguish, because often they both share the linguistic feature of modality (Toolan, 2001, 131). According to Semino and son (2004, 13), FIS often creates ironic effects, because it is perceived by the

reader as distancing him/her from what the character said. An example is when Mr. Woodhouse in Austen's Emma worries about a ball. His inference of Emma's disappointment is somewhat diminished by his pessimistic conviction that they should stay at home anyway: "and as for the ball, it was shocking to have dear Emma disappointed; but they would all be safer at home" (Austen. [1816] 1985, 41). Here the reporting clause is missing and the deictic orientation is that of reporting (past tense and back-shifted pronouns).

DS and FDS are the modes of speech presentation in which a character's words are shown. The effect is that of mimesis. DS contains both a reporting and a reported clause, which is marked in inverted commas. DS brings with it a further faithfulness claim as it reports verbatim the speech act value, the grammatical structure and the words of the utterance as well as its propositional content. An example is " 'Emma knows I never flatter her,' said Mr. Knightley" (Austen, [1816] 1985, 42), where the reported clause is distinguished from the reporting clause ("said Mr. Knightley") by inverted commas. The effect of DS is that of vividness and dramatization, as the exact words of the character are given (Toolan, 2001, 120). Toolan also points out that "the choice of direct speech reporting is also to accept a scenic slowing of pace, an enhanced focus on the specificity and detail of an interaction, and a greater pressure on the author to make such text redet mingly interesting" (2001, 129). According to Leech and Short (1981/2007) and Semino and Short (2004), the norm for speech presentation in their corpus of twentiethcentury narrative fiction is direct speech.

FDS, on the other hand, must contain the direct reported clause (mostly in inverted commas), but it does not need to include the reporting clause. FDS is most faithful as it presents the words of the character/or the original with no obstruction from the narrator or the reporter. As most investigations of DS and FDS have differentiated between DS and FDS, Semino and Short (2004, 49) have maintained that distinction in their annotation of the various discourse presentation categories, although they see the difference between DS and FDS Cand likewise between DW and FDW) as a variation of the same category. Also. Semino and Short (2004) notice that it is not uncommon for a stretch of direct speech presentation to move to free direct speech presentation which, therefore, aggravates the annotation process and highlights the danger of making arbitrary decisions. Although Short (2007a) suggests the inclusion of FDS in the category of DS, Busse (2010a) suggests annotating both categories separately and maintaining the traditional distinction

Writing Presentation

Writing presentation, in which the anterior discourse or the original be piece of written discourse, is similar in function to speech presentation.

However, generally, it is stronger in relation to faithfulness claims because writing has been valued more highly than spoken language. In Semino and Short's (2004) corpus, writing presentation is more frequent in the news reports and (auto)biographies than in narrative fiction and also the quantitative distribution of the respective categories for writing presentation is different from those for the other discourse presentation scales (Semino and Short, 2004, 48).

In writing presentation, similar options to those for speech presentation are suggested. NW includes those cases where the narrator reports that the character engages in writing. NRWA is the most frequent category in the Semino and Short (2004, 105) corpus and describes the written act and the value of the writing act, but no further reference is made to the content. Often a prepositional phrase followed by noun phrases occurs. An example would be the summary of the advertisement which searches for Oliver Twist in Dickens's ([1837 1966) Oliver Twist and which Mr. Bumble discovers while reading the newspaper: "And then followed a full description of Oliver's dress, person, appearance, and disappearance: with the name and address of Mr. Brownlow at full length" (Dickens, [1837] 1966, 110).

IW contains the writing act, the topic and the propositional content, but not necessarily the precise written words of the utterance. IW is usually introduced by a reporting clause or a reporting signal. An example of IW is "A letter arrived from Mr. Churchill to urge his nephew's instant return" (Austen, [1816] 1985, 263), in which "a letter arrived" serves as a reporting clause and "to urge his nephew's instant return," because it is clausal, is the reported clause.

FIW may contain features of DW and W and often omits the reporting clause. The effects are similar to that of FIS where the narrator often uses this mode of discourse presentation in an ironic way. An example would be "Mrs Churchill was unwell" (Austen, [1816] 1985, 263) which follows the information that "A letter arrived from Mr. Churchill to urge his nephew's instant return" (Austen, [1816] 1985, 263) so the reader knows that this information must have occurred in the letter that was sent. DW, in line with its counterpart speech category, gives a verbatim account of a written document, but often uses the reporting clause to introduce it. FDW can be equated with genuine writing, as the reporting clause often does not occur.

Thought Presentation

The thought presentation scale contains similar categories to those already discussed, but the functions of the parallel modes of thought presentation are critically different. This is due to three factors that differentiate thought from Writing and speech presentation. When thought is presented, faithfulness cannot be observed because human beings do not have access to the thoughts of others.

Therefore, the summary aspect cannot be applied because we cannot assume that there is an anterior thought act. In addition, it is difficult to determine whether and, if so how, thoughts are construed in linguistic form because although thoughts are part of cognition, it is not conclusive that "all cognition is thought" (Short, 2007a, 236). Therefore, the different forms of thought presentation should be seen as a series of "effect scales" (Short, 2007a, 231).

In the first of the thought categories, internal narration (NI), the character is displayed to be engaged in a specific act of thinking, but neither the propositional content nor the specific thought act are rendered. It is a report of mental or cognitive states and describes the character's cognitive or emotional experience without real reference to his thoughts. NI can also describe a particular emotional reaction. In Semino and Short (2004), NI encompasses all cases where a narrator reports a character's cognitive and emotional experience but not their perceptions. The verb "puzzle," for instance, in the following example by Hardy, alludes to the mental state of the character: "seeing how puzzled Phillotson seemed, Jude said as cheerfully as he could [...]" (Hardy, [1881] 1975, 192). Semino and Short (2004) initially established a parallel between NI and NV on the speech presentation scale because, according to them, both occupy some intermediate position between the straightforward narration of actions and events. The classification of this category as thought presentation, however, is a far from settled issue. Other scholars apart from Short and Semino have also discussed NI from different perspectives such as, for instance, Simpson (1993,245) who is dubious regarding the classification of NI as discourse presentation. Cohn (1978), in turn, had previously suggested that this phenomenon should be understood as a form alpsychonarration" (she adds NRTA and IT into this category). Palmer (2004) of has illustrated that the depiction of conscience and consciousness is not always verbal, but depicted in the mode of psychonarration.

While Semino and Short (2004, 46) opt against labelling these phenomena under the heading of narration because that would exclude its essential mental activity component, Toolan (2001, 119, 143) describes these presentations as "reports of mental or verbal activity which do not purport to be a character's articulated speech or thought." That N ned speech or thought." That Nl does not actually contain characters' mental or verbal activity leads Toolan to argue for the category of NI to be excluded from the thought presentation scale and he placed within the narration proper instead. This reading aligning the narrator- sees NI "as a statement that the narrator makes World of his or her characters" (Short, 2007, 236). Instead Shot 2007. 220 introduces a new category, Narrator's Presentation of Reference to thought (NT), which constitutes more straightforward thought presentation

equivalences of NV and NW and also takes the equivalent position to NW and NV on the thought presentation scale (Short, 2007a, 237). An example would he "he began to think" in "As he left the room, Lord Henry's heavy eyelids drooped, and he began to think" (Wilde, [1890] 1994, 68).

If NI is considered as thought presentation it should be positioned next to NT. If it is seen as part of narration, as Leech and Short (1981, 341-2) and Toolan (2001, 142) have done, and seen as the narration of internal states or events (parallel to the narration of extemal events), then it should be placed on the left hand side of NT and included within N. As such, examples of NI would not present thoughts, but rather the narrator's statements of the internal world. Short (2007a) suggests that what is labeled as NI in the corpus analyzed for Semino and Short (2004,237) should be re-examined because these examples "appear to cover a wide range of different kinds of phenomena" (2007a, 235).

NRTA expresses the thought act, the topic, but not the exact thoughts. Similar to NRSA, the thoughts appear to be summarized, but it is one of the most inexact categories because a summary cannot really take place within thoughts. After Frank Churchill's departure, Emma's deploring attitude towards this is conveyed in NRTA: "and foresaw so great a loss to their little society from his absence" (Austen, [1816 1985, 266). Next, IT consists of the thought act, the topic and the propositional content, but not necessarily the precise details of the thought act. When Emma, in Austen's Emma, engages in matchmaking between Mr. Elton and Harriet, the narrator presents Emma's thought in IT: "she thought it would be an excellent match; and only too palpably desirable, natural, and probable, for her to have much merit in planning it" (Austen, (1816 1985, 63). FIT, which, contrary to FIS, is extensively used in the fictional corpus investigated by Semino and Short (2004, 13), is a very open category, because, in terms of the words, grammar, syntactic, orthographic and deictic distinctions used, it contains features from DT and IT. Some other scholars prefer to amalgamate both categories in what is known as free indirect discourse (FID). Jefferson (1981, 42), for instance, suggests that FD is a mixture of proper narrative or proper speech and thought. Toolan (2001, 135) points out that this amalgamated category can be characterized in multiple ways: as substitutionary narration, as combined discourse, as a contamination, tainting or coloring of the narrative or as a dual voicing.

As previously stated, there are clear functional and effectual differences between the categories of speech and though presentation, despite fact that the various linguistic strategies that realize each form are parallel in structure, as Leech and Short (1981) pointed out and Short and Semino (2004) verified. These differences in functions result from the different nature of the

discourse presentation modes as such, namely that it is impossible for a witness to truthfully report the thoughts of others. Nevertheless, recourse to more direct thought presentation is a more direct way of entering into the character's mind as the narrator is not controlling the thought world of the character. In this end of the scale, DT tends to be construed with a reporting clause and the reported clause presented in inverted commas. The final category, that of FDT, only shows the direct thoughts of the character or thinker and is considered the freest way of reporting the thought process of characters.

Of all the categories described here, free indirect discourse (FD) has been the one studied most frequently for the last 100 years, especially, although not exclusively in narratology. Pascal (1977), McHale (1978) and Bakhtin (1981) have, for example, emphasized the "dual voice" and the effects of FD. Banfield (1982) analyses the implications of FD for narration which she considers "unspeakable" because, unlike discourse (which is both communicative and expressive), narrative does not contain a genuine addressee (nor any textual trace of one), nor a genuine speaker (cited in Toolan, 2006C). Stanzel (1984), in turn, stresses that FID typically comes with internal character focalization and defines FD as mimetic diegesis, because it tells aspects of a character's own words.

To sum up, a few issues regarding the different modes of discourse presentation need highlighting. Generally speaking, the modes of discourse presentation can be viewed along reduced or increased character alignment, although effect or function, especially for the different categories of speech and thought presentation, are different. Speech and writing presentation, on the other hand, are considered to be similar in relation to their associated effects (Semino and Short, 2004,50).

Focalization

The studies devoted to explaining the notion of focalization are not only abundant but also varied as to their informing disciplines and main influences (Herman, 2002, 2009; Toolan, 2001; van Peer and Chatman, 2001). Focalization is a term that emanates from narratology but its impact has been very much felt in stylistic circles too. Although the existing attempts at a definition are, similarly, countless, the origins of such endeavors can be located in the structuralist tradition of which Gerard Genette is one of its most famous proponents. Genette (1972, 1980) coins focalization partly as a reaction to point of view, for the latter, Genette claims, appears to merge two narrative phenomena that need to be understood separately, the now classical distinction of "who sees?" versus "who tells?." Rimmon-Kenan (1983, 2002)

acknowledges Genette's distinction but takes issue with his reasoning as to why the substitution of focalization is necessary:

The story is presented in the text through the mediation of some prism,' perspective,' 'angle of vision,' verbalized by the narrator though not necessarily his. Following Genette (1972). I call this mediation 'focalization' ...). Genette considers 'focalization' to have a degree of abstractness which avoids the specifically visual connotations of point of view. (Rimmon-Kenan, 2002, 72)

Genette's preference for the term focalization, thus, originates from, on the one hand, the fact that point of view appears to overlook the "seer vs. narrator" dichotomy, and, on the other, because point of view may be perceived as semantically overloaded with reference to visual aspects. Other scholars (Rimmon-Kenan, 2002; McIntyre, 2006) have pointed out that focalization is not exactly free from optical references either, so the apparent abstractness that justifies the substitution has not been widely accepted by the academic community. Rimmon-Kenan, nonetheless, incorporates the very useful distinction between "who sees" and "who speaks," so her discussion of the term focalization is, in practice, rather close to that of Genette. More recent approaches, however, tend to avoid dwelling too much on this issue and prefer a general definition, as here exemplified by Jahr: "the submission of (potentially limitless) narrative information to a perspectival filter" (Jahn, 2007,94).

The Genettian model distinguishes three main types of focalization: nonfocalization or zero focalization, internal focalization and external focalization. The first type gives the control of the narrative to the "teller," the narrator, so that filtering of information via any other mediator is virtually discarded; narratives with omniscient narrators would be prototypical examples. In internal focalization the story is viewed through the eyes of some internal participant in the story known as "reflector character." This situation constrains the perspective from which the story is presented to that of the reflector. If that reflector's perspective is kept steady throughout the narrative we can talk of fixed focalization; if the original reflector's angle is alterated with the perspective facilitated by further reflectors, then focalization is said to be variable; finally, if the same story is recurrently being told from the viewpoint of many reflectors, that narrative is characteristically displaying multiple focalization. External focalization indicates that mediation is kept to a minimum; so readers (or audiences in the cases of drama or cinema, for instance) are only given information which is externally accessible, that is, information on characters' attitudes, thoughts or emotions are not narratively contemplated. Extemally focalized narratives could have simply been captured

by a camera or a tape recorder. For instance, instances of dialogue without reporting clauses or stage directions are obvious cases of external focalization.

Genette's classification has been much criticized and reworked although most of the subsequent proposals still acknowledge its value as, at least, the original point of departure. Besides the reservations expressed above by Rimmon-Kenan, Bal (1997) has separately voiced some forceful and convincing criticisms towards the Genettian model; her most important departure concerns the reduction of focalization types to two, internal and external. Bal's critique affects primarily the zero and external categories subsumed under external focalization as, for her, focalizers can either be integral components of the story or external elements without a role in the story-world. Bal objects to Genette's vague indications as to "who is seeing what" in his zero and external groups, but she also calls into question the existence of totally unmediated external or zero focalization as such a level of neutrality is simply impossible to achieve.

Modern narratology enjoys now the advantage granted by temporal perspective on all of these issues, so its practitioners are capable of highlighting flaws and lacks in the original proposals of the past. For instance, narratology considered itself "a timeless and culture-independent discipline" (Jahn, 2007, 94), yet it has nowadays become apparent that "their seemingly neutral theoretical models may have been shaped by cultural and historical contingencies" (Jahn, 2007, 94). To cultural and historical determinants, Jahn (2007, 102) adds the role that cognitive sciences are playing in our understanding of narratological concepts. Where more traditional positions rely mainly on the textual components of narratives, cognitive considerations are increasingly incorporating the role of the reader as active participant in the creation of mental worlds to which they are transported in the process of reading (deictic shift theory) or which they cognitively project (text world theory, mental spaces theory). Finally, it needs to be acknowledged that discussions on the nature of focalization are never too far apart from those on point of view. As a general tendency, point of view seems to have been the term of choice adopted by stylisticians (McIntyre, 2006; Short, 1996; Simpson, 1993) whereas narratologists have opted for the former coinage. It appears that the particularly strong linguistic basis that stylisticians endow their analyses with becomes the main point of divergence between the two camps. Having said this, though, the two groups represent akin disciplines, not discrete and totally dissenting schools of thought.

Rimmon-Kenan seems to be a case in point to illustrate the affinities existing between stylistic and narratological approaches. In her discussion of the

various types of focalization, she concurs with Bal in describing an external as well as an internal perspective, but she also claims that these two angles need to be conceptualized in relation to the so-called facets of focalization (2002, 78). These facets work on three main planes directly influenced by Uspensky (1973), the spatio-temporal, the psychological and the ideological planes. Fowler (1986) also benefits from the latter's taxonomy although his research has bome fruit in relation to point of view, not focalization types. Further distinctions are based on the genre primarily chosen by narratologists, narrative fiction, whereas stylisticians have managed to spell out the peculiarities of the use of point of view use as much in prose, poetry or drama (Short, 1996; McIntyre, 2006). Rather than creating a false sense of opposition between these two disciplines, focalization and point of view should be viewed, instead, as sides of the same coin each one capable of bringing to the fore distinctive aspects of textual style.

Foregrounding

4grounding is deviation and parallelism, and 4grounding @ at u. This "sentence" displays multimodal and linguistic elements that go beyond the (linguistic) signs usually expected to be used in a reference book: The cardinal number "4" has replaced the prefix "fore." Users of SMSs or chatrooms will know that this is an abbreviation frequently appearing as a communicative device in new media. To decode the meaning of the cardinal number in this sentence the reader has to utter it. What is exploited here is the homophony between the cardinal number "4" and the prefix "fore." Similarly, "u" is an abbreviation, which stands for the personal pronoun "you" and the icon " " is used as a visual realization of the concept of "smiling."

Many scholars draw on visual features to explain foregrounding as elements of a text which stand out somehow or are deviant from (or parallel to what one would expect or what is conventional. Short (2007c), in his weh be course, for example, uses similar and deviating fonts as well as colors to highlight the concepts and the terms "parallelism" and "deviation AB example which illustrates foregrounding is the title of Dylan Thomas's mom Grief Ago" (1936). This is an example of linguistic deviation at the level of lexico-grammar, because the collocation of grief" with "ago" is rare in Standard English. The adverb "ago" frequently collocates with expressions of time, as in "a week ago," and not with emotions like "grief." At the phonological level elements such as rhyme and alliteration are examples of foregrounding

Foregrounding is closely related to the Russian Formalist concept of "defamiliarization" (ostranenie, i.e., "estrangement" or "making strange") which was introduced into literary criticism by Sklovskij (1917). Šklovskij

the works by CDA proponents such as Norman Fairclough (1989) or The van Leeuwen (2005a), for example. Halliday's grammar is particularly suited for CDA because of its orientation towards context, that is, situational, generic and ideological. Furthermore, its three-dimensional approach to language the textual, interpersonal and ideational provides CDA with a broad range of grammatical tools for analysis as well as a theoretical framework. These allow the analyst to disclose the ideologically loaded as well as constructed nature of discourses and enacted] hegemonic genres-specific ways of using language to achieve purposes of social domination" (van Leeuwen, 2006a, 290).

Critical discourse analysts like Ruth Wodak, for example, also use other methods, such as argumentation strategies and forms of conversation analysis. Van Dijk (1993), however, has also illustrated that a multidisciplinary approach is necessary which "chooses and elaborates theories, methods and empirical work as a function of their relevance for the realization of sociopolitical goals" (van Dijk, 1993, 252). This also affects the ways that Critical discourse analysts engage with critical social theory as proposed by Foucault, Bourdieu, Habermas, Harvey or Giddens to name but a few. CDA shows how discourses create power structures, but key concepts like that of ideology are sometimes defined and used in slightly different ways by the various analysts. For example, van Dijk (1993, 258) considers ideologies to be "worldviews" that create "social cognition": "schematically organised complexes of representations and attitudes with regard to certain aspects of the social world, e.g. the schema [...] whites have about blacks." Fairclough's (1997, 26 view of ideologies is, in turn, based on Marxism. The topics that have been explored by critical discourse analysts range from racism and antisemitism (Wodak et al., 1990), immigration and asylum seeking van Leeuwen, 1999), the business world and the language of politics (Fairclough, 1995), to gender, education, doctor-patient scheme and so on.

Stylistics shares with CDA the focus on the identification of styles as well as a number of similar tools for analysis, although CDA's data is usually nonliterary language. In turn, Hallidayan grammar has been crucial for stylistics analysis, but the stylistic analysis is often less politically motivated.

Blank verse is the name given to unrhymed lines in iambic pentameter, as we see it in much Elizabethan drama as well as in poetry such as Milton's Paradise Lost ([1667) 1983) and Wordsworth's "The Prelude" ([1805] 1983). In much of Shakespeare's drama, noble people speak in blank verse, whereas prose is employed as the natural speaking idiom of characters who belong to the lower classes. In Hamlet (1604-5 in Evans, 1997), additional meaning is created by metrical form when it comes to the style of the lines spoken by Hamlet himself. Following the conventions mentioned above, Hamlet's lines are in blank verse at the beginning of the play, as are those of other nobles, but when he changes and appears to be mad later in the play, he starts speaking in prose. Interestingly, however, he only does so when in the company of other people, but switches back to blank verse when alone, which may be seen as a metrical indication that rather than being truly mad, he only pretends to be so.

Free verse is a term employed to refer to poetry consisting of verse lines whose rhythmic pattern is not organized into metrical feet. In addition to this, most poetry in free verse has irregular line length and is not structured by means of end rhyme. Today, much English poetry is written in free verse.

Narrative

Establishing exactly what constitutes a narrative is not an easy enterprise. We can start by considering the following definitions:

One will define narrative without difficulty as the representation of an event or a sequence of events. (Genette, 1982, 127;

The representation [...] of one or more real or fictive events communicated by one, two or several [...] narrators [...] to one, two or several narratees. (Prince, 2003a, 58)

Simply put, narrative is the representation of an event or a series of events. "Event" is the key word here, though some people prefer the word "action." Without an event or an action, you may have a "description," an "exposition," an "argument," a "lyric," some combination of these or something else altogether, but you

won't have a narrative. (Abbott, 2008, 13)

These are but three of the many attempts by scholars to come up with a comprehensive description of narrative. Besides the above, Toolan (2001, 1) also suggests that the layman's understanding of narratives as processes requiring a "teller," something to tell a "tale"), and someone to tell it to the "addressee') leaves out some of the traits that should be integral to this notion. For instance, should narratives be reduced to these three elements, then, it would be hard to distinguish narratives from other communicative events such as conversations with friends, phone calls or celebratory speeches. For Toolan, one aspect in which narratives differ from the latter three is in the salient role

played by the teller of the story, so that processing the narrative is as much to learn about the occurrences in the tale as it is about experiencing how the telling entity presents it. Thus, Toolan's contribution to a definition states:

A narrative is a perceived sequence of non-randomly connected events, typically involving, as the experiencing agonist, humans or quasi-humans, or other sentient beings, from whose experience we humans can 'learn.' (Toolan, 2001,8)

Not only does this new proposal refer to the necessary presence of "events" as Genette, Prince and Abbott do above too, but also, the purposeful sequentiality of such events is emphasized. Therefore, narratives presuppose that something needs to happen in a particular order, irrespective of whether the order in which events occur and the actual presentation (textual or otherwise) of those occurrences coincide; moreover, those happenings need to involve some cognizant beings from whose experiencing humans can learn. Here "leam" appears to be used in the very broad sense of the type of benefit gained by recipients of that narrative (emotional, psychological, informative or cognitive). According to this definition, a ballet performance, as much as novels, short stories, oral narratives, folk tales, pantomime, films or comic strips can be categorized as different types of narrative. Definition endeavors, consequently, are not particularly scarce. In an attempt to tackle this issue, Mary-Laure Ryan (2007) has resorted to list a series of basic characteristics which determine the prototypical or marginal nature of a particular text as narrative, rather than its inclusion or not in the category since, as she explains herself, narratives are, by definition, "fuzzy":

Rather than regarding narrativity as a strictly binary feature, that is, as a property that a given text either has or doesn't have, the definition proposed below presents narrative texts as a fuzzy set allowing variable degrees of membership, but centered on prototypical cases that everybody recognizes as stories. (2007, 28)

Her list of conditions for narrative prototypicality includes:

Spatial dimension: 1) Narrative must be about a world populated by individual existents.

Temporal dimension: 2) This world must be situated in time and undergo significant transformations. 3) The transformations must be caused by non-habitual physical events.

Mental dimension: 4) Some of the participants in the events must be intelligent agents who have a mental life and react emotionally to the states of the world. 5) Some of the events must be purposeful actions by these agents.

Formal and pragmatic dimension: () The sequence of events must form a unified causal chain and lead to closure. 7) The occurrence of at least some

of the events must be asserted as fact for the story-world. 8) The story must communicate something meaningful to the audience. (Ryan, 2007, 29)

In relation to this set, a ballet performance could, at least for some, be acknowledged as a more unorthodox member of the category of narrative under condition 6. The kind of causality generally experienced in the more traditional forms of narrative fiction could be said to be missing in particularly experimental performances of ballet more concerned with emotional reactions from the audience than with the sequentiality of the various acts and scenes. As Ryan states, we might still have enough components to include dance formats as narratives although we might need to situate them at the more marginal end of the cline.

Traditionally, narratologists have drawn a basic distinction between two aspects of narratives, differently termed by the various theorists, but which essentially encompass similar (if not totally equivalent) features. An initial differentiation was made between "fabula" and "sjuzhet" by the Russian Formalists (Propp, [1928] 1968; Tomashevsky, 1965). The French structuralists preferred to use "histoire" and "discours" respectively, whereas more contemporary narratologists have opted for "story" and "discourse" (Chatman, 1978). The dichotomy, thus, is established between fabula histoire/story on the one hand, and sjuzhet/discours discourse on the other. The first set conveys the sense of the chronological sequence of events as these would have logically occurred, whereas the second refers to the actual manifestation of those events, not necessarily in fact, hardly ever) in the same order in wir ch they took place. Wales makes a distinction between the "deep structure" of a narrative versus the "surface structure" (2001, 367). The terms here mentioned are not free from controversy either so, as analysts we need to be aware of the variation in terminology and choose the most appropriate to serve our interests.

The work on narrative issues undertaken by the Russian Formalists can be illustrated by reference to Propp's (1928] 1968) structuralist morphology. Based on his analysis of 115 Russian fairy tales, Propp suggests a framework of events (functions) as they occur in those fairy tales. Similar to a corpus linguist his main aim is to identify recurrent patterns and features (constants) against the background of deviating, random or unpredictable elements, which he calls variables (Toolan, 2006a, 461). He identified 31 recurrent functions, which are manifested in a fixed sequence, although some of those can also be seen as pairs. Therefore, the functions of characters in a story remain constant. Some examples of those functions are, for instance, the first one which identifies that "one of the members of a family absents himself from home (an extreme exponent of this function represented by the death of one of the parents)," or function sixteen which stresses that "the hero and villain join in direct combat" (Toolan, 2006a, 460). Among the seven basic character roles also singled out as constants in fairy tales are the following: villain, donor/provider, helper, princess

(+father), dispatcher, hero (seeker or victim) and false hero. Characters may take on more than one role and one role may also be represented by more than one character. Propp's framework has been used especially in anthropological studies, and, strikingly, applied to genres as diverse as children's stories and crime series, although it has also been criticized for being too reductionist and for ignoring the various levels of detail that are part of a story. Yet, there exists a substantial degree of agreement on what readers (intuitively) regard as essential to a narrative, which suggests that some structure based on common sense does indeed exist. Although what could be called "narrative competence" and "culture-specific abilities" or "intuitive knowledge" remain highly debatable notions, Propp's model can also be read as an example of how readers are able to rely on intuitions and to deduce or summarize the plot.

Barthes's ([1966] 1977) functional characterization of narratives illustrates the French structuralist take on issues relating to the concept of narrative. He distinguishes between three major levels of narrative structure: (a) functions (as in Propp); (b) actions (roughly, equivalent to characters-cf. Greimas's (1966) actants) and (c) narration (equivalent to discours or discourse). In Barthes's framework, functions create coherence in narratives and can be further subdivided into functions proper and indices. Indices give the reader information about the characters' state of mind and the general atmosphere.

Finally, it is also worth mentioning that the analysis of narratives has not been circumscribed to the fictional variety. The most famous take on natural narratives is that of Labov and Waletzky (1967) and Labov (1972) in what has come to be known as the sociolinguistic approach. In his investigation of the narratives of ordinary people and of personal experience, Labov (1972, 359-60) sees narratives as "one method of recapitulating past experience by matching a verbal sequence of clauses to the sequence of events which [...] actually occurred." One of Labov's ultimate aims in describing narratives from a sociological perspective was to relate the narrator's social characteristics class, gender, age, ethnicity, geography—to the structure of the narrative. Labov and Waletzky (1967, 12) suggest an analytical framework which isolates "the invariant structural units which are represented by a variety of superficial forms." The clause is seen as the fundamental grammatical unit, which is accompanied by semantic functions. Clauses are combined with one another and grouped into sections with different functions. They answer for example, questions, such as "What happened next?" or "What's the point? "

Therefore, narratives can be said to follow a specific pattern perfectly encapsulated by Labov's famous "diamond" picture which describes the progression of an oral narrative. Oral narratives are said to be built around the following six concepts: "evaluation," "resolution," "coda," "abstract." "orientation" and "complication action." The orientation, for instance, provides

information about the setting or the context. The abstract "is an initial clause that reports the entire sequence of events of the narrative" (Labov, 1997, 402). The coda is a functional device for returning the verbal perspective to the present moment" (Labov and Waletzky, 1967, 39). There may also be the "climax," and the "denouement," which convey the resolution of the action. Clauses which are not chronologically ordered would illustrate the complicating action and provide the referential function which reports a next event. The relevant question is that of "What happened?." The evaluative function of the narrative which answers the question "What's the point?" is realized through clauses which tell the readers what to think about a person, place, etc.

Narrative is crucial to a stylistic investigation. Yet, because the production, reception and use of narrative is seen as a meta-code or a human universal (and a feature which distinguishes human beings from animals in an even more pronounced way than creative language usage as such), narrative has been explored from other linguistic angles as well, e.g., within the areas of sociolinguistic investigation and psycholinguistics, as well as cognitive linguistic approaches.

Point of view

As is the case with focalization, much has been written on the notion of point of view (Bray, 2007; Fowler, 1986, 1996; Jeffries, 2000; Leech and Short, 2007; McIntyre, 2004, 2006; Sotirova, 2004, 2005). These two concepts have been discussed and analysed in parallel ways in narratology and stylistics respectively. "Point of view" appears to have won the bat le for stylisticians, especially as it has traditionally been the term of choice for some of the founding and most famous practitioners of the discipline: Fowler, Leech, Short, Simpson and Wales. The latter (Wales, 2001), for instance, underscores the variety of meanings this notion encompasses:

1. Point of view in the basic aesthetic sense refers to "angle of vision" [...].

2. Even in ordinary speech we use point of view in the figurative sense of the way of looking at a matter, rather than a scene, through someone's eyes,

or thoughts [...].

3. Point of view in the figurative sense entails not only the presence of a conceptualizing character or focalizer, but also a particular way of conceptualizing: a world-view or ideology [...]

4. An aspect that has been particularly developed relates point of view to the larger discourse situation of narrator-text-reader. (Wales, 2001, 306–7)

In practice, a very general definition of the term understood as the perspective or filter through which events are perceived would need to incorporate much of the content present in the four senses described by Wales. Practitioners of stylistics have traditionally used the point of view typology suggested by

Fowler (1977, 1986, 1996) who borrowed it in tum, from that of Uspensky (1973). Fowler (1996, 162) distinguishes three types, spatiotemporal, ideological and psychological point of view. The first type refers to how readers position the story in space and time. In relation to the former, for instance, readers need to get a sense of the various locations of the storyline and deictic markers such as adverbs ("here," "there"), demonstrative determiners and pronouns ("this" and "that"), deictic verbs ("bring." "take") and locative expressions (prepositional phrases) help to anchor the narrative spatially. Similarly,

temporal parameters allow us to place the story alongside a timeline, despite the fact that events can and very often do appear temporally scrambled in the form of flashbacks or flashforwards. 0 Temporal representation, in particular, has been of special interest to narratologists (Genette, 1980; Rimmon-Kenan, 2002) who generally approach the relationship between textual time and story time via three notions: order, duration and frequency. Order considers the possible asynchronisms between the storyline and its textual presentation. Asynchronous textual manifestations, called "anachronies" by Genette (1980), can introduce events that happened prior to the current moment of the story (analepses or flashbacks) or anticipate situations that will occur later on (prolepses or flashforwards). Duration refers to the ratio of textual space to temporal length. For instance, the pace of a story may appear rushed or quickened if certain temporal gaps are noticeable because they have not been given any textual space whatsoever; these instances of temporal vacuum are known as "ellipsis." The opposite effect is achieved when events that would not logically stretch in time are awarded considerable textual presence, what is known as "descriptive pause."summary" is achieved by condensing a lengthy period of time in a relatively small textual space. Finally, by means of the "scene" technique, story and text duration are presented as very similar, as in dialogue or in plays.

The final way of dealing with the ratio textual versus story time is via the notion of frequency. Frequency refers to the number of times that a particular incident in the story is textually represented: singulative frequency is the default parameter so events would be made textually manifest as often as they occur in the story, that is, we are told once what happens once, twice what happens twice and so on. Repetitive frequency, on the other hand, emphasizes a particular event by textually representing it more than once, as when various characters convey the same circumstance from their own perspective, with as many recountings as there are characters. Iterative frequency, finally, spares us the need to unnecessarily encounter situations that would happen routinely on more than one occasion (for instance, narrators do not inform their readers of une number of times characters presumably wake up, have showers, or eat lunch, unless these happen to have some special significance for the story).

Linguistically, temporal point of view is indicated via deictic adverbs ("now," "then"), tense (present, past) and temporal expressions (prepositional phrases). Fowler's spatio-temporal point of view seems rather close, although not totally equivalent because the temporal component is not so clearly established there, to sense number one in Wales's definition above.

The second type of point of view suggested by Fowler is ideological point of view defined as "the system of beliefs, values, and categories by reference to which a person or a society comprehends the world" (1986, 130); it is sense number three in Wales's classification that appears evoked now. Fowler subsequently considers who might be behind the transmission of a particular ideology in texts, whether the author via the figure of the implied author, the narrator, the author surreptitiously speaking through the narrator or simply the characters. Such considerations would place ideological point of view on a similar plane to sense four in Wales's definition. Fowler (1996, 166) highlights two ways in which ideological point of view is linguistically indicated, one explicit and a second more indirect. Modality is one of the ways in which ideology is explicitly projected in texts and can be realized in the form of the prototypical modal verbs or via modal adverbs or sentence adverbs, evaluative adjectives and adverbs, verbs of knowledge, prediction and evaluation and generic sentences. The second, less conspicuous, way in which ideological point of view is manifested has actually developed into a wellknown way of describing the linguistic (verbal and otherwise) realizations of narrators', authors' and characters' consciousness, the so-called mind style:

The world-view of an author, or a narrator, or a character, constituted by the ideational structure of the text. From now I shall prefer this term to the cumbersome "point of view on the ideological plane." (Fowler, 1996,214)

There has been plenty of work on the way mind style functions and the markers that point to individual manifestations in texts. Some of the linguistic indicators that have been named as mind style projectors include transitivity patterns, metaphor use, under- and over-lexicalization and certain syntactic structures. Psychological point of view, finally, "concerns the question of who is presented as the observer of the events of a narrative, whether the author or a participating character, and the various kinds of discourse associated with different relationships between narrator and character" (Fowler, 1996, 169-70). This final take on point of view is more clearly reminiscent of the distinction "who sees" vs. "who tells" traditionally discussed in narratology in relation to focalization and, as a matter of fact, Fowler does liken the two terms (1996,169).

As stated above, stylisticians seem to have preferred "point of view," although they have not discarded "focalization" totally, especially as it relates to

the role of characters as "focalizers" or "mediators." One important factor that seems to clearly differentiate the two main disciplines dealing with issues of point of view and focalization is the primarily linguistic slant of stylistics in relation to its sister discipline, narratology. A particularly illuminating example of how linguistics can help practitioners decide on the various types of point of view is provided by Short (1996) who, instead of proposing a further taxonomy, gathers a list of likely viewpoint indicators comprising of schema-oriented language, value-laden language, given versus new information, deixis, representations of thought and perception and psychological sequencing. Short (2000) and others (McIntyre, 2004, 2000) have extended the list to also include graphology, presupposition and Grice's co-operative principle. Simpson (1993), in turn, has attempted to combine narratological aspects and linguistic indicators in his "modal grammar of point of view in narrative fiction."

Referent

In one sense of the word, the referent is the extra-linguistic entity to which a linguistic expression, as well as other types of sign, refers. The referent may be imaginary as well as real. While the referent of the lexical item "book" is thus the object of a book in the external world, the referent of "unicom" is the object of this particular fantasy creature in an imagined external (i.e., extralinguistic) world. When describing the meaning of a given word or expression, the referent must be distinguished from the signified, which is the abstract notion, rather than the object, to which a signifier refers.

Secondly, in the study of cohesion, reference is a linguistic phenomenon which involves a referring item such as a pronoun and a linguistic item referred to, that is, the referent. In the utterance "I know a good photographer. He works in Manchester," "he" is the referring item and "a good photographer" is the referent. Alternatively, it might be argued that "the" and "a good photographer" are a case of "co-reference," that is, of two referring items which both refer to the same real-world (or fictional) referent. Text-internal reference may either be anaphoric or cataphoric. In anaphoric reference, the referent precedes the referring expression as in the example above. In cataphoric reference, the referent follows the referring expression as in the first line of lan McEwan's novel Black Dogs (1992) "Ever since I lost mine in a mad accident when I was eight, I have had my eyes on other people's parents," where "mine" refers forwards in the text to "parents." This extract illustrates well the process of referent-tracking that readers engage in when reading in order to make sense of the text. In the search for a cataphoric referent of the pronoun "mine," readers may first tend to wrongly assume that "mine" refers to "eyes," since the nominal nature of the latter makes it qualify as the first possible referent in the text succeeding the referring item. As we read along, however, we realize that our capacity for referent-tracking has fooled us, as probably intended by the author.

Online Reading

Stylistics

About author: Mustafa Gad

0 comments:

Labels

Popular Posts

Online Reading

Stylistics

About author: Mustafa Gad

0 comments:

Follow Us

Labels

Popular Posts