Cambridge University Press
9780521853033 - Style - Language Variation and Identity - by Nikolas Coupland
Excerpt
‘Style’ refers to a way of doing something. Think of architectural styles and the striking rustic style of house-building in rural Sweden. That particular style – what allows us to call it a style – is an assemblage of design choices. It involves the use of timber frames, a distinctively tiered roofline, a red cedar wood stain and so on. We can place this style. It belongs somewhere, even if the style is lifted out of its home territory and used somewhere else. It has a social meaning. The same is true for styles in all other life-domains. Cultural resonances of time, place and people attach to styles of dress and personal appearance in general, to styles in the making of material goods, to styles of social and institutional practice, perhaps even to styles of thinking. We could use David Machin and Theo van Leeuwen’s (2005) idea of ‘social style’ to cover all these. The world is full of social styles.
Part of our social competence is being able to understand these indexical links – how a style marks out or indexes a social difference – and to read their meanings. The irony is that, if we ourselves are closely embedded in a particular social style, we may not recognise that style’s distinctiveness. Reading the meaning of a style is inherently a contrastive exercise. You have to find those red cedar buildings ‘different’ in order to see them as having some stylistic significance. This is the old principle of meaning depending on some sort of choice being available. But style isn’t difference alone. When we use the term ‘style’ we are usually attending to some aesthetic dimension of difference. Styles involve a degree of crafting, and this is why the word ‘style’ leaks into expressions like ‘having style’, ‘being in style’ or ‘being stylish’. The aesthetic qualities of styles relate, as in the case of the Swedish red cedar buildings, to a process of design, however naturalised that process and its results might have become in our experience. We talk about ‘style’ rather than ‘difference’ when we are aware of some holistic properties of a practice or its product. A style will ‘hang together’ in some coherent manner. Engagement with style and styles, both in production and reception, will usually imply a certain interpretive depth and complexity. Although we are considering ‘style’ as a noun at this point, when we refer to ‘a style’ and to ‘styles’ (plural), and giving styles a quality of ‘thing-ness’, the idea of style demands more of a process perspective. I think we are mainly interested in styles (noun) for how they have come to be and for how people ‘style’ (verb) meaning into the social world. ‘Styling’ – the activation of stylistic meaning – therefore becomes an important concept in this book.
This general account of style can of course be applied to linguistic forms and processes too. We are all familiar with the idea of linguistic style, and most people will think first of language in literary style. Literary style relates to the crafting of linguistic text in literary genres and to an aesthetic interpretation of text. This book is about style in speech and about ways of speaking, not about literary style, although it would be wrong to force these areas of study too far apart. The book is about style in the specific research context of sociolinguistics, where concepts very similar to ‘social style’ have been established for several decades. The general sociolinguistic term used to refer to ways of speaking that are indexically linked to social groups, times and places is dialects. Dialects are social styles. Some dialects are in fact rather like red cedar timber buildings, redolent with meaningful associations of rurality and linked to particular geographical places. They have strong cultural associations, especially when we look at them contrastively. Dialectologists have traditionally looked for boundaries between dialect regions, and traced the evolution of dialects over time and the consequences of dialects coming into contact with each other (Chambers and Trudgill 1999).
We are likely to think of dialects in this sense as being the social styles of yesteryear, largely out of step with the social circumstances of contemporary life. But dialect differences are of course a characteristic of modern life too. Dialects are evolving social styles and they can be read for their contemporary as well as their historical associations – associations with particular places (geographical dialects) and with particular social groups (social dialects). Dramas associated with dialect are played out as much in cities as in rural enclaves, and sociolinguistics for several decades has enthusiastically teased out the complexities of language variation in urban settings. The human and linguistic density of cities invites an analysis in terms of ‘structured difference’. Cities challenge the view that one discrete social style (e.g. a dialect) is associated with one place, which was the basic assumption in the analysis of rural dialects. It has become the norm to consider cities as sociolinguistic systems that organise linguistic variation in complex ways. But understanding the social structuring of styles, even in the sophisticated manner of urban sociolinguistics, is not enough in itself. We need to understand how people use or enact or perform social styles for a range of symbolic purposes. Social styles (including dialect styles) are a resource for people to make many different sorts of personal and interpersonal meaning. As I suggested might be generally true for intellectual interest in style, what matters for linguistic style is more to do with process than with product, more to do with use than with structure. Stylistic analysis is the analysis of how style resources are put to work creatively. Analysing linguistic style again needs to include an aesthetic dimension. It is to do with designs in talk and the fashioning and understanding of social meanings.
So this is not a book about dialectology either. My starting point is certainly the sociolinguistics of dialect, as it has been carried forward by variationist sociolinguistics in the tradition of William Labov’s research. This is where the term ‘style’ was first used in sociolinguistics, and one of my aims for the book is to map out the main steps that sociolinguists have taken using the concept of style. This will initially be a critical review, focusing on the limited horizons of style research in variationist sociolinguistics. The positive case to be made, however, is that, under the general rubric of style, sociolinguistics can and should move on from the documenting of social styles or dialects themselves. It should incorporate the priorities I have just sketched – analysing the creative, design-oriented processes through which social styles are activated in talk and, in that process, remade or reshaped. This means focusing on particular moments and contexts of speaking where people use social styles as resources for meaning-making. It means adding a more active and verbal dimension (‘styling social meaning’) to sociolinguistic accounts of dialect (‘describing social styles’).
To set the scene for later arguments and debates, several core concepts need to be explored in this introductory chapter. First we need to consider variationist sociolinguistics and its general approach to style. Then we will look back at the early history of stylistics (the general field of research on style in linguistics), to appreciate the climate in which sociolinguistics first came to the idea of style. The idea of social meaning then comes up for initial scrutiny. Looking ahead to the more contemporary research that this book mainly deals with, we will then consider research methods and the sorts of sociolinguistic data that we can deal with under the heading of style research. The wider relevance of style to contemporary social life, which can be characterised by the term ‘late-modernity’, is then reviewed. Finally in this chapter, I give a short preview of the structure of the rest of the book.
Sociolinguistics is, as they say, a broad church. The blander definitions of sociolinguistics refer to studying language ‘in society’ or language ‘in its social context’. Other definitions focus on studying linguistic diversity or language variation. What these simple definitions have in common is that they give priority to language, then add some summary idea of what aspect of language is to be given priority (its variability) or what sort of data is to be given priority (social manifestations of language). Definitions like these have to be understood historically. It was once important to stress ‘social contexts’ in defining sociolinguistic priorities in order to challenge types of linguistics where actual occurrences of spoken language were not given priority. Even though most people would agree that using language is an inherently social process, sociolinguists needed to make a case for observing language as it is used in everyday life and for not relying on intuited or fabricated instances of language. Stressing variability has been important in order to resist the ideological assumption that what matters in language is linguistic uniformity and ‘standardness’. William Labov used the notion of secular linguistics to describe his approach to language variation and change. The idea was that studying variable language forms, ‘non-standard’ as well as ‘standard’ forms, challenges what we might think of as the high priesthood of theoretical linguistics and its reliance on idealised linguistic data. It also challenges the belief that ‘standard’ language is more orderly and more worthwhile than ‘non-standard’ language.
But the study of language variation and change has been in the mainstream of sociolinguistics for four decades. Variationist sociolinguistics, as the approach developed by Labov is generally called, has developed its own powerful principles of theory and method (Chambers 1995/2003; Labov 1966, 1972a, 1972b, 1994, 2001a; Chambers, Trudgill and Schilling-Estes 2004). In this book I intend to take the considerable achievements of variationist sociolinguistics for granted, and to ask what it has not achieved, particularly in relation to the notion of style and the active dimension of styling. So, as I have mentioned, my orientation is a critical one, although I intend it to be constructively so. The negative part of my argument is that variationist sociolinguistics has worked with a limited idea of social context – and styling is precisely the contextualisation of social styles. The survey designs of variationist research, which have been remarkably successful in revealing broad patterns of linguistic diversity and change, have not encouraged us to understand what people meaningfully achieve through linguistic variation. Variationist sociolinguistics has produced impressive descriptions of social styles, but without affording much priority to contextual styling.
What then are the general features of the variationist approach? Sociolinguistic surveys of language variation give us detailed descriptions of how linguistic details of regional and social accents and dialects are distributed. (‘Dialect’ is a general term for socially and geographically linked speech variation, and ‘accent’ refers to pronunciation aspects of dialect.) Speakers are not fully consistent in how they use accent or dialect features. Their speech will often, for example, show a mixture of ‘standard’ and ‘non-standard’ forms of the same speech feature. Nor are individuals within any particular social category identical in their speech. So the sort of truth generated in variationist research is necessarily one based in generalisations and statistical tendencies. These are ‘probabilistic’ truths, expressing degrees of relative similarity and dissimilarity within and across groups of speakers and social situations. The convention is to produce averaged statistical values (e.g. percentages of people’s use of a particular linguistic feature in a particular social situation, or factor loadings in statistical tests) to represent patterns of linguistic variation. So, accent variation between two different groups of speakers is usually represented as the difference between one statistical value (perhaps a percentage) and another.
Variationist research has very expertly shown that ‘speaking differently’ has to be defined in several stages. Stage one is typically to identify a group of people who share a geographical characteristic, such as living in the Midlands city of Birmingham in England, or for that matter Birmingham in Alabama in the Southern USA. Within this territory or ‘community’ of people who have lived in the city for all or most of their lives, sub-groups are identified based on social criteria. This sort of classification isolates, to take a random example, the category of ‘young females in Birmingham with working-class jobs’, distinguishing them from other social categories. In a second stage, the research samples the speech of the different groups, usually through extended one-to-one sociolinguistic interviews. The researcher then counts how often a particular speech feature is used.
For example, in the English Birmingham, the issue might be how often each speaker pronounces the diphthong vowel in words like right and time with a phonetically backed and rounded starting point. In this example, the local Birmingham pronunciation [ɔI;] is in opposition to [aI;] which is the less localised and more ‘standard’ variant in England. Phonetic forms occupying intermediate positions between these variants might also be recognised. Variant forms of sociolinguistic variables tend to be influenced by the details of their linguistic placement. For pronunciation variables (linked specifically to a speaker’s accent, then), the positions that different pronunciation forms occupy in the stream of speech-sounds, and the sets of words that they occur in, are factors that are likely to impact on the frequency with which they are used. These patterns might affect everyone’s speech. A typical finding would then be that most speakers in the sample would in fact use a mixture of different pronunciation forms – e.g. using both ‘standard’ and ‘non-standard’ variants of this sociolinguistic variable (ai). But overall frequencies of use would very probably differ across speakers and sub-groups when statistical averages are taken.
At the end of the process of categorising and counting the distribution of various linguistic variants in a body of data, a type of statistical truth would emerge. It might allow us to say that, overall, Birmingham speech does indeed have some distinctive tendencies of pronunciation – different from the speech of other regions and from ‘standard English’ pronunciation. That is, descriptively speaking, Birmingham speech is a relatively distinctive social style. The descriptive evidence would go some way towards distinguishing the city as a ‘speech community’, even though the ‘standard’, less-localised forms of speech crop up in Birmingham too. But people living outside the city would use some of the local or ‘non-standard’ feature less often than those living in the city, or not at all. Looking at how speech is socially organised within the city, we would probably be able to say that the speech of particular social sub-groups in Birmingham differs in some statistical respects. Perhaps, overall, women in Birmingham use the [ɔI;] feature in words like right and time less often than men do. Perhaps women with more prestigious jobs use it less than women with low-prestige jobs. So there are social styles, at least in a quantitative sense, associated with these groups too.
Labov, however, doesn’t use the term ‘style’ in this sense. He refers to what I am calling ‘social styles’ of speech simply as ‘social variation’. He reserves the terms ‘style’ and ‘stylistic variation’ for a further sort of language variation that can be detected in sociolinguistic interviews (e.g. Labov 1972b). This is when he is able to show that, again in a statistical sense, individual people speak ‘less carefully’ at some points in an interview than they do at other points. When they are being ‘less careful’ or more relaxed they will typically use features of the local style more frequently than in their supposedly normal interview speech. In this way Labov introduced the idea of ‘stylistic variation’ to refer to ‘intra-individual’ speech variation – variation ‘within the speech of single individuals’. This became a very familiar claim in community-based studies of language variation and change, and we will look at it in much more detail in Chapter 2. But it is important to note that, although Labov is mainly concerned with social style at a community level, his original insight about stylistic processes related to the individual speaker and to particular social contexts of speaking. That is, he was interested in what happens when an individual speaker delivers a version of a social style in a range of particular speaking situations. This proves to have been a seminal insight. As we shall see, however, the survey methods that Labov pioneered tend not to give priority to the local processes through which this happens. They orient much more to styles than they do to styling. The convention of basing variationist research on speech in interviews clearly limits the range of social contexts in which styling can be observed and analysed.
Several other sociolinguistic traditions, beyond variationism, are fully sensitive to contextualisation processes and have been so from the earliest days of sociolinguistics. The ‘active contextualisation’ perspective on social style that I am arguing for in this book is already established in other parts of sociolinguistics, and was central to Dell Hymes, John Gumperz and others’ conception of the ethnography of speaking (Hymes 1962, 1996; Bauman and Sherzer 1989; Gumperz and Hymes 1972). The theoretical tension that we have to deal with in later chapters is in fact well summed up by the contrasting implications of the terms ‘speech’ and ‘speaking’. The variationist study of social styles/ dialects has oriented to speech and to speech data, when it also needs to orient to speaking and to the styling of meaning in social interaction. This is not an oversight or even a limitation of variationist sociolinguistics in its own terms. Variationism has simply set itself other primary objectives, linked to understanding language systems and how they change, rather than understanding social action and interaction through language. The objectifying priorities of variationist sociolinguistics show through in much of its core terminology. The word ‘variation’ itself implies an analyst’s viewpoint, looking down at arrays of variant forms distributed over some spatial matrix. What ‘varies’ is the community’s or the speaker’s language system; more locally, what ‘vary’ are sociolinguistic variables (linguistic units of variable production) defined in the system. This organisation isn’t accessible to, or even directly relevant to, people engaged in speaking and listening, although it is the variationist’s main concern. What matters to people is the meaning that language variation might add to their discursive practices – what people are trying to mean and what they hear others to be meaning.
Formal category systems and taxonomies used by researchers in many fields of inquiry often imply equivalence between categorised units, along the lines of ‘this item is one of this type and goes here, and that item is one of that type and goes there’. All research that is based on coding and counting will make assumptions of this sort, and variationist sociolinguistics does this too in some respects. Variant forms of sociolinguistic variables are defined as being equivalent in their referential meanings. In the (English) Birmingham example, the phrase right time has the same linguistic (referential) meaning however it is pronounced, and [aI;] and [ɔI;] are, to that extent, equivalent in their meaning. Whatever the speaker’s accent, the utterance seems to convey the same basic information. But this approach reduces the scope of the term ‘meaning’ and tends to wash out issues of value as they attach to variable language in actual use. When said in a Birmingham accent, the utterance and the speaker might conceivably be held to be less convincing or authoritative, for example. The social meaning of the utterance, depending on how it is phonologically styled, might interconnect in significant ways with other social aspects of the speech event in which it is embedded.
Bridging between survey orientations and practice orientations in the sociolinguistics of variation seems an obvious development, even though the objectives and assumptions of (broadly) Labovian and (broadly) Hymesian sociolinguistics have traditionally been quite separate. But the separation of these two agendas is in many ways artificial. There is a certain oddness in not addressing social interaction as a medium for variation research, in addition to its commitment to social surveying and to reaching generalisations at that level. There is no inherent clash between ‘macro’ and ‘micro’ levels of variation analysis. One important theme in later chapters is that local processes of meaning-making depend on the affordances that socially structured variation in some sense provides, even though we need to be far more precise than this about how levels of analysis inter-relate. Speaking is the basic modality of language, where linguistic meaning potential is realised and where social meanings of different sorts are creatively implemented. If we decide to engage with the idea of social meaning, however we precisely define it, social meaning will not be something separate from the activation and interpretation of meaning in acts of speaking. The term ‘discourse’ (despite the many different senses in which it can be used – see Jaworski and Coupland 2007) is a useful shorthand for this wider concern. The research agenda around style can therefore also be referred to as the analysis of ‘dialect in discourse’.
Quantitative analysis of the distribution of speech variants among groups of speakers is an abstraction away from the social process of speaking and of making meaning in context. It is of course an entirely legitimate research method, suited to its own purposes of generalising about language variation and change. But investigating variation in the context of social interaction is simply looking at language variation in its primary ecosystem of discursive meaning, and it can therefore claim to be a sociolinguistic priority. A more institutional argument is that there should be benefits to any one tradition of sociolinguistic research in reaching out to other traditions. So much of sociolinguistics nowadays is grounded in analyses of discourse and social interaction that, once again, it would be strange for variationism not to move into that arena. This move might allow us to find other, more integrative, sorts of sociolinguistic truth.
It should already be obvious that the term ‘style’ has significant but largely different histories in sociolinguistics and in other fields. In the sociolinguistics of variation, style has been a very limited concept and a peripheral concern. In his overview of variationist sociolinguistic research Jack Chambers writes that ‘style is an important independent variable but it is never the focal point (Chambers 1995: 6). As we will see in Chapter 2, stylistic variation has been treated quantitatively in sociolinguistic surveys in exactly the same way as social (or social class-related) variation is treated. It has been a matter of demonstrating that ‘intra-individual’ variation exists and that the nature of such variation can be explained by some simple principle or other. In this section, in order to gain some perspective, we return to some early non-sociolinguistic treatments of language style. Naturally enough, there are many points of contact and overlap between early sociolinguistic treatments of style and early stylistics. But those early emphases and interests have in fact persisted much longer in variationist sociolinguistics than they have in stylistics itself. Modern stylistics has blended into different forms of discourse analysis, prefiguring some of the general arguments I am making in this book.
The discipline label ‘stylistics’ was popularised in the 1950s, and it came to be thought of as a discrete field of linguistics or applied linguistics. ‘General stylistics’ (Sebeok 1960) was interested in all forms of language text, spoken and written, distinguished from the sub-field of literary stylistics. Early stylistics was dominated by linguistic structuralism, which emphasised the structural properties of texts at different levels of linguistic organisation (phonological, grammatical, lexical, prosodic). It gloried in the technical sophistication of linguistic description, at a time when linguistics was still developing momentum. Stylistics was largely based on taxonomies – lists of language features, levels and functions. For example, a very simple hierarchical analysis of English style was offered by Martin Joos in his strangely titled book, The Five Clocks (1962). The ‘clocks’ were levels of formality in spoken and written English, which Joos labelled ‘frozen’, ‘formal’, ‘consultative’, ‘casual’ and ‘intimate’. It was based on an intuition about degrees of familiarity/intimacy between people which, Joos argued, impacted on communicative style. The detail of how Joos meant these terms to be applied is not particularly important here, but the ‘clocks’ idea endorses a linear scale of ‘formality’. Formality or communicative ‘carefulness’ is assumed to dictate a speaker’s stylistic choices or designs. As we’ll see, this is how Labov came to operationalise sociolinguistic style too.
Roman Jakobson, in a famous lecture delivered in 1958 (Jakobson 1960, reprinted in Weber 1996a), is often credited with giving the first coherent formulation of stylistics. Jakobson’s theme was the relationship between poetics (aesthetic response to language and text) and linguistics. His argument was that the investigation of verbal art or poetics is properly a sub-branch of linguistics. He reached this position by establishing that the poetic function of language, which he defined as ‘the set … towards the MESSAGE, focus on the message itself’ (Jakobson 1996/1960: 15; reprinted in Weber 1996a: 10–35), is a general function of all language use. It is not restricted to poetry and other literary texts. Jakobson argued that, if language always has a poetic function, linguistics must account for it, and that it could and should therefore account for poetry and other artistic forms too. The most original aspect of Jakobson’s paper is his attempt to list all the main functions of language. The poetic function stands alongside the referential function (the cognitive ordering of propositional
© Cambridge University Press