SIL Electronic Book Reviews 2006-005

Introduction

The book under review arose from a workshop at Stanford University and has been hailed to become the standard work on stylistic variation. It consists of an introduction and 16 contributions to the study of style in variation, that is, specifically intra-speaker constraints in sociolinguistic variation. Lists of the volume’s 21 figures, 27 tables, and 17 contributors as well as 32 pages of references and a 4-page index frame the content part of the book. The topic of “Style and Sociolinguistic Variation” is approached from four different angles, represented by the four parts of the volume: “Anthropological approaches,” “Attention paid to speech,” “Audience design and self-identification,” and “Functionally motivated situational variation.” Each part contains one or two feature articles by a prominent proponent of the field, to which commentary articles respond.

Content

The introduction by Rickford and Eckert starts with distinguishing linguistic versus social versus stylistic constraints, and then proceeds to give an historical overview of research in stylistic variation, starting with Labov’s 1966 “The Social Stratification of English in New York City.” Before introducing the individual articles of the volume, they pay particular attention to how Bell’s (1984) focus on audience was broadened by Finegan and Biber (1994) to concentrate on situation, and by Coupland (1980) to include speaker’s identity. Particularly, they bemoan the dwindling interaction between variation research and ethnography of speaking, and claim that “the very definition of style must expand” (p. 6). The introduction concludes with 12 pages of abstracts of the 16 articles of the volume.

Irvine’s feature article “‘Style’ as distinctiveness” starts off the anthropological approach to stylistic variation. She draws parallels between style of speech and lifestyle and insists on the necessity to analyze style within systems, taking ideologies and aesthetics into account. In her investigation of the defining distinctions between style, register, and dialect, she makes use of data from Javanese speech styles, Wolof speech varieties, and German dialects in southern Hungary. She shows how register, originally defined in terms of situations, and dialect, originally defined in terms of persons or groups, cannot be clearly demarcated from each other. Style, by contrast, she takes to involve more subtle degrees of distinctiveness (e.g. within one speaker), while at the same time extending beyond linguistic phenomena. Based on then still ongoing research together with Susan Gal (only partly published in Irvine and Gal 2000), Irvine then introduces the three semiotic processes of iconization, recursivity, and erasure. For each of these, she gives examples from her data. She concludes that investigations of style should treat it as creative process. In such investigations, she has “found it important to place less emphasis on the specific features of a style…and more on the contrasts and relationships between styles” (p. 42).

The commentary by Ervin-Tripp “Variety, style-shifting, and ideology” takes up the three processes of iconization, recursivity, and erasure, and asks to what extent these are ideological. Ervin-Tripp sees language ideology manifested particularly in style acquisition and in code- or style-switching. She then goes on to elaborate on the latter, investigating types of style shifts in monolinguals. Her data is in AAVE and comes from speeches and interviews of two African-American civil rights leaders in the 1960s, Stokely Carmichael and Dick Gregory. Having identified certain linguistic features of style-shifts and shown how the contrasting groups which such style shifts are meant to refer to are far from stereotypical, Ervin-Tripp concludes that style-switching exemplification of recurrent iconization is not simple but “requires realigning a complex array of potential contrasts” (p. 56).

The second feature article “The ethnography of genre in a Mexican market: form, function, variation” by Bauman focuses on genre analysis. Bauman expands the framework of genre from dealing primarily with textual products to focusing on discursive practice. He condenses this understanding of genre analysis into the research question, “How does the generic organization of linguistic means serve as a resource for the accomplishment of social ends in the conduct of social life?” (p. 59). Based on soundscape recordings from the weekly market in San Miguel, Mexico, Bauman investigates the verbal creation of commodity value through shorter “calls” and extended “spiels” (terminology borrowed from Lindenfeld 1990). Apart from their economic function to draw customers, these oral commercials display an additional poetic function, marked by fluency, cohesion, and the participatory involvement of passers-by. Having hinted at an inventory of basic elements and their combinatory possibilities as well as given examples of collaborative performances, Bauman then builds his case for treating calls and spiels as genres (in addition to differentiating contexts of use and categories of users) by highlighting quality differences of advertised products, differences in structural complexity, and diverging functions. He concludes that it would be dangerous to reduce speech style to a single defining dimension or criterion as this “obscures the multiple indexical resonances that accrue to ways of speaking in social life” (p. 77).

In his response to Bauman, Macaulay, as indicated by the title of his article, raises “The question of genre.” He sees a number of problems with Bauman’s approach: first, Macaulay questions whether the identified speech elements are useful as analytical categories. In particular, he questions the comparability of the examples (18 calls versus only one spiel), not least in light of one example of a hybrid category. Second, there is the problem that an anthropologist’s taxonomies may not match the perception of the local people. Third, it is far from easy to identify speech events which are equivalent, thus forming a valid basis for comparative analysis. However, Macaulay does commend Bauman for having demonstrated how to base a study of style on a previous subcategorization of speech genres, as claimed to be necessary by Bakhtin (1986).

Part 2 “Attention paid to speech” is spearheaded by its only feature article, “The anatomy of style-shifting” by Labov. In it, Labov starts by differentiating two approaches to style shifting, namely, taking it as a naturalistic phenomenon versus using it as a controlled measuring device. For the former, the problem occurs how to obtain representative data of group interactions with varying interlocutors. Also, Labov addresses the problem of how to disentangle effects of audience design versus those of audio-monitoring. He finds that contexts of diminished audio-monitoring come closest to vernacular speech. He then reports on the “Decision Tree,” used to categorize and analyze spontaneous speech, consisting of the following eight categories: response, narrative, language, group, soapbox, kids, tangent, and residual. Of these, response, language, soapbox, and residual exhibit careful speech, while narrative, group, kids and tangent exhibit casual speech. Labov then launches into a detailed, and partly highly statistical, description of a Philadelphia neighborhood study, focusing especially on the features DH (alternation of stops, fricatives, and affricates in initial position), ING (alternation of apical and velar consonants for the unstressed syllable /iN/), and NEG (the percent of negative concord preceding an indeterminate). In the analysis, the factors of age, gender, and social class are controlled. As an astonishing result of the study, all eight categories of the Decision Tree came out on a par. While originally hoping to rank the eight categories with regard to their stylistic importance, Labov had to conclude that this was not possible, thus corroborating the equal importance of all eight.

The first of three replies to Labov is Baugh’s “A dissection of style-shifting.” Baugh starts with a brief historical overview of what he calls the four stylistic trends in Labov’s work. His major discussion focuses on field methods for which he strongly recommends reading Labov (1984). As possible factors worthwhile to look out for, Baugh mentions literacy barriers to and cultural influences on data collection, and specifically the interaction between a fieldworker’s own linguistic variation and the data to be investigated (a particular variant of the observer’s paradox). With regard to Labov’s Decision Tree, Baugh sees the need to expand it, as the figures for the residual category of “Careful Speech” seem to indicate further differentiation between careful and casual speech. Finally, Baugh discusses a possible application of stylistic analysis to teacher testing where “pathological speech tests…tend to over-identify African Americans and Latinos as having abnormal language development, when problems lie within the testing instruments themselves” (p. 118).

In “Style and social meaning,” Eckert tries to connect the constructed world of Labov’s sociolinguistic interview and the larger stylistic world. In particular, she looks at the potential interaction between subevents of the Decision Tree and notices that different parameters of standard versus vernacular speech can be found within the same subevent. Moreover, she argues that the differences between Labov’s parameters (DH, ING, and NEG) are not necessarily comparable in their origin, e.g. whereas ING correlates with formality, NEG is influenced by stigmatization. Eckert goes on to discuss aspects of consciousness of style as well as the stylistic construction of a self. Drawing examples from fieldwork in a Detroit suburban high school and among pre-adolescents in Northern California, she notices a certain foregrounding of stylistic variables in the construction of stylistic icons. Such foregrounding results in highlighted meaning, and it is this negotiation of social meaning that Eckert sees in need of being paid attention to for meaningful research in stylistic variation.

The final reply to Labov’s article is Traugott’s “Zeroing in on multifunctionality and style.” Traugott comes from an historical linguistic perspective which traditionally focuses more on the individual as well as on the intersection between structure and use. In her ensuing discussion of grammaticalization as creative language use, she notices the relevance of distinguishing between open- and closed-class items (as done in Prince’s 1987 study of one individual’s Yiddish). For example, concerning Labov’s variable DH, “the voiced fricative /ð/ might have different effects from other fricatives and affricates considering its status not only as a stem, but also as an almost unique signal of definiteness and deixis” (p. 129). As a consequence, Traugott advocates that style research take into account different linguistic functions like socio-physical, epistemic, and speech act, and in order to emphasize her point, she provides examples for the different uses of the four adverbs obviously, in fact, actually, and anyway. She finds that the frequency of different uses of the same item has implications for style-shift and, consequently, that “[s]imply counting tokens of the word is not enough” (p. 134). In conclusion, Traugott demands that attention be paid to the difference between lexical and grammatical items as well as to “design” and not only to speech, thus building a bridge to the next part of the book.

Alan Bell, the main proponent of part 3’s theme “Audience design and self-identification,” starts off this part with his feature article “Back in style: reworking audience design.” He begins with his definition of sociolinguistics as looking for explanations of regularities or patterns in the linguistic choices of a situated speaker. Based on his doctoral research into style-shift of New Zealand newsreaders, Bell’s theory of “audience design” (Bell 1984) basically claims that “style is oriented to people rather than to mechanisms or functions” (p. 141). After giving a detailed overview of the theory, Bell reports on a project involving interviews in New Zealand which were specifically designed to test audience design hypotheses (only ethnicity and gender were varied with all other factors being held as constant as possible). The results confirm audience design’s tenets that language usage either functions as identity marker or indicates sensitivity to outgroup members. In his conclusion, Bell attempts an integration of audience design (focusing quantitavely on patterns in interpersonal linguistic behavior) and referee design (focusing qualitatively on the linguistic choices of a speaker how to represent his self). In order to achieve that, Bell advocates a 3-layer approach to stylistic research, adding a co-occurrence approach (Biber and Finegan 1994) to the two design approaches.

In her reply which addresses both the preceding article by Bell and the following article by Coupland, Yaeger-Dror investigates “Primitives of a system for ‘style’ and ‘register’.” Based on data on prosodic variation of negation in English, she gives examples of quantitatively measurable instances of referee design (contrary to Bell’s claim), and simultaneously refutes Coupland’s simplification of design distinction. Particularly, Yaeger-Dror claims that “both audience design and referee design are quantitatively measurable, but whether a specific case of variation is one or the other must be determined qualitatively” (p. 172). She then goes on to discuss additional parameters to be considered, like topic, purpose, planning, setting, and performance frame, as well as other factors which might influence the identification of style primitives, like gender, ethnicity, power, age, solidarity, and “face.” In conclusion, while affirming her above claim, she concedes that a definite categorization of the additionally mentioned parameters into primitives over against only “metaphorical” still awaits further investigation.

The second feature article of part 3 is Coupland’s “Language, situation, and the relational self: theorizing dialect-style in sociolinguistics.” His main tenet is that “[s]tyle needs to be located within a model of human communicative purposes, practices and achievements, and as one aspect of the manipulation of semiotic resources in social contexts” (p. 186). He thus aims at avoiding both extremes of either overgeneralizing style as “everything” or narrowing it down to a single quantifiable sociolinguistic parameter. After presenting a list of ten traditional sociolinguistic assumptions which he explicitly challenges, Coupland turns to those aspects which previous sociolinguistic approaches purportedly ignored, like communicative goals, functional complexity, issues of identity and selfhood, or attitudinal style, and Hymesian (1974) “ways of speaking,” which he equates with patterns of ideational selection in a Hallidayan sense. In his theorizing, among other aspects, Coupland refutes the illusion of nonmotivated style-shifting and criticizes calculation across different dimensions, especially the lack of differentiation between situational and social categories which consequently obscures local stylistic variation. He also makes extensive reference to Bakhtin, e.g. the conflict-orientation of styles. Coupland especially advocates a focus on people who use language rather than on texts as language itself; for him, stylistic variation is person variation. After propounding further his multidimensional view, which includes borrowing from literary criticism and defining the relational self as a continually re-created autobiography, Coupland gives a brief illustration from a case-study of a Cardiff disc jockey’s (DJ) speech. Under the assumption that the narrative of self-identity is inherently fragile given the challenges of “high” modernity, Coupland concludes with the claim that any data is more complex than their analyses make us believe.

With “Couplandia and beyond,” Giles replies to Coupland from a sociopsychological perspective. He counters Coupland’s criticism with examples of more “subjective” dimensions in his own style research. Giles then goes on to caution against focusing too much on individual style shifts and recommends rather to take into account a speaker’s skill of targeting various audiences simultaneously (especially relevant in Coupland’s DJ data). Also, in Giles’ opinion, a style shift has to be perceived as such by the addressee/listener in order to count as style shift. Defending his own communication accommodation theory (CAT), Giles reasons that convergence is often partial rather than full in order to remain plausible, and that group images and presentations need to be distinguished from individual ones. As example, he mentions transitional shifts between authority and empathy in police officers’ bidialectal style usage. In conclusion, he adds a number of fundamental questions to Bell’s sociolinguistic research definition (see above), including: “Can we not agree to cohese CAT (and its satellite models) with relevant aspects of audience and referee design so as to avoid cross-disciplinary redundancies?” (p. 219).

A second reply to Coupland comes from Rickford in “Style and stylizing from the perspective of a non-autonomous sociolinguistics.” While agreeing with Coupland’s criticism of traditional sociolinguistic assumptions, Rickford calls for the need to move from description to explanation, and also cautions that the existing (and possibly necessary) differences in conception preclude a seemingly demanded unity of direction in sociolinguistic research. Particularly, he counters Coupland’s criticism of unidimensional approaches and admonishes sociolinguists to identify primary dimensions on which to build their models, rather than secondary or tertiary. Still, he commends communication science for its distinction of objectives into instrumental (or task), relational (or inter-personal), and identity (or self-presentational), first made by Clark and Delia (1979). Giving examples from the African American worship and Bible reading tradition as well as from a Guyanese DJ, Rickford raises questions of style-reading and style-perception by different audiences, concluding that there is a place for self-report of performers.

The last and fourth part of the book is spearheaded by Finegan and Biber’s [F&B] “Register variation and social dialect variation: the Register Axiom.” They claim that neither Bell’s model of dialect (or inter-speaker) variation nor Preston’s model of linguistic (or internal) variation can explain the observed patterns. As a model which is also explanatorily powerful, they propose the “Register Axiom” in which register (or intra-speaker) variation is regarded as primary. In this, they build explicitly on Kroch’s 1978 theory of social dialect variation. Centering their definition of register on situations of use, F&B state their hypothesis that “social dialect patterns…derive from a combination of (a) their communicatively motivated distributions across registers; and (b) the differential access of social groups to praxis in those registers” (p. 244). They then investigate the “classic sociolinguistic finding” about the distribution of linguistic features across social groups and situations along the lines of economy, exemplified by the features contraction, THAT omission, pro-verb DO, and IT, and elaboration, represented by the features attributive adjectives, prepositional phrases, and type/token ratio. These seven features, with their percentages calculated from various large English corpora, are then plotted against three parameters of functional variation, viz. planning, purpose, and shared context. The results support F&B’s claim that the mandate of ease is more frequent with lower status, and the mandate of clarity with higher status. After an overview of further explanatory support for their model from the literature, they give evidence which shows that registers and social dialects do not concur with their respective feature distribution, and consequently casts doubt on Bell’s Style Axiom. In conclusion, they present their Register Axiom again which states that the frequency of use in a social group depends on that group’s access to registers which favor that use.

The first reply to F&B is Milroy’s “Conversation, spoken language, and social identity.” Apart from finding overarching models problematic in general, Milroy takes issue with a number of F&B’s theoretical constructs and distinctions. After criticizing their use of the term “function,” she shows how features which occur across registers like IT or THAT omission need to be subcategorized. Milroy also insists on taking historical developments into account; e.g. features like double negative once belonged to literary registers. She finds the notions of elaboration and economy and their application to social dialect variation particularly problematic, giving examples from different studies of English. She then repeats the same for F&B’s mandates of ease and clarity. Her recommendation is to at least introduce the distinction between dialogic (or spoken) versus monologic (or written) into the model, even though, “in view of the heterogeneous character and mixed conceptual heritage of much of [F&B’s] work, I very much doubt whether a unitary model of the kind they propose is appropriate” (p. 278).

The second reply of this part and at the same time concluding paper of the entire volume is Preston’s “Style and the psycholinguistics of sociolinguistics: the logical problem of language variation,” in which he compares F&B’s model with several others, including his own. In his attempt to answer the question, “Why do Bell and I discover one set of relationships in surveying (and conducting) empirical work and F&B another?” (p. 281), Preston points to differences in the definition of “style” and reminds the reader that the factors chosen for investigation predetermine both results and analysis. Schematizing F&B’s versus Bell’s model (into which he incorporates models by Labov and himself) in comparative diagrams, Preston finds that F&B’s notion of “economy” is difficult to determine and that a correlation of linguistic features to text types is dubious methodologically. Focusing then on a comparison of F&B and Labov, Preston finds fault with assumptions about certain social groups’ incomplete control over their vernaculars, and he also calls for taking the influence of literate forms into account more directly. He then goes on to criticize F&B’s choice of features as not being “variables” as they do not offer alternatives, like an investigation of imperative versus nonimperative forms in direction giving would have done. After discussing three ways of determining variation within one grammar or between different grammars, Preston attempts to integrate all concerns into one figure and concludes with conciliatory notes on the metaphor of the blind scholars and the elephant.

Evaluation