Corpora and Discourse
Studies in Corpus Linguistics (SCL) SCL focuses on the use of corpora throughout language study, the development of a quantitative approach to linguistics, the design and use of new tools for processing language texts, and the theoretical implications of a data-rich discipline.
General Editor
Consulting Editor
Elena Tognini-Bonelli
Wolfgang Teubert
The Tuscan Word Center/ The University of Siena
Advisory Board Michael Barlow
Graeme Kennedy
Douglas Biber
Geoffrey N. Leech
Marina Bondi
Michaela Mahlberg
Christopher S. Butler
Anna Mauranen
Sylviane Granger
Ute Römer
M.A.K. Halliday
Jan Svartvik
Yang Huizhong
John M. Swales
Susan Hunston
Martin Warren
University of Auckland Northern Arizona University University of Modena and Reggio Emilia University of Wales, Swansea University of Louvain University of Sydney Jiao Tong University, Shanghai University of Birmingham
Victoria University of Wellington University of Lancaster University of Liverpool University of Helsinki University of Hannover University of Lund University of Michigan The Hong Kong Polytechnic University
Stig Johansson
Oslo University
Volume 31 Corpora and Discourse. The challenges of different settings Edited by Annelie Ädel and Randi Reppen
Corpora and Discourse The challenges of different settings
Edited by
Annelie Ädel University of Michigan, USA
Randi Reppen Northern Arizona University, USA
John Benjamins Publishing Company Amsterdam / Philadelphia
8
TM
The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.
Library of Congress Cataloging-in-Publication Data Corpora and discourse : the challenges of different settings / edited by Annelie Adel and Randi Reppen. p. cm. (Studies in Corpus Linguistics, issn 1388-0373 ; v. 31) Includes bibliographical references and index. 1. Discourse analysis--Data processing. 2. Corpora (Linguistics) I. Ädel, Annelie. II. Reppen, Randi. P302.3.C6683 2008 401'.410285--dc22 isbn 978 90 272 2305 0 (Hb; alk. paper)
2008006978
© 2008 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa
Table of contents
1. The challenges of different settings: An overview Annelie Ädel and Randi Reppen
1
Section I Exploring discourse in academic settings 2. ‘...post-colonialism, multi-culturalism, structuralism, feminism, post-modernism and so on and so forth’: A comparative analysis of vague category markers in academic discourse Steve Walsh, Anne O’Keeffe and Michael McCarthy 3. Emphatics in academic discourse: Integrating corpus and discourse tools in the study of cross-disciplinary variation Marina Bondi 4. Interaction, identity and culture in academic writing: The case of German, British and American academics in the humanities Tamsin Sanderson
9
31
57
Section II Exploring discourse in workplace settings 5. “Got a date or something?”: An analysis of the role of humour and laughter in the workplace meetings of English language teachers Elaine Vaughan 6. Determining discourse-based moves in professional reports Lynne Flowerdew 7. //→ ONE country two SYStems //: The discourse intonation patterns of word associations Winnie Cheng and Martin Warren
95 117
135
Section III Exploring discourse in news and entertainment 8. Who’s speaking?: Evidentiality in US newspapers during the 2004 presidential campaign Gregory Garretson and Annelie Ädel
157
Corpora and Discourse
9. Television dialogue and natural conversation: Linguistic similarities and functional differences Paulo Quaglio 10. A corpus approach to discursive constructions of a hip-hop identity Kristy Beers Fägersten
189 211
Section IV Exploring discourse through specific linguistic features 11. The use of the it-cleft construction in 19th-century English Christine Johansson 12. Place and time adverbials in native and non-native English student writing William J. Crawford Author index Corpus and tools index Subject index
243
267
289 291 293
JB[v.20020404] Prn:11/04/2008; 10:36
F: SCL3101.tex / p.1 (47-111)
The challenges of different settings An overview Annelie Ädel and Randi Reppen
Corpus-linguistic studies of discourse Corpus linguistics has, over the past few decades, undergone a transformation from a “little donkey cart” to a “bandwagon” (Leech 1991: 25), and is now at a point at which it “is becoming part of mainstream linguistics” (Mukherjee 2004: 118). Mainstream linguistics, however, is very broad and multifaceted, and some subfields are more amenable to corpus-linguistic methodology than others. If we disregard some basic research issues, such as access to a suitable corpus that gives a reasonably representative sample of the population studied, there are certain generalizations we can make about the compatibility of corpus-based methods with the research questions posed in different linguistic subfields. For example, while lexicographers are often able to use corpus-assisted methods in answering their particular questions about language in relatively straightforward ways, discourse analysts – whether working with speech or writing – are likely to spend a great deal of time finding possible solutions for computerizing their methods. Discourse phenomena, with their frequent dependence on and sensitivity to context, co-text, and interpretation, require rather complex solutions and often a great deal of intervention on the part of the researcher. Despite the potential difficulties of automatizing data retrieval and analysis, researchers interested in discourse have started to adopt corpus-linguistic methods more seriously over the past few years – a trend to which the current volume bears witness. This is, however, a very recent development. At the end of the twentieth century, Biber et al. (1998: 106) described the state of the art as follows: “although nearly all discourse studies are based on analysis of actual texts, they are not typically corpus-based investigations: most studies do not use quantitative methods to describe the extent to which different discourse structures are used, and relatively few of these studies aim to produce generalizable findings that hold across texts.” Two other textbooks on corpus linguistics published around the same time –
JB[v.20020404] Prn:11/04/2008; 10:36
F: SCL3101.tex / p.2 (111-170)
Annelie Ädel and Randi Reppen
McEnery & Wilson (1996) and Kennedy (1998) – both point to the comparatively marginal application of corpus-linguistic methods in discourse studies. However, a couple of years into the new century a slightly different picture of the compatibility of computer-assisted methods with discourse-level phenomena was presented. Comparing the state of the art in 2002 to the early days of corpus linguistics, Conrad (2002: 86) gives a positive characterisation, stating that, “[a]s corpus linguistics first developed, it was often thought that it could not be applied to language phenomena that extended beyond clause boundaries. As the field has matured, it has instead become apparent that many studies within corpus linguistics address discourse-level concerns, many showing association patterns or the interactions of variables that would not be apparent without corpus-based techniques.” At this point in time, we are happy to be able to say that things really are changing. For readers who wish to explore why this might be, Partington (2004) offers a summary of explanations (such as the widespread inclusion of text extracts rather than full texts in standard corpora) for the historically slight application of corpuslinguistic methods in studies of text and discourse. As a demonstration of recent shifts in this area, the present volume brings together researchers from diverse areas of text and discourse, all of whom demonstrate the viability of corpus-based research and corpus-assisted tools for discourse studies.
Finding discourse-relevant data It is interesting to consider the search methods used by the different researchers in this volume to locate linguistic forms in a corpus – usually, in the case of discourse analysis, forms that are linked to a particular function. We believe that a description of commonly used retrieval methods can help others in reflecting on their own studies and the options available to them. Four main methods were used by the authors of these chapters, which we believe to be representative of the field. The most typical search method can be called one-to-one searching, which involves investigating a linguistic form through a search term that only yields relevant hits. A good example of this is Crawford’s time and place adverbs here and now in Chapter 12, where there are no spurious hits, and the entire set that the researcher intends to examine is captured. To use more technical vocabulary, precision and recall are both at 100%. The ease of capturing relevant examples, however, does not necessarily mean that no more work remains for the researcher, who will often go on to examine the different discourse functions or semantic distinctions of the search term in question. Other search methods, however, need to be used when there is not a simple one-to-one mapping between a search term and the body of relevant hits in a corpus. To mention just a couple of complicating factors familiar to all linguists,
JB[v.20020404] Prn:11/04/2008; 10:36
F: SCL3101.tex / p.3 (170-233)
The challenges of different settings
individual linguistic forms can be polysemous, while specific functions of language (such as politeness) can be realized by many different linguistic forms. The second search method can be called sampling (Ädel 2003). It involves the use of one or more search terms that are good examples of the linguistic phenomenon in question. The drawback is that not all instances of the phenomenon, but only a subset, will be captured, although one advantage is that the search terms used tend to yield a high number of relevant hits. When using this method, the researcher cannot claim to have covered all bases or to have mapped out a linguistic function in its entirety, but many valuable insights can still be provided, especially if the search term is a good indicator of the phenomenon under study. Chapter 5 provides a good example of sampling, with Vaughan being able to draw interesting conclusions about the role of humour in the workplace based on occurrences of laughter. Vaughan uses occurrences of laughter, indicated in the transcriptions, as a “proxy” (cf. Garretson & O’Connor 2007: 89) for humour. The third search method can be called sifting (Ädel 2003), since once the initial hits have been retrieved, they need to be sifted through – meaning that a certain proportion will be manually discarded. Using this method, the researcher often needs to put a great deal of time into checking the retrieved data (before the actual analysis can begin). The advantage of this method tends to be that, once the sifting has been done, the remaining set covers all or most of the potential forms of the linguistic phenomenon one is looking for. An example of this method is found in Chapter 9, where Quaglio uses an extensive inventory of linguistic forms that tend to be associated with face-to-face conversation. A small subset of these includes so and really used as informal intensifiers (but crucially, not anaphoric so and not really as a news recipient). Although this is part of a multi-dimensional analysis (Biber 1988) that both finds and interprets the co-occurrence of a selection of linguistic features, some of the forms involved can still be said to be retrieved by sifting. The fourth and final method can be called frequency-based listing. It involves the use of a frequency list (of individual words or collocations), specifically based on the corpus under investigation, as a starting point. Using such a list, the researcher goes on to select the relevant search terms that occur with high frequency. This way, the search terms will be tailor-made for the corpus and the particular discourse studied. It is an effective way of using corpus-assisted methods to spot persistent patterns in a specific dataset. A nice example of this method is found in Chapter 2, where Walsh, O’Keeffe & McCarthy are able to identify exactly which expressions of vagueness to focus on based on a frequency list of multi-word clusters. Having identified the relevant expressions, they can go on to concordance and analyze them. Of course, we live in an increasingly hybridized world, and it would probably be foolish to expect to find only pure examples of each method. Two or more of
JB[v.20020404] Prn:11/04/2008; 10:36
F: SCL3101.tex / p.4 (233-295)
Annelie Ädel and Randi Reppen
these search methods are sometimes combined. The study by Garretson & Ädel reported in Chapter 8, for example, uses both sampling and sifting. Sampling is the overall method: by listing what they call “reporting words” (e.g. the verb lemma STATE, the noun statement, and the phrase according to), they attempt to capture instances of hearsay evidentiality in their data. Sifting is employed when individual words in the list are ambiguous or polysemous, as in the case of states – a highly frequent string in the US newspaper data. The analyst is required to retain examples like the association states that misconceptions continue to affect law and reject examples like two dozen states that allow early voting, either by manual elimination or through complex computational algorithms. Any automatic or semi-automatic corpus-based method is restricted to considering surface realizations (whether actual linguistic forms, or units identified by annotation) – and herein lies the challenge for studies of functional categories. The present volume offers many interesting examples of how this challenge can be met.
Overview of the chapters Rather than organizing the book according to the different methods researchers used for analyses, we chose as the main organizing principle the different contexts of language use. One of the main strengths of this book is its exploration of discourse in various settings, covering discourse in academia, in the workplace, in news and entertainment. Thus, the four sections of the book primarily reflect the different settings of the discourses analyzed. The theme of the first section is “Exploring discourse in academic settings”. The section begins with Walsh, O’Keeffe, and McCarthy taking a close look at the use of vague language in a range of speech events recorded at universities in the Republic of Ireland and Northern Ireland. The chapter brings to light some interesting uses of vague language and how the use of vagueness varies depending on the discourse context. The next two chapters focus on language in academic journals. First, Bondi examines stance and engagement as realized through keyword adverbs in a corpus of English-language journal articles in history and economics. A selection of the adverbials (significantly, undoubtedly and invariably) is studied more closely, from the perspective of collocation and patterns of semantic preference as well as pragmatic and textual functions. Next, Sanderson looks at journal articles drawn from five different disciplines in the humanities and written in German, American English and British English, focusing on the use of pronouns that mark interactivity between writer and reader. Various types of sociological information about the authors were encoded, which enabled her to check the relative influence of variables such as linguistic background, discipline, age, and gender.
JB[v.20020404] Prn:11/04/2008; 10:36
F: SCL3101.tex / p.5 (295-341)
The challenges of different settings
The theme of the second section is “Exploring discourse in workplace settings”. This section examines language in the workplace, both the contexts of business and public reports, and the context of professional meetings. The section begins with Vaughan’s in-depth look at the roles humor plays in institutional interactions of teacher meetings. Using a corpus from two different settings of teacher meetings recorded in Mexico and in Ireland, Vaughan discovers interesting patterns in the use of laughter. The following two chapters explore a variety of aspects of the use of English in Hong Kong. Using a small, specialized corpus of professional reports, Flowerdew analyzes discourse moves, focusing particularly on problemsolution patterns. She also examines a couple of keywords (the lemmas problem and impact) and how they co-pattern with structural units in the texts. In the next chapter, Cheng and Warren end the second section with “a first attempt at examining the relationship between the phraseological characteristics of language and the communicative role of discourse intonation”. They present an innovative investigation of patterns of discourse intonation in frequent three- and four-word combinations based on a corpus of spoken English in Hong Kong. The theme of the third section is “Exploring discourse in news and entertainment”. This section exhibits the greatest diversity of genres, including newspaper reports, a television series, and internet-based discussion boards on hip-hop. As diverse as the genres, so are the techniques used to examine discourse. In Chapter 8, Garretson and Ädel tackle the highly political issue of how hearsay evidentiality is reported in news articles related to the 2004 US presidential election. In a detailed look at how campaign language is reported and attributed, they lead the reader through unexpected insights into how different newspapers report the speech of different individual and collective entities. The next chapter takes us from the serious world of reporting presidential campaigns to a popular American situation comedy, Friends. In Chapter 9, Quaglio provides a detailed linguistic investigation of Friends, comparing it to a large corpus of natural conversation. It is a data-driven investigation which combines multidimensional methodology with a frequency-based analysis of a large number of linguistic features associated with the typical characteristics of face-to-face conversation. Quaglio indicates how the language of this television show may prove to be a resource for ESL and EFL teachers. The section concludes by moving from the language of television to the internet postings of hip-hop fans. In Chapter 10, Beers Fägersten carefully examines how identity is constructed in the virtual environment of message board postings. She guides the reader through the linguistic construction of identity – through the use of specific openings and closings, slang and taboo terms, and “verbal art” – in this highly specialized use of language. The theme of the fourth and final section is “Exploring discourse through specific linguistic features”. Johansson traces the uses of it-clefts diachronically. Using several corpora of diachronic and present-day English, she looks across
JB[v.20020404] Prn:11/04/2008; 10:36
F: SCL3101.tex / p.6 (341-411)
Annelie Ädel and Randi Reppen
several different registers to reveal how the use of it-clefts has changed over time. The greatest frequency and the greatest number of variations on the prototypical it-cleft pattern are found in manuscripts from trials, where the functions of identifying and clarifying are shown to be important, especially to verify the identification of a person, thing or place. In the final chapter, Crawford analyzes the time and place adverbs here, there, now and then in three corpora of learner writing in English and compares that with corpora of English speech and writing produced by native speakers. The adverbs are analysed quantitatively and qualitatively in order to test the hypothesis that the learner writers’ language use is closer to that of native-speaker speech rather than native-speaker writing. Although the investigations represented in this book are quite narrowly focused on English, the reader will learn a great deal about different varieties of English, for example diachronic, international, learner, and non-standard varieties. Not only does this volume offer a rich sample of the spoken and written discourse around the world that takes place in English – with the interesting exceptions of references to German in Chapter 4 – but it also offers a range of topics and methods. The different approaches to the use of corpora are as diverse as the topics investigated. It is our hope that this will encourage other researchers to continue to use corpora in new ways, addressing questions in ways that were previously difficult to imagine.
References Ädel, A. 2003. The Use of Metadiscourse in Argumentative Writing by Advanced Learners and Native Speakers of English. PhD dissertation, University of Göteborg. Biber, D. 1988. Variation across speech and writing. Cambridge: CUP. Biber, D., Conrad, S. & Reppen, R. 1998. Corpus Linguistics: Investigating Language Structure and Use. Cambridge: CUP. Conrad, S. 2002. Corpus linguistic approaches for discourse analysis. Annual Review of Applied Linguistics 22: 75–95. Garretson, G. & O’Connor, M. C. 2007. Between the humanist and the modernist: Semiautomated analysis of linguistic corpora. In Corpus Linguistics Beyond the Word: Corpus research from phrase to discourse, E. Fitzpatrick (ed.), Amsterdam: Rodopi. Kennedy, G. 1998. An Introduction to Corpus Linguistics. London: Longman. Leech, G. 1991. The state of the art in corpus linguistics. In English Corpus Linguistics. Studies in honour of Jan Svartvik, K. Aijmer & B. Altenberg (eds), 8–29. London: Longman. McEnery, T. & Wilson, A. 1996. Corpus Linguistics. Edinburgh: EUP. Mukherjee, J. 2004. The state of the art in corpus linguistics: Three book-length perspectives. English Language and Linguistics 8(1): 103–119. Partington, A. 2004. Corpora and discourse, a most congruous beast. In Corpora and Discourse [Linguistic Insights: Studies in Language and Communication 9], A. Partington, J. Morley & L. Haarman (eds), 11–20. Frankfurt: Peter Lang.
JB[v.20020404] Prn:8/02/2008; 12:22
F: SCL31P1.tex / p.1 (61-88)
Exploring discourse in academic settings
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.1 (48-136)
‘...post-colonialism, multi-culturalism, structuralism, feminism, post-modernism and so on and so forth’ A comparative analysis of vague category markers in academic discourse Steve Walsh, Anne O’Keeffe and Michael McCarthy Newcastle University, UK / Mary Immaculate College, University of Limerick, Ireland / University of Nottingham, UK
The use of vague language is one of the most common features of everyday spoken English. Speakers regularly use vague expressions to project shared knowledge (e.g., pens, books, and that sort of thing) as well as to make approximations (e.g. around sevenish; he’s sort of tall). Research shows that many of the most common single word items in a core vocabulary form part of vague language fixed expressions (e.g. thing in that kind of thing). This chapter will address the use of vague language in a new corpus of academic English, the Limerick-Belfast Corpus of Academic Spoken English (LIBEL CASE). The LIBEL corpus consists of one million words of spoken data collected in two universities on the island of Ireland, one in the Republic of Ireland and one in Northern Ireland. Analysis of the LIBEL corpus identified forms and functions of vague language in an academic context and these findings are compared with two corpora of everyday spoken language from the Republic of Ireland and the United Kingdom, the Limerick Corpus of Irish English (LCIE) and the Cambridge and Nottingham Corpus of Discourse in English (CANCODE). Cross-corpora comparison allowed us to look at how forms and frequencies of certain vague language expressions vary across casual and formal/institutional contexts. Within the academic data we build on Walsh’s work (see for example Walsh 2002, 2006) to show how vague language use is relative to mode of discourse at any given stage of classroom interaction. We suggest that these qualitative differences are a valuable means of understanding the complex relationship between language and learning.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.2 (136-164)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
.
Introduction: Vague categories
The use of vague language is one of the most common features of everyday spoken English. Speakers regularly use vague expressions to project shared knowledge (e.g., pens, books, and that sort of thing) as well as to make approximations (e.g. around sevenish; he’s sort of tall). Research shows that many of the most common single word items in a core vocabulary form part of vague language fixed expressions (e.g. thing in that kind of thing). Carter and McCarthy (2002), who looked at five million words of spoken British English data, show that vague language items are among the core vocabulary items (see also O’Keeffe et al. 2007). Multi-word units which mark vagueness, such as and things like that, that sort of thing, occurred with greater frequency than many single word items. Degrees of variation exist in how vague language is defined. Channell (1994) restricts it to ‘purposefully and unabashedly vague’ uses of languages while Franken (1997) distinguishes between ‘vagueness’ and ‘approximation’. Zhang (1998) makes a case for four separate categories: ‘fuzziness’, ‘generality’, ‘vagueness’ and ‘ambiguity’. Chafe (1982) puts vagueness and hedging in the same category of ‘fuzziness’ – all of which are seen as ‘involvement devices’ more prevalent in spoken rather than written language. The notion of vagueness as an involvement device is consistent with the view that vague language is a core feature of the grammar of spoken language (Carter & McCarthy 1995, 2006; McCarthy & Carter 1995; O’Keeffe et al. 2007). As Carter and McCarthy (2006) note, vague language is a strong indicator of assumed shared knowledge which marks in-group membership insofar as the referents of vague expressions can be assumed to be known by the listener. This is consistent with Cutting (2000), who illustrates how discourse communities use vague language as a marker of in-group membership. The interactive aspect of vague language is important to our focus in this chapter where we examine the use of vague language in the learning context of university discourse. In this domain, the use of vague language is part of meaning making within specific learning contexts or modes (see Walsh 2006: 111). We will focus on one type of vague language, namely vague category markers (hereafter VCMs). These non-lexicalised categories are created within interactions, at the moment of speaking. The categories contain exemplars followed by a vagueness tag (and so on, and that kind of thing, et cetera, and things like that) and the listener(s) is/are expected and assumed to fill in, or implicitly understand the reference. The example in Extract 1 is taken from a drama lecture in the Limerick Belfast Corpus of Academic Spoken English (LIBEL CASE;1 see details in Sections 3 and 4): . Hereafter, LIBEL CASE will be shortened to LIBEL.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.3 (164-217)
A comparative analysis of vague category markers
Extract 1 . . . And I suppose my understanding of critical theory and critical aah critical studies I suppose as such emanate from or are the key social critiques of our time which have emanated from the work of the Frankfurt School. So at the moment it’s you know ahh critical theory is dominated by ideas of postcolonialism multi-culturalism structuralism feminism post-modernism and so on and so forth. Here the exemplars are post-colonialism, multi-culturalism, structuralism, feminism, post-modernism, and the tag which creates the VCM is and so on so forth. Extract 2 is an example from a corpus of casual conversation (the Limerick Corpus of Irish English, LCIE, see below) between friends who are chatting. Speaker (1) creates a VCM but the listener does not understand the exemplar. Hence the category is not created and needs further explanation. In the process of explanation, another VCM is created: Extract 2 (see Appendix for transcription codes) <$6>: He just made up words like he just made up I don’t know what. <$1>: Is that not artistic license like? amm coinage and stuff like that? <$6>: What? <$1>: Coinage. <$6>: What’s coinage? <$1>: When you are writing poetry and stuff you can make up your own words. <$4>: Yeah I mean yeah. <$6>: Like say sarcasamistic like? <$1>: Yeah
you are a poet and you don’t know it my friend? <$5>: Ah snozberry. <$1>: Yeah. <$4>: Fantastic. This is a good example of how meaning is negotiated interactively within a conversation. While the first VCM which speaker (1) uses over-extends the range of assumed shared knowledge between the speakers by using the exemplar coinage, the second VCM uses a much more general exemplar, poetry, which is obviously within the range of shared knowledge of the group.
. Previous research into vague categories Vague categories can be divided into lexicalised and non-lexicalised types. Lexicalised categories are those which provide superordinates or prototypes encoded as a single, lexical item, for example bird, furniture, machinery. Until recently,
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.4 (217-270)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
most research into the nature of categories has been concerned with these lexicalised categories within the field of semantics; see in particular the work of Rosch and her associates (Mervis & Rosch 1981; Rosch 1978; Rosch et al. 1976), who demonstrated that the categories they studied had a graded structure and that at the centre of each category was a prototype that exhibited the highest concentration of characteristic properties compared with members at the periphery which contained fewest characteristic properties. Non-lexical categories are ad hoc rather than prototypical. The concept is attributed to the work of Barsalou (1983, 1987), though links may be seen in the work of Cruse (1986) on what he called lax hyponymy (the non-institutionalised arrangements of items into instantial categories at the time of speaking). The question as to whether categories are stable or subject to change is addressed in particular by Barsalou (1983, 1987), who talks about the dynamic nature of ad hoc category formation, for example places to look for antique desks. In such examples, categorisation is non-lexicalised and without clear boundary, challenging the notion that categories are stable, easily recognisable and arrived at ‘pre-textually’ (after Overstreet & Yule 1997a). Overstreet and Yule (1997a: 85–86) reflect that: If only common (i.e. lexicalised) categories are studied then little insight will be gained into the discourse processes involved in categorisation when a single lexical item is not available to the discourse participants for the referential category.
Building on the ad hoc categories of Barsalou (1983), they stress the spontaneity of categorisation and the context-dependent nature of the categories themselves when one looks at examples from actual discourse as opposed to stylised examples. Overstreet and Yule (1997a: 87) suggest a continuum from lexicalised to non-lexicalised categories based on the degree to which categories are (a) conventionally and linguistically established, and (b) constrained by contextual factors. In the literature, the tags which help create these ad hoc categories go by different terms such as ‘general extenders’ (Overstreet & Yule 1997a, 1997b); ‘generalized list completers’ (Jefferson 1990); ‘tags’ (Ward & Birner 1992); ‘terminal tags’ (Dines 1980; Macaulay 1991); ‘extension particles’ (DuBois 1993); ‘vague category identifiers’ (Channell 1994; Jucker, Smith & Lüdge 2003); ‘imprecise language’ (Biber 2006) and vague category markers (O’Keeffe 2003, 2006; Evison et al. 2007). In this chapter we adhere to the term vague category marker (VCM). The questions of interest for this chapter are: do VCMs manifest themselves in spoken academic discourse, and if so, to what ends, and do such phenomena differ from or resemble uses of vague language in everyday causal conversation? This last question is important, since special registers in spoken language are often best characterised by the degree to which they resemble or depart from the typical linguistic features of everyday conversation. We enter this investigation via the notion of classroom modes (based on Walsh 2006), a set of ways of communicating be-
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.5 (270-329)
A comparative analysis of vague category markers
tween teachers and students which recur in the academic corpus, and which seem to have clear pedagogical foci in relation to overall goals in educational settings.
. Classroom modes In this section, a framework for analyzing spoken academic discourse at university level is presented and exemplified. The framework, SETT (Self-Evaluation of Teacher Talk, Walsh 2006), emphasizes the fact that interaction and classroom activity are inextricably linked, and acknowledges that as the focus of a learning event (e.g. lesson, seminar, or workshop) changes, so interaction patterns and pedagogic goals change. When language use and pedagogic purpose are considered together, different contexts emerge, making it possible to analyze the ensuing discourse more fairly and more objectively (see, for example, van Lier 1988; Seedhouse 2004). Under this variable view of contexts, student and teacher patterns of verbal behaviour can be seen as more or less appropriate, depending on a particular pedagogic aim. Characterizing university teaching in this way is not intended to offer an all-encompassing description nor a means to ‘code’ interaction patterns. Rather, the intention is to offer a framework and a metalanguage which may be used to interpret interaction in the context of third-level classrooms. Like other writers who adopt a variable view of classroom context (see, for example, Seedhouse 2004) the SETT framework, presented below, also adopts a variable approach. Specifically, the design of the framework rests on four assumptions. Firstly, all classroom discourse is goal-oriented: the prime responsibility for establishing and shaping the interaction lies with the teacher; secondly, pedagogic purpose and language use are inextricably linked – it is impossible to consider one without taking account of the other; thirdly, any higher education classroom context is made up of a series of micro-contexts (termed modes) which are linked to the social, political, cultural and historical beliefs of the participants (cf. Kumaravadivelu 1999); fourthly, micro-contexts are co-constructed by teachers and students through their participation, through face-to-face meaning-making and through a process of ‘language socialization’ (Pavlenko & Lantolf 2000). A mode is defined as a ‘classroom microcontext which has a clearly defined pedagogic goal and distinctive interactional features determined largely by a teacher’s use of language’ (Walsh 2006: 111). A modes analysis recognizes that understanding and meaning are jointly constructed, but that the prime responsibility for their construction lies with the teacher. The original SETT framework is based on a corpus of 14 English for Specific Purposes lessons, totalling approximately 12 hours or 100,000 words. The framework has since been applied to a much larger corpus of one million words of academic spoken English recorded in two universities on the island of Ireland. This
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.6 (329-486)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
corpus, LIBEL (Limerick and Belfast Corpus of Academic Spoken English), is composed of spoken academic data, collected at Queen’s University Belfast, Northern Ireland, and the University of Limerick, Republic of Ireland, from the following contexts: lectures, seminars, small group tutorials, oral presentations and conference papers. 50% of the corpus was collected in each site and its design matrix spans subject areas and colleges within the two institutions so as to achieve internal comparability and overall representativeness (see www.mic.ul.ie/ivacs). Table 1. Overview of number of hours collected to date (LI = Limerick, BEL = Belfast) Discipline
LI
BEL
Arts and Humanities Social Sciences Science Engineering and Informatics Business
36 26 5 11 3
6 15 17 9 2
Based on the initial corpus findings, qualitative samples of the data were analysed by working from concordance lines. In the qualitative stage, a CA methodology was used, which centred on turn-taking mechanisms in relation to teachers’ perceived goals of the moment and their stated written lesson aims. Interaction patterns were found to vary according to instructional activity; for example, establishing procedures to complete an activity resulted in a very different pattern of interaction to that of open-class discussion. The different patterns manifested themselves in the turn-taking, sequence of turns and topic management. According to Heritage, interactants’ talk is ‘context-shaped’ by a previous contribution, and ‘context-renewing’ by subsequent ones; understanding is indicated by the production of ‘next’ actions (1997: 162–163). In other words, participants both contribute to and demonstrate understanding of the interaction through the ways in which turns are managed. In this way, it is possible to characterize both the relationship between talk and actions, and assess the extent to which the ‘talk-in-interaction’ is appropriate to the shifting agenda and pedagogic goals of the moment. Following this procedure, it was possible, by analyzing the corpus, to identify four patterns, or four micro-contexts, called modes: managerial mode, classroom context mode, skills and systems mode, and materials mode. Each mode has distinctive interactional features and identifiable patterns of turn-taking related to instructional goals. While other modes could almost certainly be identified (depending on the specific context), these four are included as being representative of the interaction which takes place in the third level classroom, because they provide clear-cut examples of different types of interactional patterning and because they are intended to be used by teachers using samples of their own data as a means of awareness raising.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.7 (486-501)
A comparative analysis of vague category markers
Table 2. Classroom Modes (Walsh 2006) Mode
Pedagogic Goals
Interactional features
Managerial – To give an instruction – To organize the physical learning environment – To refer students to materials – To introduce or conclude an activity – To change from one mode of learning to another
– A single, extended teacher turn which uses explanations and/or instructions – The use of transitional markers – The use of confirmation checks – An absence of student contributions
Materials
– To provide input or practice around a piece of material – To elicit responses in relation to the material – To check and display answers – To clarify the focus of the material when necessary – To evaluate contributions
– Predominance of IRF (Initiation, Response, Feedback) pattern – Extensive use of display questions – Content-focused feedback – Corrective repair – The use of scaffolding
Skills and systems
– To enable students to produce correct answers – To enable students to manipulate new concepts – To provide corrective feedback – To provide students with practice in sub-skills – To display correct answers
– – – – – – –
The use of direct repair The use of scaffolding Extended teacher turns Display questions Teacher echo Clarification requests Form-focused feedback
Classroom context
– To enable students to express themselves clearly – To establish a context – To promote dialogue and discussion
– – – – – – –
Extended student turns Short teacher turns Minimal repair Content feedback Referential questions Scaffolding Clarification requests
The four modes, together with teachers’ interactional features and typical pedagogic goals, are summarized in Table 2. Owing to the multi-layered, ‘Russian doll’ (Jarvis & Robinson 1997: 225) quality of classroom discourse, any classification is not without its problems and the present one is no exception. Tensions between and within modes do exist: rapid movements from one mode to another, termed mode switching; brief departures from one mode to another and back again, termed mode side sequences; the fact that some sequences do not ‘fit’ into any of the four modes identified. These have all posed problems for description. Moreover, the analysis is further complicated by the homogeneous and heterogeneous quality of classroom contexts (Seedhouse
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.8 (501-595)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
2004); within a mode, every interaction is both similar to other interactions (homogeneous) and yet a unique encounter (heterogeneous).
. Data and methodology For this investigation, we draw on three spoken language corpora, LIBEL, from an academic setting, and two comparable corpora, CANCODE and LCIE, composed of casual conversation from Britain and Ireland. Table 3 summarizes these data. Table 3. Description of data used in the study Corpus
No. of words
Description
Limerick-Belfast Corpus of Academic Spoken English (LIBEL)
500,000 words2
– Consists of lectures, small group tutorials, laboratories and presentations – Collected in two universities on the island of Ireland: Limerick and Belfast3 – Data from common disciplinary sites (see Table 1)
Cambridge and Nottingham Corpus of Discourse in English (CANCODE)4
1 million words (a subset of the 5-million-word corpus)
– Consists of casual conversations between family and friends in Britain and Ireland – Designed to reflect spoken genres, speaker relationships and context (see McCarthy 1998)
Limerick Corpus of Spoken English (LCIE)
1 million words
– Designed as a comparable corpus to CANCODE – Consists of casual conversations between family and friends in Southern Ireland (see Farr et al. 2002)
In this chapter, we draw on two methodologies not always seen as complementary, corpus linguistics and conversation analysis. These have much to offer each other as they provide both quantitative and qualitative insights respectively (Carter . At the time of writing, LIBEL comprises one million words, 500,000 of which are fully transcribed. Its breakdown across disciplines in terms of number of hours transcribed is: Arts & Humanities 32%; Social Sciences 32%; Science 17%; Engineering & Informatics 15%; and Business 4%. . Note that while Limerick and Belfast are geographically on the same island (of Ireland), they come under two different jurisdictions: (1) The Irish Republic and (2) The United Kingdom and Northern Ireland, respectively. . CANCODE was a joint project between the School of English Studies, University of Nottingham, UK, and Cambridge University Press (with whom sole copyright resides). No part of the corpus may be used or reproduced without the permission of the copyright holder.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.9 (595-700)
A comparative analysis of vague category markers
& McCarthy 2002; O’Keeffe 2006; Walsh & O’Keeffe 2007). Applied to the corpora in Table 3, Wordsmith Tools software (Scott 1999) was used to produce word cluster (or chunks) frequency lists, that is to say, lists of recurrent strings of pre-selected extents (e.g. three-word clusters, four-word clusters). These quantitative data were sorted so as to identify VCMs in each corpus. This process involved concordancing individual high-frequency chunks operating as VCMs, and extensive manual reading of sample files. When we look at the micro-contexts, or modes, we employ CA to help understand the ways in which vague language is manifested in each mode, and the contribution VCMs make to the enactment of the modes. A brief summary of the transcription conventions used appears in the Appendix. Table 2 should be used as a reminder of the interactional features and pedagogic goals of each of the four modes.
. Analysis The quantitative findings based on the three corpora are illustrated below. These show the most common VCM forms and their frequencies in the three datasets. These forms are based on cluster analyses using Wordsmith Tools. First of all, at the level of geographical variation, these results point out that British English speakers’ use of VCMs is greater than that of Irish English speakers. However, closer examination shows that variation is accounted for mostly across less than half of all of the forms (i.e. it is these forms that diverge most): and all, and/or [something/anything/everything] (like that), and/or stuff (like that), (and) (all) this/that sort/kind of thing, and (and) (all) this/that sort/kind of thing. Overall, at a quantitative level, greater variation is evident between Irish and British English (i.e. CANCODE and LCIE) than between LCIE and the register-specific LIBEL data. At the level of contextual variation, or register, variation is accounted for by the higher frequency of use of certain forms in the academic data. These are: et cetera and and so on (and so forth)(like that). The various combinations of the form and so on (and so forth)(like that) account for 48% of all VCMs in the LIBEL data while the next most frequent form, et cetera, makes up 12% of all uses of VCMs in the academic data. In comparison, both the British and Irish casual conversation data draw more on a wider range of forms. If we remove the above forms from the overall count, we see that the total for LIBEL would be considerably lower than either of the casual conversation datasets.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.10 (700-834)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
Table 4. VCM forms resulting from cluster analysis (normalised to occurrences per million words)5 Form
LIBEL
LCIE
CANCODE
and so on (and so forth)(like that) et cetera (et cetera) and/or [something/anything/everything] (like that) and (all) (of) that and/or stuff (like that) or something (and) (all) this/that sort/kind of thing and things like that (and) (all) this/that sort of thing/stuff and all6 and all the rest (of it) this that and the other Total
524 136 126 77 67 61 52 46 21 13 4 2 1129
103 57 198 190 193 440 66 49 24 97 23 8 1448
60 30 1024 270 602 513 128 61 123 13 17 7 2848
As the concordance line extracts for the high frequency items et cetera and and so on show (Figures 1 and 2), the LIBEL VCMs were not found to be specific to particular disciplines. Note, in the case of et cetera, the strong preference for reduplication of form. In the results presented in Table 4, reduplications were counted as single VCM occurrences (i.e. a cross sectional hatching et cetera et cetera was counted as one vague category, marked by the form et cetera et cetera). However, as a percentage, 40% or all et cetera VCMs were reduplicated by speakers in LIBEL. This compares with 21% reduplication of et cetera in LCIE and 28% in CANCODE. As we have discussed above, various studies show that VCMs are used in casual conversation as involvement devices and are markers of the shared worlds of the speakers in a conversation. They draw on participants’ socio-cultural commonage and have an overall effect of marking in-group membership. In order to find out more about how and why speakers use VCMs in academic discourse, we turn now to a qualitative analysis which uses the four modes as its framework. . Only vague category uses of each form were counted. Rounded brackets ( ) mark words which may occur in the phrase and forward slashes / refer to either or options. For example, ‘(and all) this/that sort of thing/stuff ’ is a combined count of all of the following possibilities: a) and all this sort of thing; b) and this sort of thing; c) all this sort of thing (not including those already counted in a); d) this sort of thing (not including those already counted in a, b and c); e) and all that sort of thing; f) and that sort of thing; g) all that sort of thing (not including those already counted in e and f); h) that sort of thing (not including those already counted in e, f and g), and so on for all the combinations as above for ‘stuff ’. . This count is not inclusive of any of the combinations counted for and (all) (of) that.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.11 (834-834)
A comparative analysis of vague category markers
them and booksellers will hide them under the counter for closer customers et cetera. of am the purely educational system for developing media and social services et cetera. student tutor interactions. Okay. How do the students teacher teacher tutor et cetera. wrong. That’s the way to do it. Arrows should be this line thickness et cetera et cetera. You’re not writing an essay. Your use of short forms ellipsis R U there et cetera. an exploder view with an isometric with a cross sectional hatching et cetera et cetera. the country. There was great deal of talk about the harvest about farm work et cetera. not necessarily parallel. At least your printing is all up and down et cetera et cetera. include the inter relationships between the cube potential et cetera et cetera et cetera. product which really is open to am you know almost any additional dimension et cetera. somebody else is registered independent is going to join this party et cetera et cetera. On ability to respond to a child. A child’s level of communication et cetera et cetera. we can come back in and put in the shade and the shadow attaching et cetera et cetera.
Figure 1. Concordance samples of et cetera from LIBEL you know the Irish having a pig in the parlour and so on. you have like a play button or a stop and a rewind next and forward and so on. to exert relative to the the actual height of the workstation and so on. from an ergonomic viewpoint in relation to ahh the amount of force and so on. we have several different lists of ahh guidance for workplace design and so on. it can also contribute to the accidents and so on. The child itself gets its better at amm going to different people and so on. Well the bottom line is you will have people who are both tall and short and so on. So amm so dexterity it’s your ability to be able to manipulate objects and so on. So actually the average tax rate could be twelve thousand to zero and so on. You have to have fresh blood going into various muscles and so on. This file again by the twenty five hours. This by the forty five hours and so on. Amm and we can also then consider the actual workplace height and so on. one cry might mean lion yeah. Another cry might mean danger yeah and so on.
Figure 2. Concordance samples of and so on from LIBEL
. Managerial mode In Extract 3 below, we are at the beginning of a small group seminar on oral history, where the lecturer is setting up an activity and organizing the seating so that the session can begin. In this extract, as in most others where managerial mode is prevalent, there is little or no evidence of vague language category marking. Instead, the lecturer makes extensive use of instructional language (if you do have access to one of those transcripts; Just make sure you sit beside someone you can look in with) to locate the teaching and learning in time and space (all you do is pull the chair over by somebody who has one; I know a lot of people weren’t here last week for very good reasons ah just all you can do is fill in whatever words of wisdom were spread around ah from other people’s notes). Managerial mode occurs most often at
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.12 (834-887)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
the beginning of a piece of teaching and is characterized in the first instance by an extended teacher turn of more than one clause and a complete absence of student turns. The focus is on the ‘institutional business’ of the moment, the core activity. Typically, there is a considerable amount of repetition and some kind of ‘handing over’ to the students which occurs at the end of each sequence. At this point, there is a movement to another mode: in Extract 3, for example, the pedagogic focus is re-aligned away from directing learning (managerial mode) to analysing a tape script (skills and systems mode). Extract 3 <$1> It’s an awful setting in the way the room is at the moment but aam if I try and [move] around a few and all you do is pull the chair over by somebody who has one. Aah Yeah okay hopefully. Ah I’d like to make sure now about the tape and the volume is the volume is there? Yeah you might need to bring it up. Anyway look right folks we’ll start. Ok it’s very awkward. It’s not the kind of set up we’d like to have because the lines are too reminiscent of what’s going to happen in a week or two but it’s not very pretty but anyway sure we’ll do the best we can. Now aam I know a lot of people weren’t here last week for very good reasons ah just all you can do is fill in whatever words of wisdom were spread around ah from other people’s notes aam and if you do have access to one of those transcripts eh all the better. Just make sure you sit beside someone you can look in with. Where vague language does occur in managerial mode, it appears to function almost as a time-saving device so that the main item on the teaching agenda can be realised with minimal disruption and minimal waste of time. Compare Extract 4 (managerial mode) below from a different lesson in another discipline. Here, the lecturer is anxious to move on to the task and to engage students with their own data which they were required to collect as part of their assignment for the semester, as part of a media class. Extract 4 Really what I want to know when you having done the interview and scribed it and looked at the content of the interview how does it relate to how you understand audiences and you now understand more about audiences, about audience agency and so on. Then you will obviously feed into your concluding points about the particular interview, about how it went, about what the content of the interview has taught you in terms of audience based research. The VCM and so on, as mentioned above, in its various forms, accounts for 48% of all VCMs in LIBEL. Here we see that it appears to serve an important classroom function in managerial mode, that is, to minimise the time spent on setting up the task and to allow the teacher ‘hand over’ to students with minimal fuss. The vague
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.13 (887-936)
A comparative analysis of vague category markers
category audience agency and so on is taken as a given, something that they already know about from recent input. The VCM in Extract 4 stands to mark shared/given knowledge which is background to the task at hand. Perhaps more importantly, creating this shared space gives learners a sense that that they can do the task and enhances their confidence. The absence of such language might make the instruction more direct and reduce the sense of ownership and collective ‘struggle’ which are essential features of higher education teaching and learning. Extract 5 is another similar example from a physiotherapy lecture where the VCM is found in the context of managerial mode, where again it is used in setting the scene for the next stage. We note the use of once again here too as an additional explicit reference to known information. Extract 5 So there may be some accidents and maybe some injuries and maybe some a strong physiological stress on the body. Especially if maybe it’s ahh a hot environment or a very cold. Once again you can use some subjective assessments to actually to assess how the person ask the person if they’re fatigued in the course of this task if a body was lumbered and ahh things like this. That’s just kind of setting the scene. And we will be coming across some ahh more points like for example Corlett’s principles in the next ahh few lectures. I’m going to use some specific points which we will consider here in relation to the machine design and the operator so that we can reduce the problems for example with repetitive strain style injuries. Okay so. Next we’re going to have a look at amm evaluating the solution. . . To sum up then, we can say that overall there is comparatively little evidence of vague language in managerial mode in LIBEL. However, large-scale quantitative studies would be needed to substantiate this fully (note that, in the T2K-SWAL, corpus evidence of vague language is found in managerial mode; see Biber 2006). We speculate that the low occurrence of VCMs in this mode is due to lecturers’ concern to establish a meaningful context where learning can take place. Any examples which do emerge in the data serve to facilitate the process of setting up (or feeding back on) an activity, or organizing learning in the most effective way so as to move to a new phase. Throughout, the prime pedagogic goal is to transmit information in the most economical way. Being able to use a VCM to refer to assumed background knowledge at the start up phase is expeditious for the lecturer. . Materials mode Materials mode centres around a phase in a lesson where there is input or practice around a piece of material. Responses are elicited in relation to the material and
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.14 (936-1004)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
concepts and comprehension are checked. This mode is not one in which we find VCMs. As Extract 6 illustrates, interaction within this mode comprises many short IRF (Initiation – Response – Feedback) exchanges. The language is very specific and vague language, of any type, is rare here. Extract 6 [<$1> = the lecturer in an English Language class. At this point in the lesson, the lecturer is checking homework that has been assigned to the class. Here we see how the task, and the student’s response to it, is the main focus. In this case the lecturer checks for the word missing from the gap fill, flat, and corrects the student’s pronunciation of the answer ‘. . .flat. Not a flight’ ] <$1> And decorate the? The? Decorate the? <$2> Flat. <$1> Flat flat flat. Not a flight. <$2> Flat. <$1> Okay. Pronounce that word. . Skills and systems mode In skills and systems mode, the interaction revolves around the core subject of the particular discipline. The main pedagogic goals are to allow students an opportunity to familiarize themselves with new skills or concepts and to provide corrective feedback. The discourse is typically tightly controlled and teachers make frequent use of display questions7 to elicit responses which are then evaluated. Meanings may be clarified in the give-and-take of the interaction through error correction, requests for clarification and confirmation checks. Where new concepts are expressed by technical language, teachers may scaffold key terminology, offering students an opportunity to gain access to a discourse community through the language of that community. Vague language does occur in this mode, as illustrated in Extract 7 (we also note, as in Extract 5, the use of again and so I’ve mentioned in Extract 7 that refer the students back to known information. Here they serve as an additional means of scaffolding and schema-building). Extract 7 Lecturer: Okay and again equity <$G?> so I’ve mentioned they’re rejuvenating the economy poverty alleviation and so on. They’re the kinds of equities that we’re the effectiveness of the redistribution of taxation <$G?> amm Okay and so on. Now we can only cover this to a certain degree. We’re very limited by the amount of time. . Display questions are those where the questioner already knows the answer. They are typically associated with classrooms and quizzes.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.15 (1004-1039)
A comparative analysis of vague category markers
In the data, the lecturer is under time pressure and uses vague language (and so on) as a means of reducing his contribution. As he recaps, he avoids the need to re-list the points which have been covered in the earlier part of the lecture, allowing students an opportunity to recall that information for themselves (rejuvenating the economy, poverty alleviation and so on). But, and perhaps more importantly, the vague language expressed in and so on does more than save time and prompt students to recall what has been covered earlier in the lecture. This example of vague language also creates a sense of shared space, common ground. The lecturer here, through his use of vague language, is actually saying ‘we all know this’. The net effect of this is to ensure that students feel included and feel ‘safe’ as opposed to feeling intimidated or excluded. This is related to the in-group membership function of VCMs that has been noted in relation to their use in casual conversation. Friends and family use them to create and sustain a sense of membership within a circle of friends or family. However, when they are used in an academic context by the lecturer in a university classroom context, it may be seen as a device used by an expert to bring novices into a discipline, to make them feel part of a given subject area of field. The creation of a shared space and the use of inclusive language are crucial to successful teaching since they create an atmosphere in which students are prepared to take risks and offer their own perspective on the content of the lecture or seminar. In Extract 8, students are made to feel included and this is part of the process of collaborative meaning-making which is so important in higher education discourse. Here, the lecturer is giving students an opportunity to answer without making them feel trapped or intimidated by the question. Extract 8 [<$1> = lecturer, <$2> = student] . . . did you put it on V H S then or or ah yeah excellent did you try and digitize it or put it on the web or anything like that? <$2> totally <$G> <$1> oh very good excellent excellent
<$1> <$2> <$1>
The VCM or anything like that offers options to the student and also creates shared space in which students feel free to respond. It is an interesting choice of form, which is more associated with casual conversation (e.g. 32 occurrences per million words in LCIE and 35 in CANCODE, compared with 17 in LIBEL). As a VCM it is very open-ended with both or and anything in its form. This may account for why it occurs more in casual conversation than in academic language. We also note that its use here marks an attempt on the part of the lecturer to not only create a vague category but to hedge the directness of the question. A more direct question such as ‘did you put it on the web?’ might have been interpreted as a criticism and not
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.16 (1039-1096)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
received any response from students – the phrase or anything like that functions as a ‘softener’, oiling the wheels of the interaction, making the question less direct and facilitating a sense of membership. In Extract 9, we see that the use of a VCM by a student allows for the tentative positing of an answer to the lecturer’s question. This hedging effect of the VCM here provides face protection for the student as well as marking the proposition as tentative. Extract 9 [<$4> = lecturer, <$7> = student, there in the final line refers to services, as opposed to agriculture] <$4> Okay. In amm nineteen eighty eight and nineteen ninety one there was a labour force survey done in each year. Now I’m just going to show you what sectors that ahh they were concerned with. Okay? Now how about someone anyone hazard a guess. Just analyse the graph analyse the bar graph now. Why do you think agriculture is so low and services is so high? Mike? <$7> I don’t know agriculture. You know fixed pay and things like that. . . . More people going to college more people coming out of college. Better jobs going there. We also note here the use of the pragmatic marker you know in conjunction with the VCM. As noted by Carter and McCarthy (2006), you know projects the assumption that knowledge is shared or that assertions are uncontroversial, and reinforces common points of reference. The use of you know plus the VCM and things like that serves to tentatively project shared knowledge on the part of the student. Jucker, Smith and Lüdge (2003) point out that vague category construction asks the hearer to construct the relevant components of the set which they evoke and, in so doing, promotes the active cooperation of the listener. In the learning context of the LIBEL data, we could say that VCMs are also a vehicle for collective meaning-making. When they are used, in skills and systems mode, on the part of the lecturer, they promote active cooperation that results in learning. When they are used by students (as in Extract 9), they also promote cooperative peer-to-peer engagement with the category and reach out to the teacher for confirmation. They therefore provide evidence of learning in action. . Classroom context mode In classroom context mode, the management of turns and topics is determined by the local context; opportunities for genuine communication are frequent and the teacher plays a less prominent role, allowing students all the interactional space they need. The principal role of the teacher is to listen and support the interaction,
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.17 (1096-1155)
A comparative analysis of vague category markers
which frequently takes on the appearance of everyday conversation. Pedagogic goals typically centre on promoting dialogue and discussion; students have genuine opportunities to express their own ideas and to make real contributions to academic debate. Student responses are usually quite long and the teacher may offer scaffolded input or seek clarification as and when it is needed. Vague language functions here in much the same way as it functions in everyday conversation, that is, as an ‘involvement device’ ensuring listener participation and promoting equity and understanding. Consider Extract 10 below. Here the teacher is trying to make a point by using a literary reference. The language of the extract is very similar to everyday conversation. The choice of VCM, and stuff, also aligns it with casual conversation. This form is not very frequent in LIBEL compared with LCIE and CANCODE (the form and stuff alone occurs 41 times per million in LIBEL compared with 141 and 167 in LCIE and CANCODE respectively; see Table 4 for other related results for stuff patterns, all of which are greater in casual conversation). Extract 10 Lecturer: . . .did any of you ever read Angela’s Ashes? [unintelligible comments from students] Yeah exactly and it’s just it’s just the poems and stuff that the Daddy keeps on you know every time he has a few drinks and he’s living abroad and he’s broke and he’s after like leaving Ireland like arrived there filled with the pox and you know like. It was just like not at all a romantic story. He gets there and then before you know he’s like standing up all the kids at night time going we’ll die for Ireland. And you know there’s was all of these like poems and and stuff like that and it was all about like will you die for Ireland? The VCMs here (and stuff, and stuff like that) ensure that the listeners feel involved and that there is empathy towards the stance that the teacher adopts, i.e. agreement. As was the case in Extract 9, we see the use of you know (like) as an additional involvement device. Through the combined use of these markers then, the teacher is able to progress the discourse, bringing everyone along together and making sure that there is a sense of purpose and direction to the dialogue. Again, the vague language being used here serves to ‘soften the blow’ of a more didactic tone. A more conversational style is also almost certain to promote good listenership (McCarthy 2002, 2003) and means that the learning will be more memorable. Classroom context mode, then, out of the four modes included in the SETT framework (Walsh 2006), offers the greatest potential for vague language since it most closely resembles everyday conversation. Note that in this mode, vague language is as likely to be used by students as it is by teachers, as exemplified in Extract 11. Here, the student asks a question, but uses vague language (and everything) as a means of creating shared space and involving the teacher-listener. The
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.18 (1155-1214)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
net effect of this is to promote understanding and to ensure that the questioner is fully understood. Extract 11 [<$6> = student, <$1> lecturer] I have a question. Yeah? I was reading in one of our books that ethnicity and race are completely different things and ethnicity you learn things and race is a is a ahh is inherent in the you know in the blood and your appearance and everything. That is that wrong? <$1> It depends on what theorist you go after. <$6> <$1> <$6>
. Conclusions We stated at the outset as the main questions posed in this chapter: do VCMs manifest themselves in spoken academic discourse, and if so, to what ends, and do such phenomena differ from or resemble uses of vague language in everyday causal conversation? To the first question, VCMs clearly do occur in spoken academic discourse (as others, such as Biber 2006 and Evison et al. 2007, have also shown). In our data they occur less frequently than in casual conversation and they appear to rely strongly on certain forms (two forms accounted for 60% of all VCMs in the LIBEL data). To the second part of this overarching question, we can say that two main functions arise in the LIBEL: (1) VCMs can be used as expeditious devices. This is particularly the case within managerial mode where VCMs are used by the teacher to help expedite the start-up phase of a class or activity. Because they provide shortcuts that mark information or concepts that can be taken as given, shared or unproblematic, they very quickly establish what is common ground and facilitate a speedy handing over to the task phase of the class; and (2) VCMs, as in casual conversation, can be used as involvement devices, where again they mark shared knowledge but to do so in a way which scaffolds learning. In skills and systems mode, for example, they operate as two-way portals. For the lecturer, they can open a door to what is key shared knowledge for this phase of the class and create a shared space around this ‘learning commonage’. For the student, they open a door to a space where it is safe to take risks. Tentative propositions can be marked using VCMs and loss of face is avoided (see Extracts 9 and 11). In classroom context mode, we find that because language use, in general, resembles casual conversation more closely (see Walsh 2006), VCMs occur along with other vague language items and mark shared, uncontested knowledge. The second function that we refer to above, the use of VCMs as involvement devices, seems to parallel their function in casual conversation. However, we need to go back to the contextual differences of the interaction. The use of VCMs, by
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.19 (1214-1256)
A comparative analysis of vague category markers
lecturers, in the LIBEL data ties in with pedagogical goals of the interactional mode within which they occur. Classroom contexts differ from casual conversations. As we noted, all classroom discourse is goal-oriented; pedagogic purpose and language use are inextricably linked; any higher education classroom context is made up of a series of micro-contexts (modes) and these micro-contexts are co-constructed by teachers and students through their participation, through face-to-face meaning-making. However, while a modes analysis recognizes that understanding and meaning are jointly constructed, it also holds that the prime responsibility for their construction lies with the teacher. Therefore, the use of VCMs as involvement devices in academic discourse and in casual conversation cannot be fully equated since the power semantic differs between the institutional setting of the university classroom and that of casual conversation. Friends and intimates in casual conversation use many types of involvement devices as they symmetrically reinforce their social relationships. University lecturers, on the other hand, use involvement devices such as VCMs to try to ‘bring their student in’ both at the local level of pedagogic goal and at the higher-order level of initiating them into their community of practice (Wenger 1998). When students use them, they are not in the power-holding role and so they function aspirationally as involvement devices. From another pedagogical perspective, we also have to recognize the importance of VCMs as vocabulary items for non-native speakers of English, either those taking classes in English, or indeed, teaching classes through the medium of English. In this respect, VCMs need to be considered as core academic vocabulary items. From a second language perspective, it is clear that the ability to understand and create VCMs is an important part of classroom language, but from the perspective of teaching/lecturing, the ability to draw on the shared and known world, as we hope to have illustrated, is a very important part of building up knowledge schema. Their prevalence in terms of high frequency chunks in casual conversation also adds to the case for including them as vocabulary items not just in English for Academic Purposes programmes.
References Barsalou, W. L. 1987. The instability of graded structure: Implications for the nature of concepts. In Concepts and Conceptual Development, U. Neisser, (ed.), 101–40. Cambridge: CUP. Barsalou, W. L. 1983. Ad hoc categories. Memory and Cognition 11: 211–77. Biber, D. 2006. University Language: A corpus-based study of spoken and written registers. Amsterdam: John Benjamins. Carter, R. A. & McCarthy, M. J. 2006. Cambridge Grammar of English: A comprehensive guide to spoken and written English grammar and usage. Cambridge: CUP.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.20 (1256-1362)
Steve Walsh, Anne O’Keeffe and Michael McCarthy
Carter, R. A. & McCarthy, M. J. 2002. From conversation to corpus: A dual analysis of a broadcast political interview. In Windows on the World: Media discourse in English, A. Sánchez-Macarro (ed.), 15–39. Valencia: University of Valencia Press. Carter, R. A. & McCarthy, M. J. 1995. Grammar and the spoken language. Applied Linguistics 16(2): 141–58. Chafe, W. 1982. Integration and involvement in speaking, writing, and oral literature. In Spoken and Written Language: Exploring orality and literacy, D. Tannen (ed.), 35–53. Norwood NJ: Ablex. Channell, J. 1994. Vague Language. Oxford: OUP. Cruse, D. A. 1986. Lexical Semantics. Cambridge: CUP. Cutting, J. 2000. Analysing the Language of Discourse Communities. Oxford: Elsevier. Dines, E. 1980. Variation in discourse – and stuff like that. Language in Society 1: 13–31. DuBois, S. 1993. Extension particles, etc. Language Variation and Change 4: 179–203. Evison, J., McCarthy, M. J. & O’Keeffe A. 2007. ‘Looking out for love and all the rest of it’: Vague category markers as shared social space. In Vague Language Explored, J. Cutting (ed.), 138–157. Basingstoke: Palgrave. Farr, F., Murphy, B. & O’Keeffe, A. 2002. The Limerick Corpus of Irish English: Design, description and application. Teanga 21: 5–29. Franken, N. 1997. Vagueness and approximation in relevance theory. Journal of Pragmatics 28: 135–151. Heritage, J. 1997. Conversational analysis and institutional talk: Analysing data. In Qualitative Research: Theory, method and practice, D. Silverman (ed.), 161–183. London: Sage. Jarvis, J. & Robinson, M. 1997. Analysing educational discourse: An exploratory study of teacher response and support to pupils’ learning. Applied Linguistics 18(2): 212–228. Jefferson, G. 1990. List construction as a task and resource. In Interaction Competence. G. Psathas (ed.), 63–92. Lanham MD: University Press of America. Jucker, A. H., Smith, S. W. & Lüdge, T. 2003. Interactive aspects of vagueness in conversation. Journal of Pragmatics 35: 1737–69. Kumaravadivelu, B. 1999. Critical classroom discourse analysis. TESOL Quarterly 33(3): 453– 484. Macaulay, R. K. S. 1991. Locating Dialect in Discourse: The language of honest men and bonnie lasses in Ayr. Oxford: OUP. McCarthy, M. J. 2003. Talking back: ‘Small’ interactional response tokens in everyday conversation. In Special issue of Research on Language and Social Interaction on ‘Small Talk’, J. Coupland (ed.), 36(1): 33–63. McCarthy, M. J. 2002. Good listenership made plain: British and American non-minimal response tokens in everyday conversation. In Using Corpora to Explore Linguistic Variation. R. Reppen, S. Fitzmaurice & D. Biber (eds), 49–71. Amsterdam: John Benjamins. McCarthy, M. J. 1998. Spoken Language and Applied Linguistics. Cambridge: CUP. McCarthy, M. J. & Carter, R. A. 1995. Spoken grammar: What is it and how do we teach it? ELT Journal 49(3): 207–218. Mervis, C. B. & Rosch, E. 1981. Categorization of natural objects. Annual Review of Psychology 32: 89–115. O’Keeffe, A. 2006. Investigating Media Discourse. London: Routledge. O’Keeffe, A. 2003. ‘Like the wise virgins and all that jazz’ – Using a corpus to examine vague language and shared knowledge. In Applied Corpus Linguistics: A multidimensional perspective, U. Connor & T. A. Upton (eds), 1–20. Amsterdam: Rodopi.
JB[v.20020404] Prn:11/04/2008; 10:37
F: SCL3102.tex / p.21 (1362-1442)
A comparative analysis of vague category markers
O’Keeffe, A., McCarthy, M. J. & Carter, R. A. 2007. From Corpus to Classroom: Language use and language teaching. Cambridge: CUP. Overstreet, M. & Yule, G. 1997a. On being explicit and stuff in contemporary American English. Journal of English Linguistics 25(3): 250–58. Overstreet, M. & Yule, G. 1997b. Locally contingent categorization in discourse. Discourse Processes 23: 83–97. Pavlenko, A. & Lantolf, J. P. 2000. Second language learning as participation and the (re)construction of selves. In Sociocultural Theory and Second Language Learning, J. P. Lantolf (ed.), 155–178. Oxford: OUP. Rosch, E. 1978. Principles of categorization. In Cognition and Categorization, E. Rosch & B. Lloyd (eds), 27–48. Hilldale NJ: Lawrence Erlbaum. Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M. & Boynes-Braem, P. 1976. Basic objects in natural categories. Cognitive Psychology 2: 491–502. Scott, M. 1999. Wordsmith Tools Software. Oxford: OUP. Seedhouse, P. 2004. The Interactional Architecture of the Second Language Classroom: A conversational analysis perspective. Oxford: Blackwell. van Lier, L. 1988. The Classroom and the Language Learner. London: Longman. Walsh, S. 2006. Investigating Classroom Discourse. London: Routledge. Walsh, S. 2002. Construction or obstruction: Teacher talk and learner involvement in the EFL classroom. Language Teaching Research 6: 1–23. Walsh, S. & O’Keeffe, 2007. Applying CA to a modes analysis of third-level spoken academic discourse. In Conversation Analysis and Languages for Specific Purposes, P. Bowles & P. Seedhouse (eds), 101–139. Frankfurt: Peter Lang. Ward, G. & Birner, B. 1992. The semantics and pragmatics of “and everything”. Journal of Pragmatics 19: 205–214. Wenger, E. 1998. Communities of Practice: Learning, meaning and identity. Cambridge: CUP. Zhang, Q. 1998. Fuzziness – vagueness – generality – ambiguity. Journal of Pragmatics 29: 13–31.
Appendix Transcription conventions <$X> speaker turn, e.g. <$1> = speaker 1, <$2> = speaker 2 etc. in order of ‘appearance’ on the recording. <$G?>... <$G2> <$G?> marks uncertain or unintelligible utterances where the number of syllables cannot be guessed. Where the number of syllables can be discerned, this number is marked, e.g. <$G2> denotes two intelligible syllables.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.1 (47-115)
Emphatics in academic discourse Integrating corpus and discourse tools in the study of cross-disciplinary variation Marina Bondi University of Modena and Reggio Emilia, Italy
The role played by mitigation in academic discourse has been widely debated in the literature, but little attention has been paid to emphatics, expressions used to intensify the degree of certainty of an utterance and to increase its illocutionary force. Focusing on the use of adverbs in journal articles and on their evaluative orientations/parameters, the chapter looks at how their frequencies, meanings and uses vary across two “soft” disciplines: history and economics. The study combines a corpus and a discourse perspective, and shows that emphatics signal “engagement” as well as “stance”, by positioning research in the context of disciplinary debate, highlighting the significance of the data or the conclusions produced, negotiating convergent or conflicting positions with the reader.
.
Introduction
This chapter is part of a wider study which aims at investigating the role played by stance markers (e.g. adverbs like actually, definitely, apparently) in academic discourse. Great interest has been shown in redefining the interactive level of discourse in the light of a plurality of analytic models of evaluative elements of discourse (e.g. Hunston & Thompson 2000). In its broadest definition, evaluation is understood as “the expression of the speaker or writer’s attitude or stance towards, viewpoint on, or feelings about the entities or propositions that he or she is talking about” (Hunston & Thompson 2000: 5). Included in this definition are forms of modality as well as a vast range of instruments of metadiscourse aimed at organising the discourse, constructing and maintaining relations between the speaker/writer and the listener/reader, as well as reflecting the value-system of the speaker and the discourse community he or she is part of. The analysis of evaluation thus links up with the area of studies on metadiscourse (e.g. Vande Kopple 1985; Crismore 1989; Hyland 1998a; 1998b; 2005). If early classifications of the
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.2 (115-165)
Marina Bondi
metadiscursive elements seemed to keep the textual and interpersonal functions (e.g. Halliday 1985) rigidly separate, more recent studies seem instead to focus on the overlapping of these two types of functions (e.g. Conrad & Biber 2000). A similar point has been made by Swales and Burke (2003: 4) in their study on the combination of polarised – i.e. “strongly positive” and “strongly negative” – v. centralized – i.e. “more neutral” – evaluative adjectives with intensifiers across academic registers. In particular, they note that the co-occurrence of these elements may reveal interesting rhetorical effects, e.g. increasing the interpersonal orientation of merely ideational statements. The analysis of stance adverbials – adverbials “commenting on the content or style of a clause or a particular part of a clause” (Biber et al. 1999: 853) – could be a case in point. From the point of view of Hunston and Thompson (2000), they can be defined as adverbials expressing the writer’s opinion as to entities or propositions in the text. The view taken here is that adverbials of stance do not only enable monologic discourse to be evaluative, but they also often assume a common ground between reader and writer in terms of what is regarded as scientifically ‘good’ or ‘bad’ at any given point in the discourse. Adverbials can thus contribute to the dialogic and argumentative features of academic discourse by constructing this common ground between reader and writer. By doing this, they also contribute to the organization of discourse and to the representation of conflict and negotiation within the discourse community. Emphatics, in Crismore’s sense (1989), or boosters, for Hyland (1998b, 2000b), are for the most part adverbs and adverbials which attribute an increased force or authority to statements: i.e. expressions used to emphasize a statement, intensifying the degree of certainty expressed and increasing its illocutionary force. This aspect has been extensively dealt with by Wierzbicka (2006) in her study of the wider cultural implications of the use of epistemic adverbs in modern English. In particular, Wierzbicka (2006: 270) illustrates the semantic peculiarities of the “confident adverbs” evidently, clearly and obviously, which appear to express varying degrees of writers’ confidence towards their own statements. Along with hedges (or mitigators), emphatics communicate both interpersonal and ideational information, allowing writers to convey judgments with greater accuracy and situate their positions in relation to knowledge and truth claims. Since new research gains approval because it is able to negotiate accepted views and ideas with those as yet unaccepted or unknown, these stance devices play an essential role in academic discourse, as communicative strategies for increasing or reducing the force of a statement, conveying conviction or caution, etc. in order to get the researcher’s views across in a convincing manner. A considerable amount of literature has been dedicated to the question of mitigation, both in studies of general communication and in the specific domain of academic discourse (e.g. Myers 1989; Hyland 1998a; Markkanen & Schroeder 1997). More recently, Poos and Simpson (2002: 17) inte-
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.3 (165-219)
Emphatics in academic discourse
grate the traditional view of hedges as signals of modesty and/or uncertainty, by showing that they may also reflect the equally important pragmatic function of displaying “more solidarity” towards “less academically indoctrinated” interlocutors. It is readily conceded that the semantic features and pragmatic functions of emphatics deserve closer study (Hyland 2000a, 2000b; Biber et. al. 1999; Conrad & Biber 2000; Precht 2003). Yet relatively few studies have addressed their role. The definition adopted for emphatics is purposely extensive due to an interest in exploring the function of these textual elements from a text-pragmatic and argumentative point of view. For the purpose of this chapter, emphatics can be defined as expressions used to increment the degree of certainty and increase or strengthen the illocutionary force of the statement. They also attribute a truthvalue or importance to what is being emphasized. The category includes a variety of tools. The most obvious forms are adverb(ial)s: As everyone knows, as we can plainly see, undoubtedly, etc. But similar functions can also be realized by superordinate projecting/inference frameworks like those analyzed by Hyland and Tse (2005) in their study of “evaluative that”: It is generally agreed that, we believe that, the key issue is, this shows, etc. All these tools share some basic pragmatic functions. First of all, they variously foreground the writer’s degree of endorsement of a statement and the degree of universality of the related belief. This is why Hyland (e.g. 2004: 16) classifies boosters primarily as expressions of stance rather than engagement. It will be easily recognized, however, that adverbs like certainly may be primarily expressions of writer’s stance, but they also tend to limit the reader’s possibility to disagree, thus becoming tools for reader’s positioning or engagement resources, i.e. tools by which writers adjust and negotiate the arguability of their utterances to their interlocutors (Hyland 2001, 2005).1 Similarly, when looking at classifications of evaluation parameters/orientations, we will notice that there is wide convergence on some meaning areas, but also that distinctions can be blurred. Following Conrad and Biber (2000: 57), we may distinguish epistemic stance – commenting on the certainty (or doubt), reliability, or limitations of a proposition, including comments on its source – from attitudinal stance – conveying the speaker’s attitudes, feelings, or value judgements – but we may still recognize that the two are part of the same meaning area (stance, or evaluation) and that distinctions will not always be clear-cut. It may not be possible, useful or accurate to distinguish the writer’s judgment about the certainty, . Using writer-orientation and reader-orientation as a basic classification tool, Merlini Barbaresi (1987: 4–6) identifies a significant semantic and pragmatic difference in the use of different emphatics in argumentative discourse. Focusing on the difference in functions of obviously and certainly, she sees the former as essentially an “epistemic modifier”, and the latter as more of an “indicator of inferability”. The two differ significantly in orientation, with the epistemic modifier being locutor-oriented and the indicator of inferability being essentially receiver-oriented.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.4 (219-265)
Marina Bondi
reliability, and limitations of the proposition from her/his attitude or value judgment about the proposition’s content (Silver 2003). As Hyland has demonstrated for hedges, “in actual use the epistemic and affective functions of hedges are often conveyed simultaneously” (1998a: ix), preventing the formation of discrete descriptive categories. A similar case can be made for emphatics.2 Approaches to evaluation that emphasize continuity across the epistemic/ attitudinal divide have often offered alternative classifications of evaluative meanings. Thompson and Hunston’s basic parameters of evaluation (the main semantic areas in which evaluation can be placed) are those of certainty, expectedness and importance or relevance (Thompson & Hunston 2000: 23–24). Lemke’s evaluative orientations also include three similar relevant categories: (a) warrantability/probability, as exemplified by adverbs like certainly and undoubtedly; (b) usuality/expectability, as in invariably; (c) importance/significance, as in significantly (Lemke 1998: 37). The study presented in this chapter focuses on the role of emphatics in academic discourse in two disciplines. This is done by looking at variation in frequencies, meanings and functions of selected adverbs, as signals of the argumentative structures of research articles in history and economics. The next section of the chapter provides a brief presentation of the material used for the study, as well as of the methodology adopted, ranging from genre analysis (with the identification of textual and generic structure) to corpus tools (with the study of lexicalizations in context). The results of the analysis will start with a preliminary overview of variation in frequency data and move on to an examination of syntactic scope, semantic preference and textual patterns of selected items spanning the range of evaluative orientations listed above (undoubtedly, certainly, invariably, significantly). The discussion of the data will focus on differences in the two disciplines and on their variety and approaches.
. Methods and material . Methodological preliminaries The study of academic discourse across disciplines, as outlined for example by Hyland (2006), is inherently comparative. Although not always explicit, comparison seems to be the main tool for research in language varieties: even when focusing on a single register, analysis evokes comparison as the basis for any conclusion on the specificity of a language variety. Explicit comparison, on the other hand, does not . See also Precht (2003) for an emphasis on the relations between expressions of affect, evidentiality and hedging.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.5 (265-360)
Emphatics in academic discourse
just substantiate claims about distinctive features: it also helps bring out elements of variation. It is a heuristic procedure, as well as an important support for claims. The methodological aim of this chapter is to explore how comparative analysis of disciplinary variation is enhanced by integrating corpus and discourse tools. Combinations of both perspectives are advocated and practiced by many, especially in the area of academic and professional discourse studies (e.g. Biber et al. 1998; Biber et al. 2004; Connor & Upton 2004; Hyland 2000a, 2002, 2004; Del Lungo & Tognini Bonelli 2004; Tognini Bonelli & Del Lungo 2005). In a wider framework, a major figure like John Sinclair has played a leading role in developing theory and practice in both fields (e.g. Sinclair 2004). The reasons for choosing not to separate the two approaches, however, may help clarify what is desirable in their integration. A discourse perspective draws attention to how interaction and argument are instantiated in textual practices which are recognized and continually redefined by discourse communities. A corpus perspective looks at words in combination and finds in phraseology the ideal starting point for the exploration of the systematic relation between text and form (Sinclair 2005). Defining one’s own object of analysis from both points of view helps relate textual practice to language choice. It is one way of making sure that statements about genre and discourse are substantiated with reference to data: attention to patterns of form highlights the existence of systematic relations and trends, besides possibilities. Integration of both perspectives, however, also ensures that corpus data are not just described, but interpreted in terms of verbal action and textual structures, beyond immediate lexico-semantic associations. The choice to start from a discourse or a corpus perspective should not be taken as a methodological statement in favour of a specific direction. Quite the opposite: corpus tools can be seen as both “catalyzing” or supporting the analysis and the interpretation in terms of discourse, and vice-versa. The presentation of the analysis may have to follow a specific sequence, but this is mainly due to the linearity of text. The interrelation of the two perspectives should be seen as a dialogic sequence, where corpus and discourse – just like participants in interaction – co-construct the development of the research process. . Material The analysis is based on two specialized corpora of journal articles, taken to be representative of research writing in two different disciplines: economics and history. The corpora are about 2.5 million words each and include all the articles published in ten journals for each disciplinary area over the course of two years (1999–2000). The journals are listed in Table 1 below, together with the acronyms that identify them in the examples that follow.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.6 (360-412)
Marina Bondi
Table 1. List of journals included in the two corpora and their acronyms Economics
History
European Economic Review (EER) European Journal of Political Economy (EJoPE) History of Political Economy (HOPE) International Journal of Industrial Organization (IJoIO) International Review of Economics and Finance (IRoEF) Journal of Corporate Finance (JoCF)
Labour History Review (LHR) Historical Research (HR) Gender & History (GH) Journal of European Ideas (JEI)
Journal of Development Economics (JoDE) Journal of Economics and Business (JOEB) Journal of Socio-Economics (JSE) The North American Journal of Economics and Finance (NAJEF)
Journal of Medieval History (JMH) Journal of Interdisciplinary History (JIH) Journal of Social History (JSH) Studies in History (SH) American Quarterly (AQ) American Historical Review (AHR)
The corpus design aims at a description of English as an international language, rather than a specific geographical variety of English. Thus mostly international journals, published both in the UK and in the US, were included. No attempt was made to separate native from non-native speakers/writers: the aim of the analysis was not to prescribe purity in writing, but to describe what is published in a variety of well-established journals in the community of historians and economists over a range of subdisciplines. The perspective adopted for the analysis paid attention both to the rhetoric and organization of text in discourse and to the language resources and the meanings realized in text. The methodology combined tools from discourse analysis and corpus linguistics. From discourse studies, the notion of genre – defined by Swales (1990: 68) as a class of communicative events sharing a common purpose – and the notion of units identified by their pragmatic function were used. When focusing on the lexical tools that allow academic writers to introduce emphasis, tools from corpus linguistics were used: in particular keywords, concordances, collocates and clusters, i.e. repeated strings of words as defined by Scott (1998). The first step of the study was an analysis of frequency data: this was meant to provide an overview of quantitative variation. The next section reviews the most common emphatic adverb(ial)s appearing in the research articles in the two fields of study and makes a few initial hypotheses as to what their use may be an indication of. The overview is based on keywords, as defined in Wordsmith Tools (Scott 1998), i.e. words that are unusually frequent or infrequent in one corpus or text when compared with a reference corpus. Key-ness indices based on comparing statistical frequencies are a measure of how much one word characterizes a corpus as against another.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.7 (412-530)
Emphatics in academic discourse
The bulk of the study is based on the analysis of concordances. After having identified a few significant adverbs to be used for closer study, the co-text of the nodes was analyzed with a view to their: a) syntactic roles, i.e. the scope of the adverb and its patterns of pre/post-modification; b) lexico-semantic patterns, i.e. patterns of collocation and semantic preference: “entities” and “processes” emphasised; c) textual patterns, i.e. pragmatic functions, argumentative moves and position in linear units of the text (introduction-body-conclusion). Their use was analyzed in the development of three logico-argumentative positions: an inferential position which pieces ideas or arguments together through verbal relations of analogy, cause-effect, specification, generalization, etc., a contrastive position which places ideas in opposition, and a concessionary and contrastive (or contrastive and concessionary) position which attenuates a contrast through partial acknowledgment or acceptance of the oppositional idea or argument. The final analysis regarded how these emphatics fit into textual patterns. To facilitate the task, the focus was exclusively on the most general distinction between introduction, body and conclusion of the article. By noting how and where the adverbs are called upon to intervene when placed in the text, certain generalizations about the strategic function of the emphatics in highlighting disciplinary differences can be put forth.
. Results and discussion . Comparing frequencies: Keywords of economics and history The preliminary overview of the study was carried out by identifying keywords through comparison of the two corpora. The adverbs in Tables 2 and 3 are listed in descending order of key-ness. They help provide a first general idea of the emphatics which are used most regularly in the two disciplines, and offer a sense of adverbial variation. Since no initial attempt was made to exclude less influential adverbs, the list contains all manner adverbs in our sub-corpora which conform to the key-ness criteria and reports their frequency in the corpus together with normalized figures (occurrence per hundred words) and key-ness index. From an initial, cursory analysis of the adverbs listed, a number of these can be said to function rarely, or never, as sentence adverbs (see below for the distinction in terms of syntactic role and scope). In economics, for instance, only significantly and typically really distinguish themselves as emphatics: typically is probably the most interesting one from the point of view of its direct relationship to the abstracting needs of a social science like economics (cf. also Bondi 2002), but significantly is also clearly related to methodological issues, in particular to the definition of statistical significance,
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.8 (530-550)
Marina Bondi
Table 2. Keywords in economics Adverb
Frequency in Economics
% in Economics
Frequency in History
% in History
Key-ness index
significantly positively substantially unambiguously typically perfectly fully
814 331 225 80 313 215 400
0.03 0.01 – – 0.01 – 0.02
214 32 41 9 134 81 245
– – – – – – 0.01
350.5 276.1 133.6 62.6 67.4 58.2 32.3
Table 3. Keywords in history Adverb
Frequency in History
% in History
Frequency in Economics
% in Economics
Key-ness index
certainly especially particularly throughout increasingly really entirely largely undoubtedly inevitably thoroughly surely evidently predominantly invariably clearly
523 833 837 527 429 244 322 391 139 112 72 150 88 76 72 707
0.02 0.03 0.03 0.02 0.02 – 0.01 0.02 – – – – – – – 0.03
144 372 412 206 143 67 117 171 31 29 13 59 23 19 17 515
– 0.01 0.02 – – – – – – – – – – – – 0.02
246.4 198.0 163.3 157.3 160.1 113.5 107.1 96.6 80.2 55.2 47.3 44.3 43.0 38.7 38.6 37.6
which seems to blur the distinction between the parameters of expectedness and significance: statistical significance is indeed based on expectedness. In history, although the keyword variety of adverbs is much greater, those which have a more extensive scope are certainly, undoubtedly, evidently, invariably and clearly, once again covering the whole range of parameters (certainty, expectedness, importance). Other adverbs, such as largely, thoroughly and especially do not normally function as sentence adverbs. Even a rough overview like this can be related to disciplinary variation. Interpreting frequencies in the light of disciplinary values may suggest that economics tends to place emphasis on a simplification of reality based on a process of abstraction (typically) and on statistics (significantly), whereas history places emphasis on frequency and accumulation of factual data (usually, largely, inevitably, thor-
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.9 (550-601)
Emphatics in academic discourse
oughly, invariably etc.), as well as their interpretation (as shown by a variety of epistemic markers). This in turn may be related to an emphasis on the singularity of events (Holmes 1997) or to forms of divergence from the tendency to abstract and generalize shown by other disciplines. . Ambiguity of functions: Focus on syntactic role/scope The study of the syntactic role and scope of adverbs often presupposes a basic distinction between emphasizers and intensifiers. Intensifiers are degree adverbs with a grading function; they are defined as degree adverbs that “scale upwards from an assumed norm” (Quirk et al. 1985: 445) or neutral point. Some modify gradable adjectives and indicate degrees on a scale (e.g. extremely cautious), while others indicate an endpoint on a scale (e.g. totally different) (Biber et al. 1999: 554–555). Emphasizers, on the other hand, contribute to the expression of modality or stance: they add to the force of the modified predicate and their syntactic scope extends over the whole predicate, they strengthen the illocutionary point of the utterance and signal that what is being emphasized is taken to be true and/or important. In expressing the semantic role of modality, emphasizers have a reinforcing effect. They add to the force as opposed to the degree of the modified predicate. As such, according to Quirk et al. (1985: 583), they do not require a gradable predicate. This does not necessarily rule out the notion that emphatics produce a semantic effect which may be similar to that of intensifiers. Moreover, when the emphasizer occurs with a gradable predicate, it “takes on the force of an intensifier” (Quirk et al. 1985: 583). The basic functions of expressing stance and degree can thus be seen to overlap. Another element which generally distinguishes emphasizers from intensifiers is their syntactic scope. It is recognized that emphasizers may take scope over the predicate or the whole sentence, while intensifiers do not (Quirk et al. 1985). And yet, there seems to be no fool-proof way of discriminating between the two in this sense either, since although intensifiers demonstrate a reduced scope, there is no set scope of ‘emphatics’, which vary greatly in correspondence with their pragmatic and argumentative roles. As Merlini Barbaresi (1987: 19) points out with respect to epistemic modifiers (e.g. certainly, inevitably, no doubt, incontestably), their argumentative scope and force of assertion “are directly proportional to the 1) degree of subjectivity/objectivity of the thesis, 2) relevance assigned to the thesis in the argumentative line, whether micro- or macro-structural”. When focusing on variation of syntactic role and scope of adverbs in our corpus, we sometimes notice a considerable difference from case to case. If initial position is mostly an indicator of a scope extending to the whole sentence, mid-position may be more ambiguous. Example (1) provides an illustration: unquestionably clearly functions as a sentence adverb (an emphasizer), whereas
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.10 (601-702)
Marina Bondi
undoubtedly may be interpreted both as a modifier (whose scope is limited to the following adjective, clear) and as an adverbial of stance qualifying everything that follows. (1) Unquestionably, the success of A History of Women, edited by Michelle Perrot and Georges Duby, as well as the substantial audience for the journal Clio, are undoubtedly clear indications of the importance of this history which has now outgrown its marginal stage. (GAH)
The scope of the adverbial may actually extend beyond the sentence, as it may participate in a macro-textual pattern. In Example (2), most obviously has a strong anaphoric quality, referring to a predictive marker (Tadros 1985) appearing earlier on, at the end of the previous sentence (several different ways). Significantly, on the other hand, functions as an adverbial modifying the verb and its object. (2) . . .the sample is skewed in several different ways. Most obviously, it is a catalogue of books published in London, and thus significantly excludes important publishers in Glasgow and Edinburgh, like William Collins and W. & R. Chambers. In addition, the Publishers’ Circular, from which the data are drawn, did not provide either a full or a representative sample of publications. (SIH)
Example (2) shows that thematic position often extends the scope of the adverb and gives it a cohesive function: thematized adverbs do not simply extend their scope forward, but they also signal the relationship between the syntactic unit they introduce and the previous text. Table 4 below offers an overview of quantitative patterns of four selected adverbs in our corpus. The table clearly shows that, in the case of these four adverbs, thematic position is on the whole much more frequent in history, but it also shows that economics and history may tend to favour different types of emphatics in thematic position: economics favours certainty adverbs, whereas history ranges rather equally over the three parameters of certainty, expectedness and importance. When looking at the same adverbs used as intensifiers, we get a complementary picture. Table 5 below shows how often the four adverbs under investigation are used to modify adjectives or adverbs. Economics clearly prefers this pattern, with the more limited scope of the adverb, and the most obvious trend is actually that of significantly, which is extremely frequent in economics, but mostly as intenTable 4. Sentence adverbs: Initial position Parameter of evaluation
Adverbs
Economics
History
Certainty
certainly undoubtedly invariably significantly
14/144 (9.7%) 3/31 (9.7 %) 1/17 (5.9%) 3/814 (0.4%)
91/523 (17.4 %) 16/139 (11.5 %) 16/72 (22.2 %) 38/214 (17.7%)
Expectedness Importance
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.11 (702-771)
Emphatics in academic discourse
Table 5. Modifying adj./adv. Parameter of evaluation
Adverbs
Economics
History
Certainty
certainly undoubtedly invariably significantly
27/144 (18.7%) 2/31 (6.5 %) 3/17 (17.6%) 569/814 (69.9%)
65/523 (12.4 %) 21/139 (15.1 %) 8/72 (11.1 %) 44/214 (20.5%)
Expectedness Importance
Table 6. Pre-modification of emphatic adverbs Economics
History
significantly
not (94/814) quite (3/814)
invariably certainly undoubtedly UNMODIFIED
almost (2/17) almost (4/144) 0/31 1011 (90.7%)
not (5/214) more (10/214) most (17/214) almost (17/72) almost (40/523) 0/139 751 (89.4%)
sifier of adjectives and adverbs, in contexts which make clear reference to statistical significance. A look at pre-modification elements also reveals interesting patterns. On the whole, pre-modification of these adverbs is rather limited. Some adverbs (e.g. undoubtedly) are never pre-modified, while others (invariably, certainly) may occasionally be graded; in this case, the tendency is once again for history to favour a wider range of shades. Once again, however, the very frequent use of significantly in economics reveals a peculiar pattern in the high incidence of negative contexts. As a general rule, pre-modification of adverbs may be related to cases of “polarization” and to parameters of evaluation. Undoubtedly is clearly the most “polarized” of our adverbs here: it does not accept shades of ‘undoubtedness’. Almost still combines with more polarized elements like invariably, especially in history. Significantly and markers of importance in general are less “polarized” than markers of warrantability and usuality; they have a higher “grading” function: things can be more or less relevant, more or less important. The majority of uses of significantly in economics, however, have statistical significance as their object, and statistical significance is typically used to establish when frequencies start being more or less significant: once the parameters are set, frequencies either have or do not have statistical significance. Many of the numerous negative occurrences of significantly have this particular meaning, and occur in patterns of analogy and contrast, as illustrated in Example (3): (3) The coefficients of the variables denoting trade agreements indicate that trade between the countries in the EEA zone is significantly greater (at the 1% level)
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.12 (771-877)
Marina Bondi
than average OECD trade. The same holds for trade between Australia and New Zealand. The NAFTA agreement does, however, indicate that trade between Canada, Mexico and the USA is not significantly greater than average OECD trade. Finally, the EU countries’ custom union with Turkey seems to have had a positive influence on trade. (EJOPE)
The specific use of significantly in economics also involves a highly formulaic use of the expression. A look at 5-word clusters (strings of word forms) in the concordance reveals frequent occurrences of “chunks” or extended collocations of language including significantly, whereas no 5-word cluster could be found for the other adverbs. Examples of clusters in negative contexts only are particularly numerous, as shown in Table 7. Table 7. Significantly: 5-word clusters in negative contexts 5-word cluster
Frequency
not significantly different from zero is not significantly different from are not significantly different from but not statistically significantly different not statistically significantly different from statistically significantly different from zero does not significantly affect the is not significantly related to not significantly different across takeover not significantly different from one significantly different across takeover amendment significantly different from zero for significantly different from zero in
29 17 10 7 7 7 3 3 3 3 3 3 3
This peculiarity of significantly can be seen even more clearly by focusing on how it is used as an intensifier. An analysis of the adjectives and adverbs it qualifies, summarized in Table 8, shows that economics, although using the expression remarkably more often than history, has a very similar range from a quantitative point of view. It also shows that the limited use of the adverb in history is mostly linked to comparative adjectives and verbs, whereas economics makes greater use of a fixed set of gradable adjectives.3 Once again the general trends studied in this section can be related to the nature of the disciplines when looking at variation from the point of view of crossdisciplinary comparison. Attention to abstraction in economics can be related to . Figures in brackets in this and the following table represent frequencies (>1) of lexical items; when no figure is provided, the frequency is 1.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.13 (877-962)
Emphatics in academic discourse
Table 8. Significantly as an intensifier Comparative adjective/adverb
Gradable evaluative adjective/ past participle
History
different (9), more (8), higher (6), lower (5), less (3), greater (3), shorter, lesser, harder, busier, larger
cautious, absent, beneficial
Economics
different (141) higher (59), lower (30), greater (28), more (24), less (23), larger (14), smaller (6), closer, below
positive (45), negative (35), exposed (16), procyclical, ameliorated
the dominance of reference to statistical norms and in a consequently limited use of emphasizers proper. Attention to detail and process in history may be related not only to the much higher and much more varied use of emphatics already noticed above, but also to the greater interest in shades of polarized elements and in the wider use of pre-modification. . Collocation and “semantic preference” A closer study of collocation and patterns of semantic preference of the adverbials also reveals variation across disciplines. The analysis was restricted to three adverbials only – significantly, undoubtedly and invariably – each representing one parameter of evaluation (importance, certainty and expectedness respectively). Concordances were studied in order to identify preference for particular types of processes. On the whole, verbs of state (be) greatly outnumber other types of verbs. This is particularly the case with economics: in the case of undoubtedly, for example, about 43.33% of the occurrences collocate with be (61.54% of which in inferential patterns) whereas history is limited to 26.43% (equally divided between contrastive and inferential patterns). A very preliminary cross-disciplinary conclusion that can be drawn from this is that economics seems to privilege emphasis on claims, whereas history is more interested in emphasizing trends. When focusing more specifically on process types, the two basic categories identified were processes of ‘change’ or ‘effect’ (increase, reduce, influence, etc.) and processes of ‘cognition’ or ‘exposition’ (relate, associate, explain, describe, etc.). Table 9 below provides an overview of the lexical items and the semantic areas involved for the three adverbs selected. The table illustrates marked differences in the preferences shown by the three adverbs. Significantly is equally associated with processes of change and of exposition: reference to processes of change is dominant in both disciplines, but much more so in economics than in history. This tendency is even more noticeable if we keep in mind that economics, as we will see below, generally makes greater
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.14 (962-981)
Marina Bondi
Table 9. Semantic preference: process types Adverb
‘Change/effect’ HISTORY
ECONOMICS
‘Cognition/exposition’ HISTORY ECONOMICS
significantly
differ(11), increase (7), alter (3), change (3), affect (2), expand (2), influence (2), enlarge (2), increase (2), alter, worsen, diverge, grow, expand, fall, improve, advance, depart
increase (30), affect (29), change (24), reduce (13), influence (8), impact (5), decrease (3), increase(3), differ (2), vary (2), compensate, contribute, deviate, fluctuate, simplify, tilt, influence, expand, improve, alter, modify.
contribute(15), correlate (4), draw on (2), figure in
relate (25), correlate (13), associate (5), explain, appreciate
undoubtedly
influence (3), cause (2), contribute to, discourage, distort, enhance, produce, is responsible for
contribute to (2), decrease (2), strengthen (2), help (2)
–
–
invariably
–
lead (2), produce, follow from
label (2), deem, describe, designate, portray, signify, know as, signify, emphasise as, appear as
–
use of metadiscursive contexts than history. Undoubtedly, on the other hand, is exclusively used with reference to change processes in our corpus. Furthermore, the table shows an interesting semantic preference for ‘exposition’ surrounding invariably in history texts. Data show that it is conveyed by the co-occurrence of the adverbial with verbal forms of ‘description’, such as labelled as, described as and appear as. However, it would also be appropriate to point out that this semantic preference goes hand in hand with an overall negative semantic prosody. Words which are shown to have a distinct semantic preference are sometimes affected in their meaning and they take such “aura of meaning” on themselves. Louw (1993), for example, shows that words like utterly – normally occurring in context with negative meanings – are heard as ironical when found in positive contexts. This is referred to as “semantic prosody”, and identified by Sinclair (1996: 87) as distinctly “attitudinal, and on the pragmatic side of the semantics/pragmatics continuum”. More specifically, we can see that 80% of
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.15 (981-1076)
Emphatics in academic discourse
the terms creating the semantic preference of ‘description’ occur in contexts where the person or object concerned is qualified in an unpleasant way. This is illustrated in the two examples below: (4) While wet nurses’ employers occasionally lauded their employees’ beneficial product, they invariably deemed wet nurses themselves impossibly troublesome – linking breastfeeding with immoral, unworthy women. (JOSH) (5) Similarly, menopause is invariably described as a diminishment of a woman’s biological potential, not as a positive change and a redirection of the body’s biological resources. (GAH)
The notion of semantic preference can be extended to pragmatic units and other elements of the relevant co-text. One major dimension to be explored could be the world of reference or the plane of discourse of the co-text. When looking at how adverbs were used, one relevant issue was, for example, whether – irrespective of the specific process they qualified – they were used in statements on discourse and the community or in statements on data and their interpretation. This analysis was meant to confirm trends observed elsewhere (Bondi 2005; Bondi 2007), which highlight a greater role of self-referential statements in economics compared to history and a tendency for history to be focused on factual narrative. A look at general figures for undoubtedly shows that the adverb is mostly used in statements on discourse and the community, accounting for 22/31 occurrences (70.9%), whereas in history the same adverb mostly refers to statements on data and their interpretation: 93/139 occurrences (66.9%) qualify statements about the object of disciplinary study. Similarly, invariably tends to be associated with statements on discourse and the community in economics (12/17 occurrences, i.e. 70.5%), whereas it is mostly associated with the object of study in history (54/72 occurrences, i.e. 75%). The data confirms the interest shown by economics in highlighting statements on discourse and the community and by history in highlighting statements on data and their interpretation. The trend is also highlighted by other markers: metadiscursive occurrences of there is no doubt that. . . /it is significant that and similar phraseology suggest a slight tendency of economics to privilege statements about discourse and a slight tendency of history to privilege statements about data. A look at phraseology also confirms that there is a clear tendency in historical discourse to make use of a much wider range of tools for emphasis, as shown in Table 10 below, where the frequencies of a number of phraselogical options are given. The quantitative data on these emphatics should be seen against the backdrop of general trends in disciplinary discourse. In general, emphasis on the discourse community and accepted methodologies is much greater in economics, whereas history emphasizes the reader’s direct contact with facts and their logical interpre-
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.16 (1076-1158)
Marina Bondi
Table 10. Selected phraseology Selected phraseology
Economics
History
it is significant that. . . There is / was no doubt. . . . . . that. . . There is little doubt that There can be no doubt that. . . There can be little doubt that. . . There is no reason to doubt. . . There is, however, hardly any doubt that. . .
3 2 1 1 0 0 1
14 24 6 11 11 3 0
tation (cf. Bondi 2005). Use of emphatics, however, is quite clearly meant to attract the reader’s attention to issues that play a major role in the line of argument of the writer and is thus more often related to references to one’s own discourse or to debate within the disciplinary community. . Pragmatic and textual functions: Focus on significantly, invariably and undoubtedly Emphatics clearly act as highlighters of key points in the line of argument of the writer. The functions they take on may often be related to their basic semantic potential and to the evaluative parameters they express. Significantly has been taken as an example of an adverb potentially referring to the parameter of importance, even if we have noticed that this is often interpreted in statistical terms in economics. A closer look at the co-text of significantly will show that, in economics, it is mainly used as highlighter of significant findings, but there is also a clear association with other emphatics and metadiscourse signalling inference and claims. The adverbial often collocates with other forms of “selfprojection” highlighting a shift from data interpretation to conclusion drawing. Among the collocates that precede it, we find numerous reference to findings: With respect to Conjecture 2B we find that behaviour is indeed significantly more competitive in Extra in the case of Cournot markets It is interesting to note that Whites significantly improve their cognitive skills as they grow older. . . Interestingly enough, * is significantly different from 1 at the 10% level for the 3-month data. . . We tested this by one-tailed Mann-Whitney U-tests and found that average quantities are significantly higher in BASIC BERTRAND than in EXTRA BERTRAND. . . The results are significantly modified if the demand for fiscal services is price-elastic. The four main findings of the paper are: (1) survival patterns differ significantly across specific industries. The implication is that redistribution fully or (in one case, significantly), compensates for the differences. . .
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.17 (1158-1205)
Emphatics in academic discourse
The results show that high-growth firms have significantly lower debt/equity ratios and dividend yields compared to. . . . . .the evolutionary act of creating a paradigm can lead to a result that deviates significantly from that ideal state. . . The table shows that in response to both more generous and longer benefits, the share of good jobs increases significantly. . . .Table 13 indicates that the pre-succession performance of the nonfamily and outsider successor firms is significantly lower than that of the family successor firms. More importantly, abnormal returns are significantly positive for firms that are below the median market value. . . . . ., and we test whether PY is significantly higher in it than in the previous and following periods. Our empirical findings indicate that the premium attached to voting stock is positively and significantly associated with the control value. . . At the very least, our results are significantly impacted whether we use lagged, contemporaneous or forward managerial ownership levels.
Other collocations following the adverbial reveal that the findings are then used for inference drawing: The relative value of commodities and the precious metals changed significantly. Thus the author of the hugely impressive study of the Spanish inflation estimated that. . . The coefficients for the lagged variance terns are not significantly different form zero, suggesting that the sizes of current and previous period residuals are not strongly correlated. . . .the estimated value of 1 is not significantly different from zero. This is consistent with our expectation that the relation should be weaker for firms in the low persistence environment. Furthermore, low load factors [. . .] are significantly associated with less differentiation in departure times. Overall, it appears that the predictions of location models with exogenous prices are supported by the results from the 1975 data. Given that most empirical estimates [. . .] are significantly less than 30, our analysis suggests that the stock market value is likely to be higher under a money rule, and. . . . . .markups of high CR4 industries are significantly procyclical. There are, however, interesting differences in the dynamics of the response of markups across the monetary measures. . . . . .the non-US country-specific portfolios are often found to be significantly exposed. These findings may be attributed to differing regulatory and supervisory requirements. . . . . .beta significantly decreases. Further analysis indicates that the size of the pre-disclosure beta, the amount of the abnormal return, the market value of the equity of the type of firm significantly affect the difference between post and pre-disclosure betas. None of the three sets of results show that education contributes significantly to individual wages. . .. This is sharply contrasted to the finding that the average rate of return to education is 12.8% for other Asian developing countries and 14.4% for all the developing economies that have been studied.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.18 (1205-1247)
Marina Bondi
Debit card growth could significantly change consumer payment patterns. We provide an analysis of debit and potential debit users in Table 2. The figure shows that the estimated correlation [. . .] fluctuates significantly throughout. Several explanations have been advanced for the sign of the correlation. The difference [. . .] is significantly different [. . .]. One possible explanation for the slope sign change is that this point represents an equilibrium ownership condition. . . .it is significantly different from zero [. . .] as well as one [. . .]. Therefore, we reject the hypothesis of the Nash behaviour (t-stat of 6.72) as well as Cartel behaviour (t-stat of 3.68). The Spearman correlation [. . .] is significantly positive (1% level). Together, these findings support the hypothesis that the probability that a firm has a completely independent and active audit committee is positively related to firm size. . . .we find that more information makes markets significantly more competitive, supporting the imitation hypothesis. This simplifies the model significantly, but the assumption also carries some strong implications for the results.
An analysis of significantly in history provides similar results, but the pattern expands on a wider co-text, often requiring more than five lines of concordance co-text. By extending the context, it is easy to see that the main function for the adverb is to highlight significant findings, but also that the pattern is complicated by lists and narrative sequences: More significantly, Homberg used his instrumental expertise to work out in practical terms Boyle’s concern with the material and transmutable elements of chemistry. Thirdly, and more significantly, the cotton unions’ choice of constituencies to contest showed poor judgment. And most significantly, industrialists’ fear of diminishing profits played and preyed upon the long-standing fear of unrestrained women. Moreover, and perhaps more significantly, Bauer rejected emancipation despite his willingness to think of Jews in religious, rather than, for example, national terms. Significantly, the submission and humility of Jesus is emphasized in the frescoes. The first scene depicts not the moment of the institution of the Eucharist but Jesus receiving Judas’s denial of his betrayal. In the second, he kneels to wash the Apostle’s feet. In the next. . . Most obviously, it is a catalogue of books published in London, and thus significantly excludes important publishers in Glasgow and Edinburgh, like [. . .]. In addition, the Publisher’s Circular, from which the data are drawn, did not provide either a full or a representative sample of publications. There is no reason to presume Bigelow’s use of ether differed significantly from the norm. Indeed, later in the century, Bigelow was an ardent defender of individualist therapeutics when reformers at Harvard wanted to increase the laboratory requirements in the medical curriculum.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.19 (1247-1301)
Emphatics in academic discourse
. . . it is not surprising that he used London as a platform [. . .]. Similarly, it is not surprising that the proportion of Gladstone’s London speeches delivered while he was in office (43 per cent) was significantly greater than the proportion of for his speeches in the rest of Great Britain [. . .]. Her claim to the full authority of Augustus is most significantly expressed in the occasional use of male titles [. . .]. However, Kantorowicz’s conclusion on [. . .] seems over-optimistic. At least to some extent, they were able to transgress normal restrictions for women; most significantly in the jurisdictional capacity granted them, and also in the fact that teaching and spiritual guidance [. . .] could be seen as female prerogatives. For just as recourse to witnesses’ depositions was significantly more common in Exchequer, so too was the sending of issues of fact in equity cases to be tried before juries at common law.
Cross-disciplinary comparison thus seems to suggest that significant findings tend to lead to inferential reasoning in economics, whereas they become part of listing and contrastive patterns in history, problematizing data and highlighting claims. If we consider invariably and undoubtedly, we can easily relate the meaning potential of each to the parameter of evaluation we started from: certainty, expectedness, importance. It is possible to relate invariably to the parameter of expectedness, where the credibility and value of an utterance is emphasized by the predictability and regularity of the trend qualified by the adverb. In the case of undoubtedly, on the other hand, the dominant parameter will be that of certainty, clearly related to the meaning potential of the adverb, with its explicit reference to epistemic stance. One of the most common intra-sentential functions of invariably is that of highlighting consistency or inconsistency within a sentence, as shown in Examples (6) and (7). (6) This practice is puzzling. If MNE and HC have similar discount rates, why does the reduction in tax rates invariably take this form rather than a uniform reduction over time? In Section 2 we argued that HC’s discount factor is typical. . .. (EJOIO) (7) . . . many such lecture courses represented an important trend towards academic democratisation throughout the eighteenth century, both within and outside British universities. Not only were women involved in this trend but by the turn of the century such lectures were almost invariably open to them. For example ... (GAH)
When we look at the inter-sentential uses of the adverb, we notice that it acts as a predictive element in forms of prospection. It can be used, for example, in highlighting a generalization predicting a list of specific examples, as in Example (8). (8) In Canada, Combined Universities CND argued that ‘the damaging of our children, and [of] countless generations to come, is nothing short of crim-
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.20 (1301-1379)
Marina Bondi
inal No one has the right to do these things’. In the United States, SANE’s dramatic newspaper ads almost invariably developed similar themes. ‘What are the risks of tests?’ asked its 10 April 1962 ad in the New York Times. It replied: ‘Radioactive fallout will increase, endangering our lives and especially the lives of our children’. Another SANE ad that year, featuring a pregnant woman, proclaimed: ‘11/4 Million unborn children will be born dead or have some gross defect because of Nuclear Bomb (GAH)
Quantitative analysis of these functions, highlighting patterns of analogy and contrast, as well as general-specific sequences, shows that invariably is used in interestingly different patterns across disciplines. Table 11 provides the basic figures. Table 11. Pragmatic functions of invariably Pragmatic functions of invariably
Economics
History
Highlighting consistency Highlighting inconsistency Predicting list of specific examples
8/17 (47.1%) 7/17 (41.2%) 2/17 (11.7%)
51/72 (70.8%) 9/72 (12.5%) 12/72 (16.7%)
The adverb is shown to be mostly used as highlighter of patterns of consistency and inconsistency across the disciplinary spectrum, but a clear trend emerges showing a much greater interest of history in highlighting consistency of facts and processes, as against an almost equal distribution across consistency and inconsistency in economics. A similar analysis of the functions of undoubtedly shows that it is often used to highlight that the writer is stating the obvious, as a premise/conclusion to further argument: not so much what should be known, as what should be easily inferred. Use of the adverb is often related to sequences of (a) explanation (cause/effect; general/specific); (b) matching/contrast. More specifically, we have identified a major function in emphasizing logical inference or specification, as illustrated in Examples (9) and (10). (9) But the very insidiousness of the process made its causes harder to discern; Malestroit was undoubtedly misled into thinking debasement the more important problem, and it is Bodin who deserves credit for pinpointing the increase of precious metals as the real issue. (HOPE) (10) ‘Some of the wordings in programmes and decisions of the Social Democratic Party seemed to be inspired by Kvinnors liv och arbete’, wrote Edmund Dahlström, one of its authors. Undoubtedly there were now clear connections being made between academic research and the political climate. (GAH)
The specification is often accompanied by contrast, so that the adverbial highlights inconsistency with a generalization offered and functions as a qualification of the general statement, as in Example (11).
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.21 (1379-1444)
Emphatics in academic discourse
(11) Inventors tend to file for patents if the expected benefit exceeds the cost. In efficient capital markets, the creator of a useful invention can borrow to finance the patent and its development. Women inventors, however, undoubtedly faced greater obstacles in obtaining funding for their inventions, and might not have been able to afford the patent fee and application process, which could amount to as much as $100 (about one-quarter of average annual non-farm wages in the late nineteenth century). (JOIH)
This contrastive element often constructs more complex sequences where undoubtedly acts very much as a marker of concession followed by contrast, as in Examples (12) and (13). (12) Finally, it would be interesting to extend this model to an infinite-horizon setting. Although we believe that our two-period model captures the essential intertemporal tradeoffs that the central bank faces, an infinite-horizon environment undoubtedly would yield more general, and richer, sets of conclusions about the central bank’s instrument-choice problem in a real-world setting with no ”concluding” period. We leave these and other interesting issues for future research. (JOEB) (13) When a new crisis hits, the previous generation of models is judged to have been inadequate (p. 58). Undoubtedly, each crisis has certain distinctive features and peculiarities. However, in light of Rodrik’s observation, it is important to determine what – if any – common elements exist between some or all of these crises, and to develop a general framework that captures these important commonalties. (JOIH)
Both examples show quite clearly that use of emphatics does not simply signal writer’s stance, but also positions the reader, by showing temporary agreement with a claim that is then clearly refuted by what follows. The reader is offered recognition, but is also led to accept the writer’s claim. Quantitative analysis of the functions listed above shows that the contrastive meanings (which may be classified as more “reader-oriented”, or more dialogic, in that they presuppose different interpretations) and the inferential/specifying meanings (more “writer-oriented” or more monologic and focused on the internal logic of the exposition) are fairly balanced. See Table 12 for the data.
Table 12. Pragmatic functions of undoubtedly Undoubtedly
Economics
History
Emphasising logical inference/specification Emphasising contrast Concession and contrast
15/31 (48.4%) 8/31 (25.8%) 8/31 (25.8%)
78/139 (56.1%) 32/139 (23.1%) 29/139 (20.8%)
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.22 (1444-1517)
Marina Bondi
Table 13. Textual position of undoubtedly and invariably Textual macro-structure Introduction Body Conclusion
Economics undoubtedly invariably 4/31 (12.9%) 22/31 (70.9%) 5/31 (16.2%)
3/17 (17.6%) 11/17 (64.7%) 3/17 (17.7%)
History undoubtedly invariably 9/139 (6.5%) 117/139 (84.1%) 13/139 (9.4%)
7/72 (9.7%) 61/72 (84.7%) 4/72 (5.6%)
Reader-oriented use of undoubtedly can be seen to be slightly higher in economics, but on the whole there is no major difference across the disciplines in the balance between writer-oriented and reader-oriented uses of the emphatic. The last phenomenon studied was the distribution of adverbs in texts. Using a rough classification of the text sections into introduction, body and conclusion, the distribution of adverbials across the sections was studied. The quantitative data are reported in Table 13. One major drawback with this kind of calculation is that the introduction as such is much more clearly marked in economics than in history; on the whole, however, they do not differ significantly from a quantitative point of view. The patterns of variation they highlight can therefore be attributed some degree of reliability. Keeping in mind that introduction and conclusion correspond roughly to 10% of the whole text on average, the data can lead to an interpretation of variation: adverbs are distributed rather regularly in historical discourse, with a slight tendency for higher figures in openings, whereas in the economics corpus they are clearly more frequent in introductions and conclusions than in the body. The data thus show that, in economics, these two adverbials are more often used in sections which are also typically related to discussion of the literature and reference to the discourse community.
. Conclusion The analysis of frequencies and patterns has shown that the use of emphatics in history is much more varied and graded than in economics.4 Economics is characterized by rather limited use of emphasizers proper, as well as by more formulaic use of language, whereas history has greater interest in shades of polarized elements and greater use of pre-modification. Patterns of semantic and pragmatic preference also reveal different trends in history and economics, both in terms of the processes emphasized and in terms of the world of reference of the moves em. Comparison across disciplines, of course, may always be made problematic by the definition of the discipline itself, which can be identified at different degrees of delicacy and homogeneity. This does not, however, make the quantitative and qualitative differences observed less relevant.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.23 (1517-1585)
Emphatics in academic discourse
phasized, which are more frequently self-referential in economics. The analysis of moves in which emphasis is found also suggests that the emphasis may be placed on highlighting different features: the significance of findings, the ease of inferability or the dialogicity of patterns of contrast and concession. There is, however, noticeable variation across the span of emphatics considered. Emphatics are shown to signal “engagement” as well as “stance”: they contribute to positioning one’s research in the context of disciplinary debate and to highlighting the significance of the data or the conclusions produced, thus becoming resources by which the author negotiates (engages with) the various convergent or conflicting positions. In research-based genres, they contribute to positioning one’s research in the context of disciplinary debate and to highlighting the significance of the data or the conclusions produced by the writer. Lexical choices and patterns are also related to the epistemology of the disciplines examined. Economics, with its emphasis on simplification, abstraction, as well as contrastive sequences focusing on discourse participants, is clearly inspired by a “rhetoric of inquiry” which identifies well-defined sections in a research article, typically organized around the patterns inspired by natural sciences (Introduction, Methods, Results and Discussion). History, with its emphasis on accumulation and interpretation of factual data, as well as causal sequences focusing on the research object, is more clearly inspired by a “rhetoric of narrative”, where readers are confronted directly with data and sequences of events and processes.
References Biber, D., Conrad, S. & Cortes, V. 2004. If you look at. . .: Lexical bundles in university teaching and textbooks. Applied Linguistics 25(3): 371–405. Biber, D., Conrad, S. & Reppen, R. 1998. Corpus Linguistics: Investigating language structure and use. Cambridge: CUP. Biber, D., Johansson, S., Leech, G., Conrad, S. & Finegan, E. 1999. Longman Grammar of Spoken and Written English. London: Pearson Education. Bondi, M. 2007. Authority and expert voices in the discourse of history. In Language and discipline perspectives on academic discourse, Fløttum, K. (ed.). Newcastle: Cambridge Scholar Publishing. 66–88. Bondi, M. 2005. Metaargumentative expressions across genres: Representing academic discourse. In Dialogue within Discourse Communities. Metadiscursive perspectives on academic genres, J. Bamford & M. Bondi (eds), 3–28. Tübingen: Niemeyer. Bondi, M. 2002. Attitude and episteme in academic discourse: Adverbials of stance across genres and moves. Textus, P. Evangelisti & E. Ventola (eds), 15(2): 249–264. Conrad, S. & D. Biber. 2000. Adverbial marking of stance in speech and writing. In Evaluation in Text: Authorial stance and the construction of discourse, S. Hunston & G. Thompson (eds), 1–27. Oxford: OUP.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.24 (1585-1693)
Marina Bondi
Connor, U. & T. Upton (eds). 2004. Discourse in the Professions. Perspectives from corpus linguistics. Amsterdam: John Benjamins. Crismore, A. 1989. Talking with Readers: Metadiscourse as rhetorical act. Frankfurt: Peter Lang. Del Lungo, G. & E. Tognini Bonelli (eds). 2004. Academic Discourse: New insights into evaluation. Frankfurt: Peter Lang. Halliday, M. A. K. 1985. An Introduction to Functional Grammar. London: Arnold. Holmes, R. 1997. Genre analysis and the social sciences: An investigation of the structure of research article discussion sections in three disciplines. English for Specific Purposes 16: 321– 337. Hunston, S. & Thompson, P. 2000. Evaluation in Text: Authorial stance and the construction of discourse. Oxford: OUP. Hyland, K. 2006. Disciplinary differences: Language variation in academic discourses. In Academic Discourse across Disciplines, K. Hyland & M. Bondi (eds), 17–45. Frankfurt: Peter Lang. Hyland, K. 2005. Stance and engagement: A model of interaction in academic discourse. Discourse Studies 7(2): 173–191. Hyland, K. 2004. Engagement and disciplinarity: The other side of evaluation In Academic Discourse: New insights into evaluation, G. Del Lungo & E. Tognini Bonelli (eds), 13–30. Frankfurt: Peter Lang. Hyland, K. 2002. Directives: Power and engagement in academic writing. Applied Linguistics 23(2): 215–239. Hyland, K. 2001. Bringing in the reader: Addressee features in academic articles. Written Communication 18(4): 549–74. Hyland, K. 2000a. Disciplinary Discourses: Social interactions in academic writing. London: Longman. Hyland, K. 2000b. Hedges, boosters and lexical invisibility: Noticing modifiers in academic texts. Language Awareness 9(4): 179–197. Hyland, K. 1998a. Hedging in Scientific Research Articles. Amsterdam: John Benjamins. Hyland, K. 1998b. Boosting, hedging and the negotiation of academic knowledge. Text 18(3): 349–82. Hyland, K. & Tse, P. 2005. Hooking the reader: A corpus study of evaluative that in abstracts. English for Specific Purposes 24(2): 123–139. Lemke, J. (1998) Resources for attitudinal meaning. Evaluative orientations in text semantics, Functions of Language 5(1): 33–56. Louw, B. 1993. Irony in the Text or Insincerity in the Writer?: The Diagnostic Potential of Semantic Prosodies. In Text and Technology: In honour of John Sinclair, Baker, M., Francis, G. & E. Tognini-Bonelli (eds). Amsterdam/Philadelphia: John Benjamins. 157–176. Markkanen, R. & Schröder, H. (eds) 1997. Hedging and Discourse. Approaches to the analysis of a pragmatic phenomenon in academic texts. Berlin: De Gruyter. Merlini Barbaresi, L. 1987. “Obviously” and “certainly”: Two different functions in argumentative discourse. Folia Linguistica 21: 3–24. Myers, G. 1989. The pragmatics of politeness in scientific articles. Applied Linguistics 10: 1–35. Poos, D. & Simpson, R. 2002. Cross-disciplinary comparisons of hedging: Some findings from the Michigan corpus of academic spoken English. In Using Corpora to Explore Linguistic Variation, R. Reppen, S. Fitzmaurice & D. Biber (eds), 3–23. Amsterdam: John Benjamins. Precht, K. 2003. Great versus lovely: Stance dfferences in American and British English. In Corpus Analysis: Language structure and language use, P. Leistyna & C. F. Meyer (eds), 133–151. Amsterdam: Rodopi.
JB[v.20020404] Prn:11/04/2008; 10:43
F: SCL3103.tex / p.25 (1693-1741)
Emphatics in academic discourse
Quirk, R., Greenbaum, R., Leech, G. & Svartvik, J. (eds) 1985. A Comprehensive Grammar of the English Language. London: Longman. Scott, M. 1998. Wordsmith Tools. Oxford: OUP. Silver, M. 2003. The stance of stance: A critical look at the ways stance is expressed in academic discourse. Journal of English for Academic Purposes 2(2): 359–374. Sinclair, J. 2005. What’s in a phrase. Lecture held at the University of Modena and Reggio Emilia, 15 November, 2005. Sinclair, J. 2004. Trust the Text: Language, corpus and discourse. London: Routledge. Sinclair, J. 1996. The search for units of meaning. Textus 9(1): 75–106. Swales, J. 1990. Genre Analysis. Cambridge: CUP. Swales, J. & Burke, A. 2003. “It’s really fascinating work”: Differences in evaluative adjectives across academic registers. In Corpus Analysis: Language structure and language use, P. Leistyna & C. F. Meyer (eds), 1–18. Amsterdam: Rodopi. Tadros, A. 1985. Prediction in Text. Birmingham: English Language Research Monographs. Tognini Bonelli, E. & Del Lungo, G. (eds). 2005. Strategies in Academic Discourse. Amsterdam: John Benjamins. Vande Kopple, W. 1985. Some exploratory discourse on metadiscourse. College Composition and Communication 36: 82–93. Wierzbicka, A. 2006. English: Meaning and culture. Oxford: OUP.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.1 (47-118)
Interaction, identity and culture in academic writing The case of German, British and American academics in the humanities Tamsin Sanderson University of Freiburg, Germany
This chapter aims to illustrate one way in which corpus-linguistic methods and specialised corpora can be combined in work on academic discourse. It reports selected findings from a study of social interaction in research articles written by German, British and US-American humanities academics, based on the 1-million-word SCEGAD corpus. While the main interest of the project was in possible cultural differences in academic discourse, statistical analysis was used to examine the influence also of linguistic background, discipline, author age, status and gender on the construction of identity and the encoding of social relations in academic writing. The findings reveal significant cultural differences, but also demonstrate the influence of variables such as discipline, gender and academic status on author-reader interaction and identity construction in scholarly texts.
.
Introduction
Academic writing has traditionally been conceived as a register lacking in personal involvement and explicit references to authorial or reader identity. Positivist theories of knowledge production in the academy encouraged a view of the individual scholar as the mere transmitter of universal, objective truths, rather than the human author of subjective, argumentative texts with a persuasive function (Aikenhead 1996: 9; see also Harding 1991; Hyland 1999). Such views were never more than a socially- and culturally-constructed illusion, an attempt to conjure an objectivity no person can ever possess because of the innately subjective nature of human perception (as noted by philosophers from Bacon in 1620 to Popper in 1979). They have obscured the quintessentially interactive nature of academic prose, with its socially-situated dialogue between author, imagined reader and reader.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.2 (118-157)
Tamsin Sanderson
Social constructivist theories of language have provided a useful foundation for re-examining traditional views of academic writing, since they emphasise the importance of language not just for conveying meaning, but also for structuring and maintaining social relationships (Berger & Luckmann 1966; Gumperz 1982; Le Page & Tabouret-Keller 1985). Thus, academic writing can be seen not as the disconnected transmission of immutable facts, but as a dynamic process of meaning construction within a defined social context. Accordingly, recent studies of academic language consider issues of identity, voice, evaluation, hedging and persuasion, and have begun to investigate specific linguistic features associated with social interaction in various scholarly genres (Duszak 1997; Hyland 1998, 2000, 2005b; Ivaniˇc 1998; Myers 2001; Thompson 2001; Mauranen 2004; Simpson 2004; Swales 2004; Harwood 2005a). With varying levels of methodological sophistication and success, contrastive studies have examined how academic authors from different cultural and linguistic backgrounds construct and manage social interaction (Busch-Lauer 2001; Hutz 2001; Zhabotynska 2001; Fløttum, Dahl & Kinn 2006). If academic writing is not the faceless genre it was long assumed to be, how and where does authorial identity manifest itself? One of the main instruments for indicating personal involvement and identity in academic texts is person reference (Mühlhäusler & Harré 1990). In British and US-American English and German – the languages examined here – the primary person-referential devices are personal pronouns. Scholarly authors make particular use of the first personal singular and plural pronouns and the second personal singular pronoun to construct individual, group and reader identity. Traditional grammatical descriptions assign fixed, and separate, speech-act roles to each person: first person pronouns are said to refer to the speaker, second person to the addressee (see for example Halliday & Hasan 1976; Lyons 1977; Comrie 1981; van Riemsdijk & Williams 1986). However, in naturally-occurring speech, there is no constant, one-to-one correlation between the grammatical form of a pronoun and its referent at any moment of speech or section of text. The same referent can be indicated by a number of different pronouns (I can be ‘I’ at one moment, but part of ‘we’ at the next). Conversely, the same pronoun can indicate a number of different referents (we are all ‘I’, depending on who is speaking). The traditional correspondence asserted between pronoun form and person is therefore little but convenience, since personal pronouns are multifunctional, and their referents context-dependent (Wales 1996: 7). In contrast, discourse-oriented approaches to person reference stress the importance of personal pronouns in encoding and managing social interaction: they are seen as playing a central role in the construction of ‘self ’ and ‘other’ (Malone 1997; Sacks 1992; Schegloff 1996), and are “indicators of the complex relationships between selves and the societies these selves live in” (Mühlhäusler & Harré
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.3 (157-201)
Interaction, identity and culture in academic writing
1990: 47; see also Wales 1996). According to the view adopted here, person reference is thus more than a device for encoding grammatical relations such as number or person: it is one of the central means by which authors relate to their imagined audiences. By considering how person reference is used in context, and which discursive functions it fulfils, we can learn much about processes of identity construction and interaction in scholarly writing. In the context of academic writing, person reference is of particular interest precisely because scholarly texts have so long been construed – or rather, constructed – as impersonal. As Aikenhead notes, the recommendation that academic texts should be impersonal follows from a positivistic view of science as “authoritarian, non-humanistic, objective, purely rational and empirical, universal, impersonal, socially sterile, and unencumbered by the vulgarity of human imagination, dogma, judgements, or cultural values” (1996: 9). In both German and English-language scholarly writing, this impersonality has been connected with avoidance of the first person singular pronoun; this convention has come to be known in German as the ‘Ich Verbot’ [I taboo] (Weinreich 1989: 132; Kretzenbacher 1991: 120; see also Hutz 1997: 232; Gläser 1998: 485). The avoidance of explicit person reference is one way in which academics attempt to conjure an impression of objectivity. However, the adoption of an impersonal writing style, most clearly signified by the avoidance of ‘I’, does not render such texts any less personal, or their authors any more objective. At times, the avoidance of surface linguistic features marking personal opinions as such is somewhat disingenuous. The lulling effect that ostensibly impersonal academic style can have on readers is not always unconscious or accidental: scholars are usually expert writers, who seek to increase the persuasive power and force of their argument by all manner of stylistic devices. By examining person reference in a corpus of academic articles, we can shed light on the relationship between the ideal-type impersonality often demanded of the scholar, and the personal identity inseparable from each academic author. Given that scholarly writing consists of opinion, argument and evaluation, are these features marked as such, or presented as impersonal truths? This tension between a practically and theoretically unattainable disconnectedness, and an actual and inescapable personal reality, conveyed in and through language, is one of the most remarkable features of academic writing.
. Methodological approach In addition to examining pertinent features of academic writing, the study aims to make a positive contribution to the methodology of corpus-based studies of academic discourse. In the section following I explain three main methodological
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.4 (201-261)
Tamsin Sanderson
issues and detail how these issues were addressed in the present study. These issues have been neglected in previous research (for a discussion of previous contributions to the field, see Sanderson forthcoming). First, an empirical study of academic discourse needs to draw upon data samples which are representative of the object of study. It is important to note that representativity is not a function of size. In order for a sample to be representative, it must contain all of the characteristics (that is, variables) present in the wider population, in roughly the same proportions as in the wider population. To be empirically sound, therefore, a sample needs to take account of all the variables present in the population being investigated, not just the one or two variables of particular interest. For work on academic discourse, this means that a study of academics from one particular culture should include for example both men and women, of different academic levels and ages, tenured and untenured. A crosscultural study also needs to take account of all of these variables, for each of the cultures it examines. Thus, in each of the three cultures investigated, the present study samples a broad cross-section of academics of both genders, all ages, at all stages of the academic careers, tenured and untenured, writing in a range of humanities disciplines. The focus of the present investigation is English and Germanlanguage humanities research writing. The corpus therefore had to sample a broad cross-section of work by native-speaker academics from these two groups, and I settled on scholars from Britain, the USA and Germany, who represent the three major English- and German-speaking cultural groups. Men and women, tenured and untenured, at all stages of the academic career, were sampled in roughly equal proportions. Five disciplines were selected, representing a cross-section of humanities research production. I have not sampled texts from one or two disciplines only, since this would not be representative of all humanities disciplines. I do not claim that my results are generalisable to academic writing as a whole, nor even to humanities research writing as a whole, since I sampled only research articles. The results of the present study are however generalisable to humanities research articles, and this is what I claim. Research articles were chosen both because they represent a defining academic genre (Swales 2004: 207) and also because their relative brevity meant that a relatively large number of them could be analysed closely and in their entirety by a single researcher. The precise texts chosen in any one study will of course depend on the area of interest and aims of each particular investigation. What is important, however, is the principled collection of texts in a corpus, as practised here, with a view to ensuring representativity. A second, vital issue considered here was that the data sampled must be generalisable to the larger population under examination. Failing this, the results of even the most well-intentioned study will have only anecdotal value. The current investigation recognises that in order for the findings of an investigation to be generalisable, the texts chosen for the corpus must constitute a random sample, which
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.5 (261-322)
Interaction, identity and culture in academic writing
has to be of a reasonable size and also has to cover the major variables contained in the broader population. In this study, these prerequisites were met through careful construction of the corpus, which is a major advantage of both specialised corpora and of corpora tailor-made for a specific investigation. As demonstrated here, future researchers will need to consider the size of their sample in relation to the total size of the population they wish to examine, and weigh this against their own time and possibly financial constraints. Since most researchers face limited resources, it is better to compromise by choosing a genre that is shorter, or examining only a few central features, rather than reducing the number of texts examined. As the present study shows, such compromises are practically feasible. A third, and final, issue which the present investigation considered and applied fully, was a detailed statistical analysis of the results. Quite simply, the human eye is a poor judge of statistical significance. For this reason, statistical tests must be applied to the findings of corpus-based studies in order to separate real from perceived differences and tendencies, in order to ensure that the conclusions reached in the study are true. Contrastive studies in particular will require a sophisticated grasp of multivariate statistical analysis if they are to discern the relative influence of multiple different variables (culture, language, discipline, etc.) on linguistic production. Presenting results as absolute numbers or a cumulative percentage measure presupposes that there is an equal number of possible occurrences in each subcorpus. In most contrastive studies, however, this condition is not met, because the subcorpora differ in size. The results therefore have to be case weighted, as they were here. The data a researcher selects are crucial to the credibility, reliability and explanatory power of a study. It is vital, therefore, that studies are based upon principled data collections, which are representative of the group or groups being examined and generalisable to the wider population. The present study reflects the author’s awareness that cultural background is not the only variable that shapes written production. This awareness motivated the extensive statistical analysis undertaken here, which was necessary in order to distinguish culture from other influential variables, and to determine the relative influence of individual variables on the various aspects of linguistic behaviour analysed in the study. The exact tests applied to the data are explained further below, but first I turn to a more detailed presentation of the corpus. . The SCEGAD corpus The analysis is based on the Synchronic Corpus of English and German Academic Discourse (SCEGAD), a 1-million-word corpus compiled by the author at the University of Freiburg in 2001–2003 for the purpose of systematically investigating native-speaker academic writing in English (British and US-American) and Ger-
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.6 (322-411)
Tamsin Sanderson
man. The corpus contains the full texts of 100 research articles: 50 were written in German by German academics, and 25 each in English by British and USAmerican academics respectively. SCEGAD therefore enables not only interlingual (German/English), but also intralingual (British/US-American English) and intercultural (German/British/US-American) comparisons to be drawn. The texts were published between 1997 and 2003 in leading journals in the following five humanities disciplines: philosophy, history, folklore, English/German literary studies and English/German linguistics. In addition to being balanced for the native language of the authors, the corpus is also controlled for gender, age and academic status. The authors were divided into six age groups, under 30, 30–40, 40–50, 50–60, 60– 70 and over 70, and four academic status levels, pre-PhD, post-PhD, full professor and emeritus professor. The corpus therefore samples a broad cross-section of humanities scholars, the youngest 28, the oldest 75, of both genders, from a variety of disciplines, who span all stages of the scholarly career from pre-PhD scholar to emeritus professor. Using SCEGAD, coupled with bivariate and multivariate statistical analyses, it is possible to examine the effect of a large number of variables, not just culture, on specific features of academic writing, and to make conclusions that are more likely to be representative of a diverse discourse community. . Phenomena examined and statistical methods The variety of phenomena which can be examined using a corpus such as SCEGAD is endless. The present paper focuses on person reference because, for the reasons outlined above, this feature is of considerable interest in academic writing. Person reference is unusual in that it is a discourse phenomenon that can be identified largely automatically; most discourse features in fact require extensive manual analysis, and this remains a major obstacle to large-scale discourse studies using corpora (see discussion in Hardt-Mautner 1995; Aston & Burnard 1998; Hunston 2004). The analysis considers both the form and the discourse functions of person reference, paying particular attention to the communicative purpose in context. The pronoun forms examined are shown in Table 1, grouped according to formal grammatical categories. The analysis centres on first and second person pronouns, or “interpersonal pronouns” (Wales 1996: 3), since these most clearly fulfil interactive and identity construction purposes. Third person pronouns generally do not serve an interpersonal function, and were therefore excluded from the analysis. However, third person references to the reader along the lines of ‘the reader may well wonder. . .’ were counted, as were oblique authorial self-references in the third person, such as ‘the author wishes to thank x’ or ‘der Forscher wurde aufgefordert’ [the researcher
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.7 (411-412)
Interaction, identity and culture in academic writing
Table 1. English and German-language pronoun forms analysed in the corpus
1st person singular
2nd person singular/plural
1st person plural
Nom. Acc. Gen. Dat. Refl. Nom. Acc. Gen. Dat. Refl. Nom. Acc. Gen. Dat. Refl.
English
German
I me my/mine me myself you you your your yourself/selves we us our/s us ourself/selves
ich mich mein/e/r/n/s/es mir mich Sie* Sie Ihr/e/r/n/s/es Ihnen sich wir uns unser/e/r/n/s/es uns uns
*This is the polite form of the German second personal pronoun: the familiar form, ‘du’, would not be used in a formal context such as an academic text.
was asked].1 In the tables which follow, the figures labelled ‘third person reference’ therefore refer not to personal pronouns, but to third person references specifically to one of the parties in the textual interaction. It should also be noted that, for the second person, singular and plural forms are conflated, because they are formally identical in both languages. The counts for second person reference include not only direct addresses to the imagined reader(s), but also instances of generic ‘you’. The quantitative results are presented in table form, showing relative frequencies and comparative differences in frequency, and also as bar graphs for different subgroups. The graphical representation is intended as a useful complement to the figures presented in table form. The results were case weighted, calculated as mean occurrences per 10,000 words, because the SCEGAD texts are on average 8,937 words long. Presenting the results as mean occurrences per 100,000 words would have artificially inflated the frequency of the phenomena examined here, whereas . The English-language tokens searched for were ‘author’, ‘writer’, ‘reader’, ‘researcher’ and ‘scholar’. In German, the tokens were ‘Autor’ [author], ‘Verfasser’ [author], ‘Leser’ [reader], ‘Forscher’ [researcher] and ‘Wissenschaftler’ [scholar]. Only those tokens which actually referred to the author or reader of the text were counted, not those referring to authors, scholars or readers in general, or some other researcher or reader. In addition, discipline-specific third person self-references to the author as a ‘philosopher’, ‘linguist’, ‘folklorist’, ‘ethnographer’, ‘historian’, etc. were also included in the analysis. All other instances of third person singular references or pronouns which did not construct authorial identity or express the relationship between author and reader were excluded from the analysis.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.8 (412-469)
Tamsin Sanderson
showing results per 1,000 words would have made the phenomena falsely seem far less common than they in fact are. The raw results were subjected to bivariate and multivariate significance tests, in order to discern whether the differences between groups that could be perceived through human observation were indeed valid, and conversely to establish whether seemingly marginal differences in frequency were in fact statistically significant. A standard statistical measure was applied, by which I regarded as significant all results up to the “p is less than 0.05” degree (p<0.05). This means that all results where there was a more than 5% chance that they could have been due to chance were discounted as statistically insignificant. Since the data did not have a standard (Gaussian) distribution, for bivariate tests the Mann Whitney test was used. Mann Whitney is a non-parametric alternative to the two-sample t-test, and is used to compare the medians of two independent groups of sampled data. The test reveals whether two samples are from the same or different populations, expressed as a degree of probability. For the multivariate significance tests (three or more independent variables), first the Kruskal Wallis test was applied in order to discover whether there were any significant differences at all between medians in the data, and then the Mann Whitney test was performed on groups of variables in pairs to establish specific significance levels. In addition to the bivariate and multivariate significance tests, a factor analysis was performed on the data. Factor analysis is used to detect structure in the relationships between independent variables – an important finding for an empirical study, where the sample will contain multiple independent variables. The results of the factor analysis produced a correlation matrix grouped for the variables which clustered together, thereby giving some indication of common ways in which sets of variables affected the observed frequencies. These statistical analyses enabled me to establish which independent variables had a discernable effect on the frequency of the selected phenomena, and which did not, as well as how the variables related to each other. The results of the significance tests were sometimes surprising, in that they were very different from what the naked eye discerned. This emphasises, once again, why statistical analysis is indispensable to all empirical studies.
. Results This section presents selected results from the corpus analysis. The results reveal that linguistic and cultural background, discipline, gender and academic status all exercise a statistically significant influence on person reference frequency. No one variable can completely explain linguistic behaviour; rather, written production is a multicausal phenomenon, influenced by a number of considerations. I discuss these results below.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.9 (469-527)
Interaction, identity and culture in academic writing
30.00 25.00 20.00
English
15.00
German 10.00 5.00 0.00 1st p.sing. pronouns
2nd p.sing. pronouns
3rd p.reference
1st p. pl. pronouns
Figure 1. Person reference according to language group: bar graph (frequency per 10,000 words) Table 2. Person reference: English and German-speaking authors compared (frequency per 10,000 words) f English 1st p. sing. pronouns 2nd p. sing. pronouns 3rd p. reference 1st p. pl. pronouns TOTAL
28.41 3.71 0.42 26.52 59.06
f German (n=1322) (n=97) (n=26) (n=1188) (n=2633)
11.76 0.00 0.14 20.91 32.85
Significance (Mann Whitney) (n=454) (n=0) (n=5) (n=779) (n=1239)
0.003 na* na* 0.042 0.019
*Frequencies in the German subcorpus were too small for the Mann Whitney test to be reliable
. Person reference across languages and cultures As can be seen in Figure 1 and Table 2, it emerges that the English-speaking authors are consistently more likely to employ all forms of person reference than are their German colleagues. This trend holds across both of the English-speaking subgroups when compared to the German-speakers represented in the corpus.2 The differences revealed here are considerable. The most highly significant difference between the two language groups lies in the frequency of first person . For reasons of space, the findings are not listed and discussed separately for each possible combination of the three groups included in the corpus, although this was undertaken in the study on which this article is based.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.10 (527-567)
Tamsin Sanderson
singular pronouns: the English-speaking authors are almost 2.5 times more likely to employ this type of person reference than are their German colleagues (p<0.005). This trend also holds when significance tests are applied to the British and US-American texts separately. In addition, while the English-speaking academics write ‘I’ and ‘we’ with similar frequency, the German authors employ first person plural pronouns nearly twice as frequently as they use first person singular pronouns, even taking into account co-authored papers. The highly significant difference between the two language groups for first person singular pronoun frequency is evidence of the ‘I taboo’ in action. No such taboo seems to operate for the English-speaking academics: indeed, the British scholars represented in the corpus claim individual identity even more frequently than they do group membership.3 When it comes to first person plural pronouns, the two language groups are also significantly different to the p<0.05 level, with the English-speaking academics demonstrating higher frequencies also of this type of personal pronoun. The German scholars are therefore not only far less likely to explicitly refer to themselves as individuals, they are also less likely to claim membership of a group. The differences for the other forms of person reference are no less striking. While the English-speaking scholars in SCEGAD directly addressed the reader as ‘you’, or used the generic ‘you’, on average 3.71 times per 10,000 words, there was not a single second person singular pronoun in the entire German subcorpus. The German scholars in SCEGAD not only refrained from directly addressing the imagined reader with ‘you’, they also preferred to perform the generic ‘you’ function with the indefinite pronoun ‘man’,4 as can be seen by comparing the following two sets of typical examples: (1) In 1946, just after the ending of the Second World War, you can still find the poet Auden . . . (LTUK)5
. For the British authors, the frequencies are 27.95/10,000 words (1. pers. sg.) and 21.13/10,000 words (1. pers. pl.). For the US-American scholars, the relevant frequencies are 28.89/10,000 words (1. pers. sg.) and 31.91/10,000 words (1. pers. pl.). . The relatively frequent use of ‘man’ in German-language academic texts was documented by Trumpp (1998: 191), who found that the indefinite pronoun was used over five times more frequently (p<0.001) in the German-language compared to the English-language research article subsection of her corpus. As two ways of fulfilling a very similar discourse purpose, it makes perfect sense that the indefinite and the second person singular pronoun should occur in roughly inverse proportions. . These abbreviations indicate disciplinary and cultural membership. LTUK, for example, marks a literary studies text by a British author; LTDE a literary studies text by a German author; PHUS a philosophy text by a US–American scolar, and so on.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.11 (567-629)
Interaction, identity and culture in academic writing
(2) Man findet diese Variante, soweit ich sehe, nicht in der Novellistik, sondern im Roman. [One finds this variant, as far as I can see, not in the novella but in the novel] (LTDE) (3) So, you cannot make the conjunction of LOT and CT compatible with psychology. (PHUS) (4) Man kann den Widerspruch durch die Annahme vermeiden, daß L-Andrea und R-Andrea identisch sind. [One can avoid the contradiction by assuming that L-Andrea and R-Andrea are identical] (PHDE)
In addition to more frequently directly addressing the imagined reader, the English-speaking authors in the corpus addressed either themselves or the reader in the third person (‘my reader’, ‘the reader’, ‘the author’, ‘der Verfasser’ [the author]) almost four times as often as their German colleagues. Unfortunately, the frequencies were not high enough for this difference to be tested for significance. Across all three cultures, third person nominal references to the reader served a similar discourse purpose, enabling the SCEGAD authors to initiate a dialogue with the imagined reader without directly addressing them as ‘you’ or ‘Sie’: (5) The reader may well wonder how considerations concerning informational transactions can, as here, be supposed to be at all relevant to the present issue. (PHUK) (6) Der folgende Abschnitt wurde in der Absicht geschrieben, den Leser zu einer rekonstruktiven Überprüfung . . . zu bewegen. [The following section was written in order to persuade the reader to undertake a reconstructive examination of . . .] (PHDE)
There was only one statistically significant intervarietal difference between the two groups of English-speaking scholars: the US-American authors were more likely to refer to themselves, and their audience, in the third person than were their British colleagues (p<0.05). The intervarietal differences are therefore nowhere near as great as those found between the English and German-speaking authors in the corpus. In sum, the English and German-speaking academics in SCEGAD demonstrate significantly different frequencies for all types of person reference analysed, with the exception of nominal forms which were too infrequent to test for significance. There is a consistent trend, which holds across both the British and USAmerican scholars, and across all person reference types, for the English-speakers in the corpus to employ more person reference in their articles than do their German colleagues. The discourse implications of these findings are discussed in more detail below.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.12 (629-679)
Tamsin Sanderson
. Person reference across disciplines Although the initial interest of the research project was in cross-cultural differences in academic discourse, since the SCEGAD corpus is balanced for a number of variables in addition to author native language and culture, it is possible to attain a fuller impression of the factors which influence person reference use than is usual in cross-cultural studies. One of these variables is discipline. The results for the British and US-American academics are given in a single column, since the differences between the two English-speaking groups were marginal (see above). The cross-disciplinary comparison, as seen in Figure 2 and Table 3, reveals that in all disciplines analysed except for history, person reference frequencies are higher in the English-language texts. This is as expected, based on the overall results examined above. However, within the general trend, there is considerable variation across disciplines: the cross-disciplinary differences between language
160.00 140.00 120.00 100.00 English Texts
80.00
German texts
60.00 40.00 20.00 0.00 PHIL
LING
FOLK
LIT
HIST
Figure 2. Person reference according to discipline: bar graph (frequency per 10,000 words) Table 3. Person reference by discipline: comparative difference in frequency (frequency per 10,000 words)
Philosophy / Philosophie Linguistics / Sprachwissenschaften Folkore / Volkskunde Literary Studies / Literaturwissenschaften History / Geschichte AVERAGE
f English
f German
Difference in f
148.76 68.94 40.48 25.95 11.16 59.06
77.79 20.25 34.37 7.39 24.44 32.85
70.97 48.69 6.11 18.56 13.28 26.21
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.13 (679-723)
Interaction, identity and culture in academic writing
groups for total person reference frequency are highly significant in the linguistics subcorpus (p<0.01) and significant in both the philosophy and literary studies subcorpora (p<0.05). The high degree of disciplinary variation is not unexpected, for there is a growing number of studies which confirm the crucial influence of discipline on person reference, as well as on engagement, involvement and hedging (Busch-Lauer 2001; Breivega et al. 2002; Harwood & Hadley 2004; Hyland & Tse 2004; Harwood 2005b). These results confirm that any study of academic discourse which does not control for the variable of discipline will paint a distorted and incomplete picture of the phenomena being analysed. The comparison across disciplines reveals some interesting individual findings. One discipline, history, does not follow the general trend. Contrary to the trend across the whole corpus, in the history subcorpus, the German academics employ person reference more often than do their English-speaking colleagues, although this difference is not statistically significant at the p<0.05 level. The explanation for these results may lie in the influence of two important intervening variables: academic status and gender. As we shall see, both high status and male gender are powerful predictors of high person reference frequency, and the German-language history subcorpus contained a higher proportion of articles written by male, high status authors than did the British and US-American subcorpora. The reason for this discrepancy is that the German-language journal published work mostly by higher-status males, whereas the British and US-American journals published a roughly equal proportion of male and female authors from a wide range of status levels. The corpus, which is a random sample of published work, accurately reflects the gender and status distribution found in the total population of articles published in the selected journals in 1997–2003.6 Because the SCEGAD corpus is controlled for a great number of variables, and because each text is marked for all of these variables, it is possible to track the influence each variable has on written production. In this case, what appears to be an exception to the general tendency – that English-speakers demonstrate higher frequencies of person reference – emerges not to be one at all, and instead can be seen to be due to the influence of intervening variables. The most highly significant difference between languages was found in the linguistics texts (p<0.01). This is interesting, since the main reason linguistics was sampled in the corpus was to test the hypothesis, proposed by House and others (Baumgarten et al. 2001; House 2003), that the dominance of English as a lingua franca covertly influences the discourse norms of other languages. The fact that German linguists largely conduct their research (reception and production) . I shall not speculate here on whether the gross gender imbalance is a result of editorial policy, or rather indicative of a smaller number of high-status female academics submitting or writing journal articles in Germany as opposed to Britain and the USA.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.14 (723-788)
Tamsin Sanderson
in English appears not to have affected their retention of native-culture discourse patterns in this case. The difference in person reference frequency between the German and English-speaking authors is also significant in literary studies (p<0.05). Literary studies is the discipline in which the German authors employ the least, and the English-speakers the second least, person reference, which is surprising, since it is certainly not the least subjective of the disciplines examined here. The literary studies scholars sampled naturally express opinions in these texts, but they comparatively seldom mark these opinions with personal pronouns. Contrast the following four typical examples, the first two from literary studies texts, the second two from other disciplines: (7) But the two positions. . . may have more in common than one might initially wish to allow. (LTUS) (8) Betrachtet man die Lyrik des Barock unter diesem Aspekt, dann fällt auch dort die Tendenz auf . . . [If one examines Baroque poetry in this regard, then the tendency makes itself noticed that. . .] (LTDE) (9) In our final analysis, we might think of Ryder’s diary as thirty-five years of psychiatrist’s notes. . . (HIUS) (10) Sehen wir von dem genannten “Mißverständnis“ ab und betrachten wir die grundsätzlichen Deutungsmuster der Theologen, so. . . [If we ignore the “misunderstanding” mentioned above and if we examine the basic interpretative patterns of the theologists, then. . . ] (FLDE)
The most striking result, however, is the high frequency of person reference in the philosophy texts, which was noted also by Hyland in his study of interpersonal metadiscourse (2000: 114). Across all disciplines, the only statistically significant difference which emerges is between the philosophy subcorpus and all other subcorpora.7 This suggests that a cross-linguistic feature of philosophical discourse may be behind the observed pattern. However, when the results are differentiated according to individual types of person reference, a more complex picture emerges. The cross-cultural similarity is in fact restricted to the relatively frequent use of first person plural pronouns (significantly higher in philosophy than in all other disciplines except linguistics and, in the German subcorpus, literary studies) (p<0.005). While the English-speaking philosophers also employ first person singular pronouns more frequently than do . In the English-language subcorpus, this difference in total pronoun frequency between philosophy and all other disciplines is significant when compared to all other disciplines except for linguistics (p<0.005). In the German-language subcorpus, the difference in total pronoun frequency between philosophy and all other disciplines is significant when compared to all other disciplines except folklore (p<0.005).
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.15 (788-862)
Interaction, identity and culture in academic writing
their colleagues in all other disciplines except for linguistics (p<0.005), the German philosophers do not use first person singular forms significantly more often than do their compatriots in any of the other disciplines examined here. Even for the German philosophers, that is, the culturally-specific ‘I taboo’ still operates. At this point, an even more complex picture emerges, which underlines the dominant influence of cultural background over disciplinary membership. It appears that the German and English-speaking philosophers are using first person plural pronouns for different discoursal purposes. The German philosophers most often employ the inclusive ‘we’, meaning ‘we humans’, as part of general statements about the nature of human perception and thought, as in the following typical example: (11) Denn wäre nicht auch das reale Außer-uns der Objekte Inhalt unserer Wahrnehmung, so wäre es für uns nicht nur nichts, sondern wir würden die Objekte auch nicht an einem anderen Ort, als wo wir uns befinden, sehen. [For if the real outside-ourselves of the objects were not also the content of our perception, then it would not only be nothing to us, but we would also not see the objects in a different place from where we are.] (PHDE)
Such cases constitute the vast majority (77.30%) of all instances of first person plural pronouns in the German articles. In the English-language philosophy texts, however, by far the most common ‘we’ referent is not ‘we humans’ (only 12.97% of all cases), but ‘we’ meaning the author and the reader (65.06% of all cases of first person plural pronouns). The following two examples from the English-language subcorpus illustrate how the English-speaking philosophers employ first person plural pronouns not to make general statements, as do their German colleagues, but to express solidarity with the reader: (12) Thus the sentence ‘¬(A – ¬A)’ leads to the contradiction ‘¬A. ¬¬A’, whence we may infer ‘¬¬(A – ¬A)’. If we could delete double negations, we should obtain the law of excluded middle. So, since double negation elimination fails intuitionistically, we can rule it out that any sentence is neither true nor false, but we cannot affirm that every sentence is either true or false. (PHUK) (13) We can, if we like, conjoin or disjoin statements about leptons with statements about trees. If we bring mathematics in, we can count both leptons and trees. (PHUS)
The English- and German-speaking philosophers represented in SCEGAD are therefore not as similar as they may first have appeared. It is true that both groups employ person reference more often than their colleagues in the other disciplines do. However, for the German scholars, this feature is due to the high frequency of first person plural pronouns referring to ‘we humans’, and can be overwhelmingly attributed to the nature of philosophical argument as it is represented in the cor-
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.16 (862-902)
Tamsin Sanderson
pus texts. Only very rarely do the German philosophers employ person reference for interactive purposes. The English-speaking philosophers behave entirely differently. When compared to their colleagues in other disciplines, and in contrast to the German philosophers, the English-speaking philosophers employ not just more first person plural but also more first person singular pronouns. Although the English-speaking and German-speaking philosophers employ first person plural pronouns with nearly identical frequency, in the English-language texts, this type of person reference is not due to any characteristic of philosophical argument, but fulfils an interactive function. Interestingly, this was precisely the function of the inclusive ‘we’ which Hyland detected so often in the English-language philosophy texts in his corpus (2000: 127). We see, therefore, that the same pronoun can serve completely different functions across cultures; numbers are not everything. While the quantitative results may have led us to declare a real cross-linguistic, discipline-specific feature (philosophers generalise with ‘human we’), it emerges that in SCEGAD the influence of discipline on this feature is marginal. It remains to be considered why the English-speaking philosophers in the corpus write in such an explicitly personal and dialogic fashion. The unusually high frequency of both first person singular and plural pronouns in these articles is a feature uniformly represented across all articles in the subcorpus. At times, the argument in these texts takes the form of a stream-of-consciousness-like personal conversation, as the following extract illustrates: (14) Next, and more important for what follows: Could that very vapour have been a solid or a liquid? I do not mean to ask whether the stuff that makes up the vapour comes in solid or liquid form. I know that it does. I mean to ask whether, say, T. S. Eliot’s yellow fog which “rubbed its back against the window pane” could have been or could come to be identical with a block of ice or a puddle of water. I do not mean to ask whether the fog could solidify into a block of ice or liquefy into a puddle of water. I know that it could. I mean to ask whether the fog could be either of these things. It seems to me that the fog could not be either of these things. (PHUS)
When not entertaining a kind of internal dialogue as in this extract, the authors express solidarity with the reader or with the (philosophical) academic community, as in extract (12) and (13) above, or employ person reference to lead the reader through the text, as shown in Hyland’s study (2000: 127). Since the German philosophers do not behave in the same way, it is unlikely that this feature is due to any innate characteristic or requirement of philosophical discourse: it is perfectly possible to write a philosophy article without employing person reference more often than in any other humanities discipline. It seems that the English-speaking philosophers simply grant themselves more freedom: they may be more extreme,
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.17 (902-968)
Interaction, identity and culture in academic writing
but they use the same linguistic devices, in the same proportions, and for the same purposes, as the English-speakers in the other disciplines do. In sum, the corpus analysis reveals that disciplinary affiliation is not as decisive as cultural affiliation when it comes to the use of person reference. Similarities within a discipline which, based on a purely quantitative analysis, initially appeared to be cross-linguistic in nature, emerge in the qualitative analysis to be due to different factors. The English-speaking authors in SCEGAD write in a more explicitly interactive fashion than the German scholars, and the philosophers demonstrate this tendency most strongly. The German philosophers write no more explicitly personal or interactive texts than do any of the other German humanities scholars, and employ first person plural pronouns only to formulate the kind of generalised statements central to the argument of their texts. The cultural differences we have already noted between English and German-speaking scholars therefore remain constant. . Person reference across genders and academic status levels Cultural background and disciplinary affiliation are not the only considerations which influence how authors present themselves and relate to their readers. Two other variables for which the SCEGAD corpus is annotated are gender and academic status. As can be seen in Figure 3, there is a consistent trend, across all subcorpora and the corpus as a whole, for the male scholars to employ person reference more frequently than do their female colleagues. This trend holds across all types of person reference investigated, though it is statistically significant only for ‘total pronouns’ (all pronoun types counted together).8 Within the general trend, however, a comparison of the subcorpora reveals some important differences. While the gender differences for the German and USAmerican scholars are not statistically significant for any form of person reference, the British scholars demonstrate significant gender differences for all person reference types except second person pronouns. The use of person reference is a stylistic feature that appears to be far more highly marked for gender among British than among US-American or German scholars as represented in SCEGAD. The striking findings from the British English subcorpus are not due to one or two anomalous texts, but rather evidence of a consistent stylistic feature of the British women’s writing in the corpus. The results are also not due to the intervening influence of other variables such as discipline. For the British female academics, it seems as though a general person reference taboo operates, and does . The two-tailed asymptotic significance value across all pronoun types is 0.044, using the Mann Whitney test (significant to the value of p<0.05). The gender difference for first person plural pronouns was very nearly significant (two-tailed significance value 0.058).
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.18 (968-996)
Tamsin Sanderson
80.00 70.00 60.00 50.00 M ale
40.00
Female 30.00 20.00 10.00 0.00 German
English
British Authors
American Authors
Whole corpus
Figure 3. Person reference by gender (frequency per 10,000 words)
so more strongly even than the ‘I taboo’ common to both genders of German scholars. In the absence of empirical studies of gender-marking in academic discourse, or even in written discourse generally, it is not possible to speculate why, for the British scholars, there should be such a marked gender difference for this phenomenon. These unexpected results are also difficult to test, because comparable gender-controlled corpora of written academic discourse do not exist. When a factor analysis was performed on the data, gender, age and academic status all formed one factor, with female gender, lower age and lower academic status all equated to less frequent use of person reference. The female scholars in SCEGAD, particularly the British women, therefore demonstrate a feature (lower relative frequency of person reference) associated with lower academic status. This raises some fascinating questions for the study of gendered language. Is Lakoff ’s “women’s language” (1975) in fact really context-dependent “powerless language”, as O’Barr and Atkins’ influential 1980 study suggested? The SCEGAD results do not support such a conclusion, since the communicative situation was constant, and the actual (if not felt) academic status levels were proportionally the same across genders in the corpus (of the British scholars sampled, 30.77% of the men and 30.00% of the women were professors). Yet Morley’s (1995) paper on the situation of female academics in British universities notes that discrimination against women in the academy is still rife, and that British women scholars remain consigned largely to the lowest status and pay levels. Is it possible that linguistic behaviour is slow to change, even when women attain high-level academic status, as have the female scholars in SCEGAD? The results point to a large and worthwhile area for future research.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.19 (996-1044)
Interaction, identity and culture in academic writing
The gender comparison is interesting for a second reason. Research on automatic text categorisation has sought to establish that women’s written language is more involved and personal than that of men: [. . .] even in formal writing, female writing exhibits greater usage of features identified by previous researchers as ‘involved’ while male writing exhibits greater usage of features which have been identified as ‘informational’. (Argamon et al. 2003: 321)
Other researchers see this dichotomous approach as “a gross oversimplification of the complexity of language and gender” (Mullany 2002: 4; see also Graddol & Swan 1989). The research agenda in this area of linguistics is coloured by polarised reactions to the gender stereotype that women are more people-oriented than men, and therefore write and speak in a more personal fashion. Holmes (1993) suggests that there are ‘universals’ of female language, one of which is the greater attention women apparently pay to the relationship-building function of conversation. Those who support such notions frequently search for confirmation of their hypothesis in a higher frequency of person reference in female language (Holmes 1993; Crawford 1995: 30; Argamon et al. 2003). In the study quoted from above, for example, Argamon and colleagues examined various linguistic features, among them pronouns, in the British National Corpus (BNC), and found that female authors employed significantly more person reference than did male authors. However, the BNC is a poor place to search for formal female language: in many written domains, few or no females are sampled, and many of the hits are in fact quotes from various novels under discussion in the corpus texts.9 Argamon et al.’s study therefore cannot claim to be representative of female language as a whole, nor can it claim to present an accurate picture of gender tendencies across different registers. In addition, brazen and unsupported claims that women employ the inclusive ‘we’ more often than men do, because women are more ‘person-oriented’, and supposedly inspired by a cooperative rather than a competitive spirit, could not be confirmed in this study (Lakoff 1975; Maltz & Borker 1982; Bailey 1992). On the contrary, the female authors in SCEGAD use fewer first person plural pronouns than do the males (p=0.058). If Holmes (1993) were right, and women were universally more interested in relationship-building in conversation, this posited
. The natural sciences subsection of the BNC, for example, contains not a single text written by a woman. Since the BNC data were collected in 1991–94, not 1891–94, this can hardly be representative of natural science publications.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.20 (1044-1089)
Tamsin Sanderson
gender universal would have to apply in academic writing too. The SCEGAD results strongly suggest that this is not the case.10 The lesson here is that researchers should exercise far greater caution when advancing ostensibly universal gender tendencies. When theories are advanced that are not based on solid data, researchers do little more than repeat and reinforce unhelpful, even harmful, gender stereotypes. One of very few empirical investigations of putative differences between male and female written language, by Hiatt (1977), demonstrated that a large variety of pronouncements and assumptions regarding women’s writing are false (although many unfortunately endure). Hiatt did discover discernable differences between male- and female-authored texts, and thus a sort of ‘feminine style’. However, this style had nothing whatsoever in common with popular stereotypes of women’s language (1977: 121).11 As regards the current study, continued research is necessary to investigate whether the SCEGAD results are indicative of wider tendencies in scholarly writing, or if they are attributable to some other factor. Certainly, it would be fascinating if academic discourse were an exception to, or even the disproof of, the widespread stereotype that women’s language is more personal than that of men. The results of this study suggest that this could well be the case. Author academic status, which is another of the extra-linguistic features for which the corpus is balanced, also exercises a significant influence on person reference frequency. As mentioned above, the corpus contains articles written by professional academics from all status levels. When statistically tested, there were significant differences between professors (current and emeritus) and nonprofessors for all types of person reference except third person reference (p<0.05). There were no significant differences between the behaviour of current and emeritus professors, or between pre- and post-PhD scholars who were not professors. The appointment to a professorship, it seems, is such an important event in the life of a professional academic that it significantly changes how scholars present themselves, and how they interact with their imagined readers, in their research articles. Again, this serves as a reminder that linguistic behaviour has multiple conditioning factors. Therefore, studies which neglect extra-linguistic variables, such as communicative situation, gender, age and relevant social factors, or control their data only for one or two variables, will not provide reliable results. Language, as . It would have been useful to have tested the SCEGAD tendencies against the BNC, but this is impossible, since the entire BNC contains (an extract from) only one single humanities periodical (The Women’s Art Magazine). . Unfortunately, Hiatt (1977) did not examine first- or second-person pronouns. The breadth of other phenomena she analysed, based on a computerised corpus of 200,000 words of prose from 100 books (50 by men, 50 by women), is impressive nonetheless.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.21 (1089-1163)
Interaction, identity and culture in academic writing
a social phenomenon, cannot be satisfactorily investigated or explained using a monocausal model. . Person reference and discourse strategies We move now to the more qualitative findings. The SCEGAD authors employ person reference for a wide variety of discourse purposes. For the analysis, all instances of person reference were categorised according to the main discourse purpose, and assigned to one of eleven (mutually exclusive) groups. These groups were compiled through a combination of inductive and deductive reasoning. First, I read through all of the articles contained in the corpus twice, marking all occurrences of person reference and the discourse purpose. The first close reading led me to posit a putative framework of discourse strategies for person reference, which was tested on the second reading. I then compared the categories to the discourse strategies discerned by other studies of person reference in academic discourse and other discourse types, to help confirm my inductively made framework. The 11 categories which best explain the discourse functions fulfilled by person reference in the SCEGAD corpus are: 1. ‘Personal opinion’: in order to express a personal opinion or make a personal comment 2. ‘TC’: text comments 3. ‘Group: academics’: for expressing group identity as an academic (historian, philosopher, etc.) 4. ‘Group: author & reader’: for expressing group identity and solidarity with the reader 5. ‘Group: humans’: for expressing group identity as a human being 6. ‘Group: other’: for expressing various other types of group identity, for example as a member of a particular research group, or group of students in personal reminiscences, or for expressing solidarity with the people one is studying 7. ‘Methodology’: for explaining the methodology employed in the article (takes the form of authorial self-reference) 8. ‘Address reader’: for directly addressing the reader (not talking about readers in general; such examples were excluded from the analysis) 9. ‘Self-advertisement’: for referring to, and thus advertising, the author’s previous work 10. ‘Acknowledgement’: for expressing thanks to various groups (colleagues, institutions, family members) for providing data, funding or personal support (among other things) 11. ‘Reference to secondary sources’: for referring the reader to work carried out by other researchers
JB[v.20020404] Prn:11/04/2008; 10:48
Tamsin Sanderson
40% 35% 30% 25% English
20%
German
15% 10% 5%
ro up
ro up :
G
G
on al op i
ni on
TC ac ad em :a ics ut ho r+ re G ad ro er up :h um an M s eth od ol A og dd y re ss re ad er Se lfad v. A ck no w led G . ro up :o th Re er fs ec .so ur ce
0%
Pe rs
F: SCL3104.tex / p.22 (1163-1206)
Figure 4. Person reference by discourse purpose: English and German-language texts (percentage of total occurrences)
Though it is not possible to discuss all eleven discourse strategies in detail here, a few individual categories will be commented on. As can be seen in Figure 4, by far the most common discourse purpose for which the German scholars in SCEGAD employ person reference is to express group identity as human beings. In contrast, the English-speaking authors most commonly use person reference to express a personal opinion, closely followed by text-comments. Both of these strategies are interactive in nature. As we saw above, the ‘human we’ expressed so often by the German authors does not serve an overtly interactive purpose. This more differentiated, discourse-oriented perspective therefore confirms the tendencies noted above. Once again, we see the ‘I taboo’ in action. In the relatively few cases in which the German scholars in the corpus advanced personal opinions explicitly marked as such with ‘I’, there often seems to be a sort of discomfort on the part of the author. This remarkably awkward sentence, from an article by a German historian, shows the author hedging his bets between the passive and the personalised active in referring to the processes of knowledge gathering in his work: (15) Diese Eigenschaften wurden zudem auch Gott selber zugeschrieben. Daraus läßt sich schließen – ich habe es getan – , daß die Herrscher des 10. Jahrhunderts “höchstwahrschein-lich”, bei Handlungen der Huld, Milde und der siegbringenden Stärke ähnliche “Erfahrungen” hatten wie Gregor VII. [These characteristics were also ascribed to god. From this it may be concluded – I have
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.23 (1206-1270)
Interaction, identity and culture in academic writing
done so – that the rulers of the 10th century ‘most probably’ had similar ‘experiences’ to Gregor VII concerning grace, clemency, and victory-bringing strength.] (HIDE)
Across all three cultural groups represented in SCEGAD, one of the most unusual discourse strategies behind person reference use was the enactment of a dialogue with the reader, either through direct personal address with ‘you’/‘Sie’ or through the periphrastic ‘the reader’/‘der Leser’ or ‘my reader’. In the German-language corpus texts, direct second person reader address did not occur at all, and oblique third person reader address occurred only once (see example (19) below). In research articles, a genre normally neither conducive nor amenable to direct personal interaction and involvement, both types of reader address seem unexpected. However, despite the fact that author and reader of a journal article are both spatially and temporally cut off from each other, there are nonetheless instances where academic authors directly address their putative reader(s): (16) If you look back to (97), you will notice that the two States involve stage level predication. In fact, the more ‘stage-like’ the predicate is, the better for the Telic Clause. (LGUK) (17) It is easy to get confused about this, because one imagines someone who has not yet learned to recognize cats reliably looking at a cat and having the same percept as you and I would have. (PHUS) (18) Not that the readers of this publication need to be educated about the politics of Standard English (I suspect I am preaching to the choir). (LGUS) (19) Der folgende Abschnitt wurde in der Absicht geschrieben, den Leser zu einer rekonstruktiven Überprüfung seiner die Extension des Handlungsbegriffs betreffenden Überzeugungen zu bewegen. [The following section was written with the intention of moving the reader to a reconstructive examination of his views regarding the extension of the concept of action.] (PHDE)
Addresses such as these can form part of various discourse strategies. Drawing on the examples above, in the first case, the pronouns are used in a directive: the author is asserting his authority as the architect of the argument, and, in an imperative, claims the right to direct the imagined reader’s attention where he sees fit. In the second example, the author uses person reference to bring the imagined reader on side, with the author assuming or anticipating that the reader shares a particular perception. This is evidence of a more cooperative approach to argument construction. Similarly, in example (18), specific attitudes and knowledge are attributed to the potential audience; the author displays her knowledge of the target audience, and at the same time, pays this audience an implicit compliment. This is at once an elegant way of building solidarity with the reader(s), and at the same time defends the author’s inclusion of material that may be regarded as extraneous
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.24 (1270-1329)
Tamsin Sanderson
or inappropriate to the specific audience. In the final example, the address to the reader comments upon the author’s own attempt at persuasion. By using persuasion as the justification for a section of the article, the author draws the imagined audience’s attention to his explicitly-stated aim. Assuming the reader wishes to be persuaded, this very direct metadiscoursal strategy might succeed; other readers may prefer to be prompted less baldly. The use of person reference for explaining methodology was particularly common in the folklore articles (52.21% of all cases in the English-language texts, 37.50% of all cases in the German-language texts). Particularly in some ethnographically-oriented folklore articles, the researchers come very much to the fore when reporting their findings. Because the process of gathering information is inseparable from the process of becoming acquainted with the subjects interviewed and studied, knowledge gathering becomes personalised, and the texts contain reminiscences about when the author first met informant x or y, or how the author experienced the foreign culture in question. A typical example from the English-language subcorpus: (20) In 1992 when I was translating Sitti Djaoerah into English in Bungabondar (Soetan Hasoendoetan 1997), my translation colleague Baginda Hasudungan Siregar, a ceremonial orator, school principal, and folktale writer and recorder, encouraged me to discuss Angkola heritage with his lineagemate, Sutan Habiaran, down the road. So I would drop by Sutan Habiaran’s house casually several times a week. He was still vigorous in his eighties and was vastly enjoying his retirement from his lumber business in Medan. He had become an antiquarian writer of no small repute, having published over ten small volumes on Sipirok adat (heritage) lore and a tome on the Siregar clan founder “back in Toba” (1974). Sutan Habiaran and I had met on my initial Saturday in Sipirok in 1974 when I was invited out to a village to attend a bone reburial ceremony. (FLUS)
The style here is almost autobiographical in nature: though the text is ostensibly about the life and habits of a specific foreign culture, it contains much information about the author’s life and habits as well. The emphasis on the extensive fieldwork carried out by the researcher may also have a self-promotional aspect (note, too, the self-advertisement in the initial sentence). The author presents herself as an experienced fieldworker, and an integrated member of the foreign community she studies. The inescapably personal nature of much folklore research is also visible in the German texts. Note however that the final case of self-reference in this extract is – characteristically for the German authors in SCEGAD – indirect, couched in the third person:
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.25 (1329-1382)
Interaction, identity and culture in academic writing
(21) Als ich wohlvorbereitet und wie immer adäquat gekleidet eine Händlerin ansprach, die gerade das Börsengebäude zum Dienstschluß verließ, fertigte mich die Dame kurzerhand mit den Worten ab: “Interessiert mich überhaupt nicht“. Sprach’s und ließ den verdutzten Ethnographen stehen. [When I – wellprepared and as ever adequately dressed – addressed a broker just as she was leaving the stock exchange building after work, the woman quickly disposed of me with the words ‘doesn’t interest me at all’. Spoke and left the dumbfounded ethnographer standing there.] (FLDE)
Following the single initial first-person self reference, this author continues in the third person for the rest of his report on his field methods: (22) Der Forscher wurde sogar im Verlauf der Feldarbeit mehrmals ausdrücklich aufgefordert, . . . zu erläutern. [In the course of the field-word, the researcher was even explicitly asked to explain . . . a number of times] (FLDE) (23) Demgemäß kleidete sich der Ethnograph stilistisch wie Börsianer; Krawatte und ärmellose Weste inbegriffen. [Thus the ethnographer dressed in stockbrokerstyle; tie and sleeveless waistcoat included] (FLDE)
The examples above suggest some discomfort on the part of this German author when explaining his methodology. The effect is rather strange, particularly when, as in the last example, the author discusses something as personal as his clothing in the third person. Yet this awkward oblique self-reference is apparently still preferable to using first person singular pronouns. The English-speaking folklorists seem to feel no such discomfort when explaining their methodology; they consistently refer to themselves in the first person. The frequent occurrence of person reference (in this case self-reference) when explaining methodology in the SCEGAD folklore texts is related to the conundrum ethnographers face when investigating and interpreting foreign cultures, of which Geertz writes in Works and Lives: The Anthropologist as Author. According to Geertz, the difficulty lies in “how to get an I-witnessing author into a theypicturing story” (1988: 84), and he recommends that “the most direct way to bring field work as personal encounter and ethnography as reliable account together is to make the diary form [. . .] something for the world to read” (84). In some of the folklore texts in the corpus, in particular those resulting from ethnographic field-work, person reference appears to be evidence of the authors’ attempts to commit themselves to what Geertz calls “an essentially biographical conception of Being There” (84). Since this is a dilemma unique to the social sciences, it therefore seems logical that it is the folklore texts in SCEGAD that show signs of it. Theoretically, the same type of situation could arise in linguistics, which is often counted a social sciences discipline. However, due to the particular journals selected for the linguistics subcorpus, and the fact that each subcorpus was balanced
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.26 (1382-1451)
Tamsin Sanderson
for subdiscipline, none of the linguistics articles in SCEGAD is comparably ethnographic in focus. However, something similar does occur in a history article which adopts ethnographic methods. The article is a case study of the effects of racial categorisation on one US-American family, and is based largely on the author’s ‘fieldwork’ interviews with the protagonists. This is frequently acknowledged in the footnotes: 5
(24)
Genealogy in family Bible in possession of X, and author’s conversation with . . . (HIUS)
(25)
18
. . . author’s conversation with X (b. 1919) . . . (HIUS)
(26)
20
. . . author’s conversation with X (b. 1928) . . . (HIUS)
(27)
31
Author’s telephone conversation with X, Brooklyn, N.Y., November 11, 1996 (untaped) . . . (HIUS)
The difference between this history article and the folklore texts is that here, the personal nature of the knowledge-gathering process is acknowledged indirectly, using the third rather than the first person, and appears only as an annotation, rather than being thematically and spatially privileged as narrative content within the body of the text. The point to note is that this way of using language – reporting on methodology using authorial self-reference – is not confined to a particular discipline, nor is it a feature of any discipline per se. Rather, it is evidence of a specific way of gathering knowledge. As a methodology, field work is by its very nature personal, and this is reflected – to varying extents – in the language used to report on it. . Managing complex persona with person reference Person reference is more than just a tool for implementing various discourse strategies. As a number of scholars have noted, personal pronouns are intimately related to the construction of identity and the maintenance of social relations (Malone 1997; Sacks 1992; Schegloff 1996; Bramley 2001). They are social words that serve social functions. Through personal pronouns, academic authors summon a range of identities (individual, group, gendered, discoursal, professional, cultural) and build relationships with their imagined readers by alluding to a range of shared allegiances (disciplinary, professional, linguistic, cultural). In SCEGAD, these identities and allegiances are most often signalled through first person plural pronouns: Hyland (2005a: 182) also found inclusive ‘we’ to be “the most frequent engagement device in academic writing”. But to whom exactly does ‘we’ refer? Jespersen (1924: 192) observed that “the word ‘we’ is essentially vague and gives no indication whom the speaker wants to include besides himself ”. Certainly, the most striking feature of the first person plural pronouns as they are used in the corpus
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.27 (1451-1501)
Interaction, identity and culture in academic writing
texts is the great diversity of groups that ‘we’ can include: as Wales (1996: 63) points out, the “actual discourse referents for ‘we’ are seemingly limitless”. Put simply, while traditional grammatical frameworks rely on a static relationship between pronoun form and referent, the grammatical distinctions of number and person do not hold in spontaneous speech or writing: in fact, Mühlhäusler and Harré state that “any pronoun can be used for any person” (1990: 133).12 Both author, reader and pronoun can be many things, even at the same time. A grammatical property of personal pronouns in both English and German – referential flexibility – offers the corpus authors all manner of possibilities for their own identity construction and their interaction with the reader. In this section, the focus will therefore be less on cultural contrast, and more on the role person reference plays in identity construction and social interaction in academic writing. First, then, some examples of how person reference is employed to highlight different aspects of authorial identity. The corpus authors often employ person reference to express their identity as part of a specific academic community: (28) As folklorists, we need not relinquish analysis of complex and contradictory subjectivities to scholars of literature and media. (FLUS) (29) Therefore, by applying the transitivity of identity, we philosophers have discovered that snow is numerically identical with water vapor. (PHUS) (30) Man berief sich auf die uns bekannten und als Vorläufer des Faches betitelten Gelehrten Jeremias Gotthelf... [One referred to teachers well-known to us and titled pioneers of the discipline, such as Jeremias Gotthelf...] (FLDE) (31) Gerade diese fuzzy category des Alltags erwies sich als enorm produktiv in der Geschichte unserer Disziplin. . . [Precisely this everyday fuzzy category has emerged as enormously productive in the history of our discipline. . .] (FLDE)
By claiming membership in this way, the SCEGAD authors draw attention to and reinforce their identity as researchers, positioning themselves as part of a traditionally-recognised discipline, and allying themselves with a respected social role as scholars. Foregrounding their membership of a particular disciplinary group can be a powerful means for authors to legitimise their opinions and lend their hypotheses credibility. This is therefore both an authority-invoking and an authority-building strategy. Using the inclusive ‘we’ to mean ‘we researchers’ also strengthens the relationship with the imagined reader, by emphasising shared membership of a particular discourse community. Hyland (2005a: 182) noted an . Of course, referential flexibility is by no means restricted to first person plural pronouns: we saw above how the third person may have a first person referent (‘the author wishes to thank. . .’), and the second person too may refer grammatically to the first (‘you could tell she was annoyed’).
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.28 (1501-1560)
Tamsin Sanderson
“enormous emphasis on binding writer and reader together through inclusive ‘we”’ in the texts in his corpus, and the same is true of the SCEGAD texts. Alluding to this kind of allegiance conjures a cooperative spirit: through the academic ‘we’, the reader is invited to “participate as something more than an audience” (Mühlhäusler & Harré 1990: 129). This strategy encourages the reader to view the text not so much as the individual presentation or personal creation of the writer, but as a contribution to the shared effort of knowledge-building within a specific academic community. Another type of group identity, which is expressed particularly by the linguists in the corpus, is membership of a particular ethno-linguistic group: (32) Als Resultat haben wir im Deutschen die häufig benutzten Hilfs- und die starken Verben, die ja alle sehr unregelmäßig sind. [As a result, in German we have the frequently-used auxiliary and strong verbs, which are all very irregular] (LGDE) (33) Zum einen ist Weltwissen zentral: In unseren Kulturkreisen geht man normalerweise, wenn es regnet, nicht unbedingt los, sondern wartet eher ab, bis der Regen nachlässt. [For one thing, societal knowledge is central: in our culture, one does not usually go outside when it is raining, but waits until the rain stops.] (LGDE)
This strategy is not related to strengthening authority, but to engendering solidarity with the reader through the assumption of a common linguistic or cultural origin. Interestingly, while such comments demonstrate an awareness that other cultures exist, they reveal the author’s assumption that the reader is from the same linguistic and cultural background: should this assumption be false, the strategy will emphasise not commonality, but difference. A foreign reader would have to interpret an intended inclusive ‘we’ as exclusive, and the membership claim will therefore have a distancing effect. It is worth noting that this type of group membership claim occurred only in the German-language texts: possibly the English-speaking scholars represented in SCEGAD assume an international audience, with whom they cannot claim solidarity solely on the basis of shared linguistic or cultural background. Another type of group identity regularly summoned by the SCEGAD authors is institutional identity as a member of a particular research group, as in the following example: (34) Die finanziellen Mittel zur Durchführung des Projekts verdanken wir der DFG, die es unter der Nummer SE 699/3–3 fördert. [We owe the financial means for this project to the DFG, which supports it under grant number SE 699/33.] (LGDE)
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.29 (1560-1612)
Interaction, identity and culture in academic writing
This type of group identity, which is often expressed as part of the acknowledgements section, occurs only in the German texts. This reflects a difference between Germany versus Britain and the USA regarding humanities research funding practices: in Germany, it is common to receive grants not for individual research, but as part of a large research group. The membership of such a research group, and the external funding this guarantees, is prestigious. Thus, in addition to acknowledging much-appreciated funding, group membership claims of this type implicitly build authority as a member of an elite group. Summoning group membership in this way can also serve a protective, hedging function: the statements in the article receive extra weight as the opinions of many people, not just the author. Some of the corpus authors juggle multiple professional identities, in this case as both a linguist and a teacher: (35) In short, if we as teachers do not confront prior understandings and grapple with the ways in which they may conflict with what we hope to teach as ‘fact’, we seem to do little to change people’s knowledge in any lasting or meaningful way. (LGUS)
This type of dual identity expression occurred in all corpus texts with a pedagogical focus, regardless of the discipline or cultural origin of the author. It underlines the authors’ involvement in the activity (teaching) they are writing about, and creates the impression that this is not some ivory-tower academic telling teachers how to behave, but one teacher talking shop with another. The atmosphere becomes more collegial and cooperative, with authors stressing that they have the same concerns as the reader, and that they too stand to benefit from their own suggestions. Academics often switch between these and many other identities, and readers must use the context to determine which identity is being highlighted at a particular time. Biber et al. (1999: 320) suggest, with reference to scholarly writing, that this switching is random: “in some cases, academic authors seem to become confused themselves, switching indiscriminately among the different uses of we”. As the extracts above indicate, the academic authors in SCEGAD do not switch indiscriminately, but present alternate group alliances according to their communicative strategy at a specific point in the text, be this authority-enhancement, solidarity with the reader, or the exact specification of a ‘we’ group referent. All three cultural groups of academic authors in the corpus demonstrate complex but deliberate, motivated language use informed by a strategic view of social interaction. Far from being confused and aimless, as Biber et al. suggest, the SCEGAD authors support Mühlhäusler and Harré’s (1990: 60) assessment of pronoun users as “social beings engaged in purposeful social behaviour in social space”. Scholarly authors are capable of summoning different alliances with remarkable alacrity and dexterity. The rapid alternation of pronoun referents was also noted in a study of meeting talk, where Fasulo and Zucchermaglio (2002: 1125)
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.30 (1612-1695)
Tamsin Sanderson
observed “high rates of pronoun switch within the same turn and even change of the sense conveyed by the pronoun across different turn units”, with ‘we’ being the most variable personal pronoun of all. Similarly, in the SCEGAD texts, the authors sometimes switch allegiances and identities even within a single sentence: (36) Folklorists may claim faithfulness to the complete story, but texts like Charles Briggs and Julian Vigil’s The Lost Gold Mine of Juan Mondragon (1990) show us what a rigorously verbatim transcript looks like. (FLUS)
In the extract above, ‘us’ is contrasted with the academic community of folklorists, to which the author certainly belongs. The ‘us’ may refer to the author and readers, or even to the wider community of academics. The same type of flexibility is present in the German pronominal system, and the German scholars in SCEGAD exploit this characteristic in the same way, and with the same routine dexterity, as the English-speaking academics: (37) Die Beobachtung, daß wir zwischen verschiedenen Formen wählen können, finden wir einerseits in der Umgangssprache, ... [We find the observation, that we can choose between different forms, on the one hand in colloquial speech...] (LGDE)
Here, the first ‘wir’ is inclusive, meaning ‘we (German speaking) people’, while the second ‘wir’ refers to the academic community, which analyses the linguistic behaviour of the former group. At times, the possible referents of ‘we’ are even threefold: in the following extract from a multiple-authored text, the ‘we’ could be either the two authors, or the authors and their reader(s), or also the academic community of historians, to which the authors also belong (or any combination of these three groups): (38) The point of a spiritual journal was to record the feelings: the penitence for sins about which we can only speculate... (HIUS)
Alongside indicating multiple allegiances, the SCEGAD authors also exploit the pronominal flexibility of English and German to reduce the personal content of an utterance. This is achieved by employing a first person plural for a first person singular referent, though a third person singular form is also used occasionally. Some typical examples from the corpus follow: (39) In this study we seek to provide an appropriate syntactic analysis for residual OV order in non-literary LME, on the basis of these findings. We also relate OV remnants to another syntactic construction [...]. (LGUK) (40) An dieser Stelle ist zu betonen, daß wir natürlich nicht das Denken des Menschen Friedrich von Staufen erfassen können und wollen. [At this point, it must be emphasised that we of course neither can nor wish to grasp the thinking of the man Friedrich von Staufen.] (HIDE)
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.31 (1695-1748)
Interaction, identity and culture in academic writing
There are numerous possible motivations for stylistic choices such as these. On the one hand, choosing not to foreground oneself may be construed as modesty, a desire to shrink into the background, emphasising the content of the article rather than its originator. On the other hand, continually couching personal opinions and actions in plural personal pronouns could also be a way of increasing the gravity of these opinions, and underlining one’s own importance: it is not for nothing that the second person singular polite forms of address in many languages are grammatically plural forms. Thus, while some researchers regard this particular type of mismatch between pronoun form and referent simply as hedging, the avoidance of the subjectivity signalled by first person singular forms in fact serves far more diverse and more complex functions. The German authors in particular employ first person plural pronouns to express even absolutely personal speech acts such as recognising or speaking: (41) Hier erkennen wir auch die Nähe zur konditionalen Verwendung von ‘wo’. [Here we recognise the proximity to the conditional use of ‘where’.] (LGDE) (42) In diesem Sinne, der diese Gruppen ohne Unterscheidung umfaßt, sprechen wir im folgenden von Friedrich Barbarossa. [It is in this sense, which encompasses these groups without differentiation, that we speak of Friedrich Barbarossa in the following.] (HIDE)
In the view of the Longman Grammar of Spoken and Written English, “by choosing the plural pronoun ‘we’ rather than ‘I’, a single author avoids drawing attention to himself/herself, and the writing becomes somewhat more impersonal” (Biber et al. 1999: 330). However, it is not at all clear that scholarly texts really become less personal when written in this style. In real terms, the writing is exactly as personal as it is when the authors refer to themselves as ‘we’ as when they do so with ‘I’: the opinions, observations and perceptions are still those of the same person, the same person has still decided how to organise the text. In terms of reader perception or general tone, it is highly unlikely that readers fail to recognise the individual author superficially camouflaged by an inclusive ‘we’ in the extracts above. In fact, it is possible that the strategy could backfire: single authors who consistently use plural forms actually risk drawing increased attention to themselves, because this style may be perceived by the reader as unnatural or affected. In an alternative interpretation, Mühlhäusler and Harré (1990: 129) suggest that the “narrative convention [of using ‘we’ for ‘I’] has the effect of a rhetorical distancing of the speaker from an overt self-reference to make the personal source of advice or knowledge or whatever it may be more palatable”. The distancing effect is certainly present in the corpus texts, but it depends on reader expectations whether the pronoun swap is perceived as more or less “palatable” than explicit self-reference, assuming of course that ‘palatability’ is actually what the writer intended to achieve.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.32 (1748-1811)
Tamsin Sanderson
Ultimately, the use of first person plural forms for authorial self-reference is a question of personal style, and style, of course, is a matter of opinion: A single author should also use ‘we’ in the common construction that politely includes the reader: ‘We have already seen ...’ But never use ‘we’ as a mere substitute for ‘I,’ as in, for example, ‘In our opinion ... ,’ which attempts modesty and achieves the reverse; either write ‘my’ or resort to a genuinely impersonal construction. (American Institute of Physics Style Manual 1997: 14)
. Conclusion Contrary to traditional conceptions, academic discourse is neither purely objective nor impersonal. As researchers such as Scollon (1994), Drescher (2004) and Hyland (2005b) have argued, academic writing is in fact intrinsically personal, in that its origin and functions – persuasion, evaluation, argumentation – involve human thought and observation. While some academics, in particular the German academics in SCEGAD, seek to camouflage their agency by avoiding the explicit surface-level subjectivity indicated by person reference, no measure of avoidance can ultimately disguise the author’s presence in the text. Personal pronouns, in particular first person singular and plural pronouns, and second person singular pronouns, are central to the construction of identity and the management of social interaction in academic texts. Academic authors, as expert writers, employ person reference for a wide variety of interpersonal functions, such as establishing rapport with the imagined reader, enhancing their authority, and both minimising and strengthening their responsibility for the truth value of an utterance. The corpus texts show how academic authors purposefully and creatively exploit the inherent flexibility of the pronominal system in English and German in order to achieve specific discourse goals. The way in which scholarly authors make use of the linguistic resources of their native language in research articles is influenced by cultural background, discipline, gender and professional status. The English-speaking academics represented in the corpus are significantly more likely to refer to themselves and to address the imagined reader in their texts. They are much more ‘visible’ as authors, and often construct their texts as an explicit dialogue with the reader. The German-speaking authors in the corpus, on the other hand, refrain from person reference in general, and first person singular pronouns in particular. They are more likely to couch their personal opinions in first person plural pronoun forms, and prefer not to address the reader directly. While the English-speaking scholars in SCEGAD are more likely to refer to themselves as individuals than as members of a group, the German authors in the corpus are twice as likely to claim group membership as they are to explicitly refer to themselves in the text. The English-
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.33 (1811-1868)
Interaction, identity and culture in academic writing
speaking authors therefore use person reference to construct their texts as at once more individual and more personal than those of their German colleagues. This does not mean that the German scholars’ texts are less personal or less dialogic than are those of the English-speaking scholars. The interpersonal content of the German scholars’ writing, and their interaction with the reader, is however not usually expressed through person reference.13 In the context of the traditional view of academic writing as impersonal, and in light of the posited ‘I-taboo’ in German academic writing, a cross-cultural study of how academic authors actually behave in their published work is interesting. It emerges that while the ‘I-taboo’ in the German texts is not absolute, the German academics represented in SCEGAD do explicitly refer to themselves significantly less often than their English-speaking colleagues. In this sense, the ‘I-taboo’ appears to hold for German academic writing in the humanities as it is represented in this corpus. If we accept the general philosophical truth that academic authors, as sentient and communicative human beings, are no more objective than any other type of human author, what is unique to academic authors is the attempt to camouflage the personal origins of their perception by avoiding person reference. The analysis in this article has shown that some groups of scholars try to hide themselves more than others do.
References Aikenhead, G. 1996. Science education: Border crossing into the subculture of science. Studies in Science Education 27: 1–52. Argamon, S., Koppel, M., Fine, J. & Shimoni, A. R. 2003. Gender, genre, and writing style in formal written texts. Text 23(3): 321–346. Aston, G. & Burnard, L. 1998. The BNC Handbook: Exploring the British National Corpus with SARA. Edinburgh: EUP. Bacon, F. 1863. Novum Organum Scientiarum. The works of Francis Bacon. Vol. 8. Trans. J. Spedding, R. Ellis & D. Heath. Boston MA: Taggard and Thompson. Translation of Novum Organum Scientiarum. Londinii: J. Billium, 1620. 25 November 2005 Bailey, R. W. 1992. Images of English: A cultural history of the language. Cambridge: CUP. Baumgarten, N., House, J. & Probst, J. 2001. Untersuchungen zum Englischen als lingua franca in verdeckter Übersetzung: Theoretischer Hintergrund, Weiterentwicklung des Analyseverfahrens und erste Ergebnisse. Arbeiten zur Mehrsprachigkeit 20. Hamburg: Universität Hamburg. Berger, P. & Luckmann, T. 1966. The Social Construction of Reality. New York NY: Doubleday.
. Kretzenbacher emphasises the importance in German academic writing of implicitly dialogic devices, such as the use of citations, and the dialogue with other researchers (1998: 137). I would add the use of indefinite pronouns, directives and rhetorical questions to this list.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.34 (1868-1973)
Tamsin Sanderson
Biber, D., Johansson, S., Leech, G., Conrad, S. & Finegan, E. 1999. Longman grammar of spoken and written English. Harlow: Longman. Bramley, N. 2001. Pronouns of Politics: The Use of Pronouns in the Construction of ‘Self ’ and ‘Other’ in Political Interviews. PhD dissertation. Australian National University, 2001. 20 December 2004. Breivega, K. R., Dahl, T. & Fløttum, K. 2002. Traces of self and others in research articles: A comparative pilot study of English, French and Norwegian research articles in medicine, economics and linguistics. International Journal of Applied Linguistics 12(2): 218–239. Busch-Lauer, I.-A. 2001. Kulturspezifik in englischen und deutschen Originalarbeiten – Medizin und Linguistik im Vergleich. In Zur Kulturspezifik von Textsorten, U. Fix, S. Habscheid & J. Klein (eds), 51–67. Tübingen: Stauffenburg. Comrie, B. 1981. Language Universals and Linguistic Typology: Syntax and morphology. Chicago IL: University of Chicago Press. Crawford, M. 1995. Talking Difference: On gender and language. London: Sage. Drescher, M. 2004. Sprache der Wissenschaft, Sprache der Vernunft? Zum affektleeren Stil in der Wissenschaft. Paper. Colloquium Wissenschaftsstile im Kontrast, Albert-LudwigsUniversität, Freiburg, 29 April, 2004. Duszak, A. 1997. Culture and Styles of Academic Discourse. Berlin: Mouton de Gruyter. Fasulo, A. & Zucchermaglio, C. 2002. My selves and I: Identity markers in work meeting talk. Journal of Pragmatics 34(9): 1119–1144. Fløttum, K., Dahl, T. & Kinn, T. 2006. Academic Voices: Across languages and disciplines. Amsterdam: John Benjamins. Gläser, R. 1998. 47. Fachtextsorten der Wissenschaftssprachen I: Der wissenschaftliche Zeitschriftenaufsatz. In Fachsprachen. Ein internationales Handbuch zur Fachsprachenforschung und Terminologiewissenschaft, L. Hoffmann, H. Kalverkämper, H. E. Wiegand, C. Galinski & W. Hüllen (eds), 482–488. Berlin: Walter de Gruyter. Graddol, D. & Swan, J. 1989. Gender Voices. Cambridge: Blackwell. Gumperz, J. (ed.). 1982. Language and Social Identity. Cambridge: CUP. Halliday, M. & Hasan, R. 1976. Cohesion in English. London: Longman. Harding, S. 1991. Whose Science? Whose Knowledge? Thinking from Women’s Lives. Ithaca NY: Cornell University Press. Hardt-Mautner, G. 1995. Only connect: Critical discourse analysis and corpus linguistics. UCREL Technical Papers 6. 16 January 2006 Harwood, N. 2005b. ‘Nowhere has anyone attempted. . .In this article I aim to do just that’. A corpus-based study of self-promotional I and WE in academic writing across four disciplines. Journal of Pragmatics 37(8): 1207–1231. Harwood, N. 2005a. ‘We do not seem to have a theory...The theory I present here attempts to fill this gap’: Inclusive and exclusive pronouns academic writing’. Applied Linguistics 26(3): 343–375. Harwood, N. & Hadley, G. 2004. Demystifying institutional practices: Critical pragmatism and the teaching of academic writing. English for Specific Purposes 23(4): 355–377. Hiatt, M. 1977. The Way Women Write. New York NY: Teachers’ College Press. Holmes, J. 1993. Women’s talk: The question of sociolinguistic universals. Australian Journal of Communication 20(3): 125–149.
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.35 (1973-2076)
Interaction, identity and culture in academic writing
House, J. 2003. English as lingua franca and its influence on discourse norms in other languages. In Translation Today, G. Anderman & M. Rogers (eds), 168–180. Clevedon: Multilingual Matters. Hunston, S. 2004. Counting the uncountable: Problems of identifying evaluation in a text and in a corpus. In Corpora and Discourse, A. Partington, J. Morley & L. Haarman (eds), 157–188. Frankfurt: Peter Lang. Hutz, M. 2001. ‘Insgesamt muss ich leider zu einem ungünstigen Urteil kommen’ – Zur Kulturspezifik wissenschaftlicher Rezensionen im Deutschen und Englischen. In Zur Kulturspezifik von Textsorten, U. Fix, S. Habscheid & J. Klein (eds), 109–130. Tübingen: Stauffenburg. Hutz, M. 1997. Kontrastive Fachtextlinguistik für den fachbezogenen Fremdsprachenunterricht. Fachzeitschriften der Psychologie im interlingualen Vergleich. Trier: Wissenschaftlicher Verlag. Hyland, K. 2005b. Metadiscourse: Exploring interaction in writing. London: Continuum. Hyland, K. 2005a. Stance and engagement: A model of interaction in academic discourse. Discourse Studies 7(2): 173–192. Hyland, K. 2000. Disciplinary Discourses: Social interactions in academic writing. Harlow: Longman. Hyland, K. 1999. Persuasion in academic articles. Perspectives 11: 73–103. 19 December 2005 Hyland, K. 1998. Hedging in Scientific Research Articles. Amsterdam: John Benjamins. Hyland, K. & Tse, P. 2004. Metadiscourse in academic writing: A reappraisal. Applied Linguistics 25(2): 156–177. Ivaniˇc, R. 1998. Writing and Identity: The discoursal construction of identity in academic writing. Amsterdam: John Benjamins. Jespersen, O. 1924. The Philosophy of Grammar. London: Allen & Unwin. Kretzenbacher, H. L. 1991. Syntax des wissenschaftlichen Fachtextes. Fachsprache 13: 118–138. Lakoff, R. 1975. Language and Woman’s Place. New York NY: Harper and Row. Le Page, R. & Tabouret-Keller, A. 1985. Acts of Identity. Creole-based approaches to language and ethnicity. Cambridge: CUP. Lyons, J. 1977. Semantics. Cambridge: CUP. Malone, M. J. 1997. Worlds of Talk: The presentation of self in everyday conversation. Cambridge: Polity Press. Maltz, D. N. & Borker, R. A. 1982. A cultural approach to male – female miscommunication. In Language and Social Identity, J. Gumperz (ed.), 196–206. Cambridge: CUP. Mauranen, A. 2004. They’re a little bit different. . . Observations on hedges in academic talk. In Discourse Patterns in Spoken and Written Corpora, K. Aijmer & A.-B. Stenström (eds), 173–197. Amsterdam: John Benjamins. Morley, L. 1995. An agenda for gender: Women in the university. European Journal of Women’s Studies 2: 271–275. Mühlhäusler, P. & Harré, R. 1990. Pronouns and People: The linguistic construction of social and personal identity. Oxford: Basil Blackwell. Mullany, L. 2002. ‘I don’t think you want me to get a word in edgeways do you John?’: Reassessing (im)politeness, language and gender in political broadcast interviews. Working Papers on the Web 3 (Linguistic Politeness and Context), 7 December, 2005
JB[v.20020404] Prn:11/04/2008; 10:48
F: SCL3104.tex / p.36 (2076-2145)
Tamsin Sanderson
Myers, G. 2001. ‘In my opinion’: The place of personal views in undergraduate essays. In Academic Writing in Context, M. Hewings (ed.), 63–78. Birmingham: University of Birmingham Press. O’Barr, W. & Atkins, B. 1980. Women’s language or ‘powerless language’?. In Women and Language in Literature and Society, S. McConnell-Ginet, R. A. Borker & N. Furman (eds), 93–110. New York NY: Praeger. Popper, K. R. 1979. Objective Knowledge: An evolutionary approach. Oxford: OUP. Sacks, H. 1992. Lectures on Conversation. Vols 1 & 2. Oxford: Basil Blackwell. Sanderson, T. Forthcoming. Corpus, Culture, Discourse: Social interaction in academic writing. Tübingen: Gunter Narr. Schegloff, E. 1996. Some practices for referring to persons in talk-in-interaction: A partial sketch of a systematics. In Studies in Anaphora B. A. Fox (ed.), 437–485. Amsterdam: John Benjamins. Scollon, R. 1994. As a matter of fact: The changing ideology of authorship and responsibility in discourse”. World Englishes 13: 33–46. Simpson, R. 2004. Stylistic features of spoken academic discourse: The role of formulaic expressions. In Discourse in the Professions: Perspectives from corpus linguistics, U. Connor & T. A. Upton (eds), 37–64. Amsterdam: John Benjamins. Swales, J. 2004. Evaluation in academic speech: First forays. In Academic Discourse: New insights into evaluation, G. Camiciotti & E. Tognini-Bonelli (eds), 31–53. Frankfurt: Peter Lang. Thompson, G. 2001. Interaction in academic writing: Learning to argue with the reader. Applied Linguistics 22(1): 58–78. Trumpp, E. 1998. Fachtextsorten kontrastiv: Englisch-Deutsch-Französisch. Tübingen: Gunter Narr. Van Riemsdijk, H. & Williams, E. 1986. An Introduction to the Theory of Grammar. Cambridge MA: The MIT Press. Wales, K. 1996. Personal Pronouns in Present-day English. Cambridge: CUP. Weinreich, H. 1989. Formen der Wissenschaftssprache. Jahrbuch 1988 der Akademie der Wissenschaften zu Berlin, 119–158. Zhabotynska, S. 2001. Author-profile in scholarly papers: Anglo-American vs. Ukranian/ Russian. In Approaches to the Pragmatics of Scientific Discourse, A. Kertesz (ed.), 73–89. Frankfurt: Peter Lang.
JB[v.20020404] Prn:8/02/2008; 12:24
F: SCL31P2.tex / p.1 (61-88)
Exploring discourse in workplace settings
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.1 (47-125)
“Got a date or something?” An analysis of the role of humour and laughter in the workplace meetings of English language teachers Elaine Vaughan Mary Immaculate College, Ireland
This chapter brings instances of humour and laughter into relief using a corpus of authentic institutional interaction of English language teachers in school staff meetings. Humour is used within the meetings as a means of showing mutual support and creating solidarity. The corpus also contains a large proportion of subversive humour, or humour which is directed against the institution, individuals in the group, the group itself and the students. Identifying humour in the data is not a simple case of finding instances of laughter or assuming that it signifies either the intention of the speaker to elicit laughter, or to be humorous. However, wherever humour is manifested, laughter frequently occurs. The methodological issue of identifying and transcribing humour is discussed.
.
Introduction
This chapter is concerned with the institutional interaction of English language teachers, and therefore is limited to a specific feature of teacher language. While the dominant trend in the research into this locale of interaction is to collect and analyse discourse that occurs in the language classroom itself (teacher-student interaction, for example), this study differs in that it reaches beyond the classroom, with the aim of exploring other types of discourse situations that language teaching professionals encounter in the workplace. Obviously, the scope of workplace discourse within an organisation is extremely broad, and consequently this study focuses on a corpus of meetings in particular, as meetings are a genre readily identifiable with the workplace (for a more detailed account of meetings as genre, see Bargiela-Chiappini & Harris 1995, 1997). In addition, due to the fact that recording took place at two different locations during the preliminary stages of the study, it was necessary to adhere to a type of interaction that, from a generic point of view at least, was comparable for both locations. The first subcorpus of meetings was
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.2 (125-188)
Elaine Vaughan
recorded in the English department of a public university in México. The university is, in the main, a technical university, with the vast majority of undergraduate students studying for science, engineering and computer engineering degrees. English language classes are compulsory for all students, and successful completion of the English course is a prerequisite for graduation. The second subcorpus was recorded in a private language school in the west of Ireland. The school offers a wide variety of courses; as well general level classes, there are business English, English for Academic purposes (EAP) and a variety of examination classes (for more detailed information on the corpus, see Data and Methodology below). Meetings were a common, if at times unpopular, feature of both workplaces. They were especially unpopular in the English department in México mainly because they were not perceived to serve any particular function, and were seen as a ‘waste of time’. The meetings in the private language school in Ireland were seen more pragmatically as a forum for the transaction of some of the more esoteric business of language teaching, such as student placement. Indeed, it is the fact that the professionals at these meetings are discussing, negotiating and managing this ‘esoteric’ business of language teaching that is the defining feature of the study. The preliminary quantitative analysis of this corpus using the wordlist function of WordSmith Tools (Scott 1999) highlighted an unexpected frequency in the data. The extralinguistic feature, laughter, that had been tagged variously as laughs, laughing and laughter (see Transcription Issues for details on the coding of these features) had a significantly high occurrence in the wordlist, and thus warranted closer attention from a qualitative point of view. Humour revealed itself as a significant feature of this study, and the way in which it builds, maintains, and more specifically, illuminates, the workplace relationships of the teachers in question will be illustrated. This unexpected frequency led to the analysis presented here, and highlights, albeit in microcosm, the compatibility of corpus-based/driven methodology and discourse analysis. Corpus methods can alert us to patterns, not just lexical or grammatical, but also, as in this case, interactional.
. Humour and the workplace setting Analysis of the phenomenon of humour has been undertaken across a wide variety of disciplines, such as philosophy, psychology, anthropology, sociology, and organisational studies. (For an extensive overview of linguistic theories of humour and laughter, see Attardo 1994.) Perhaps inevitably, defining humour itself is problematic; there is some overlap in terms (wit, joking) and many studies focus on laughter specifically. Laughter has been proposed as the language of humour (Zijderveld 1983), though the equating of laughter with humour has also been criticised (Adelswärd & Öberg 1998; Osvaldsson 2004). What emerges from much of
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.3 (188-228)
The role of humour and laughter in the workplace
the earlier work is the conception of humour as a “social lubricant” which “makes the routine flow of life possible” (Martineau 1972: 103) and this is relatively uncontested throughout a literature that has tended to focus on the functions of humour in interaction. Graham et al. (1992) synthesize much of this research, in spite of its diversity of provenance, and highlight three broad theoretical perspectives which have informed the analysis of humour: superiority theories, incongruity theories, and relief or arousal theories. Superiority theories suggest that laughter is always directed at someone and therefore the person that laughs is expressing their superiority; incongruity theories focus on cognitive processes involved in humour; and relief or arousal theories focus on the premise that laughter, as a physical event, is the release of nervous energy. They present an extremely broad-ranging overview of humour research in a variety of disciplines, and detail twenty-four functions of humour in interaction based on this research. Particularly salient for this study is the research cited by Graham et al. examining the relationship between humour and group cohesiveness. Pogrebin and Poole (1988; cited in Graham et al. 1992: 166) present the following three principal functions of group humour: (1) it allows members of the group to share common experiences and define their working ideology, (2) it promotes social solidarity and (3) it helps the group to cope with a variety of forces outside of their control. Humour in groups is summarised as serving the social functions of “defining and re-defining a group, clarifying status relationships, and easing the tension of new or novel stimuli” (ibid: 167). Research within the discipline of organisational studies has emphasised this potential of humour to provide interesting insights into not only group cohesion, but also the cultural values of the workplace, and even how social status within a workplace is negotiated (Duncan 1982). Vinton (1989) has also described how by assimilating and mirroring the type of humour used in the workplace, employees integrate into new work situations more successfully. Language-based research into the role of humour and laughter in the institutional context has also underlined its largely solidarity-based, collaborative nature. Adelswärd (1989) examines the social significance of laughter in a variety of institutional contexts and finds, among other things, that “mutual laughter is a sign of rapport and consensus” (p. 107), a finding that is also reflected to a certain degree in this study (see Discussion and Results). However, this is not to say that the use of humour is always by its nature benign. As Hay (2000: 716) puts it, every attempt at humour expresses solidarity but also involves constructing a position of respect and status within a group. Although Hay’s humour research focuses on casual conversation, and highlights its solidarity-based, power-based and psychological functions, this view of the functions of humour can be extended to the workplace. Rogerson-Revell highlights how in current linguistic research into business communication there has been a shift in emphasis from the structural organisation of professional talk “towards a more pragmatic and functional analysis, focusing on
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.4 (228-298)
Elaine Vaughan
the strategic use of linguistic resources to achieve certain outcomes” (2006: 7). An example of how humour can be used strategically to achieve a particular purpose can be seen in Collinson’s (1988) study of humour as a conduit to a masculine sense of identity in shop-floor relations. He suggests that the men were required to “give and take a joke, to swear and to retain their domestic authority” (p. 197). He also highlights the use of humour as a means to “control those perceived to be not pulling their weight”’ (ibid.). In a similar vein, Bonaiuto et al. (2003) investigate the organisation of humorous sequences in group negotiations and conclude that humour “enables people to cautiously avoid the use of obvious criticism” (p. 214), while at the same time providing a frame to undermine the proposals of others. Zajdman (1995) similarly considers humour as a “strategy” which can mitigate a face-threatening act (FTA), and illustrates the advantages that can accrue to the speaker in using humour strategically. Even where self-deprecating, or “selfdirected” humour is used, it contains the circular message “I am weak. I admit it. To admit means to be strong. Therefore I am strong” (p. 338). The unifying theme in these studies is the idea that through this ostensible perception of humour and group laughter as solidarity based as well as solidarity building, a subtle means of expressing and wielding power in interaction is provided. Holmes (2000: 161), reporting on the Language in the Workplace Project in New Zealand, highlights the lacuna which exists in humour research of taperecorded material from authentic workplace interaction. She also provides a definition of humour that has informed subsequent studies (see Rogerson-Revell 2006; Mullany 2004), and is adopted here: Utterances are defined as humorous by the analyst, “on the basis of paralinguistic, prosodic and discoursal clues, as intended by the speaker(s) to be amusing and perceived to be amusing by at least some participants” (2000: 163). Her theoretical framework combines insights from politeness theory (Brown & Levinson 1987) and Critical Discourse Analysis (CDA) (Fairclough 1995). The concept of ‘face’ is crucial in Brown and Levinson’s work. They define face as “the public self-image that every member [of society] wants to claim for himself ”. It consists of two related aspects (1) negative face – the need for freedom from imposition and (2) positive face – the need for enhancement of a positive self-image (Brown & Levinson 1987: 61). These two basic face needs are satisfied by two different styles of politeness. Negative politeness is action aimed at non-interference and non-imposition. Positive politeness is action aimed at building on indexes of solidarity such as in-group membership (Blum-Kulka 1997: 143). Brown and Levinson’s contention that humour (joking) addresses positive face needs and thus engenders solidarity is obviously most influential (though this has been contested elsewhere, see Austin 1990; Eelen 2001; Mullany 2004), as well as CDA’s concept of “repressive discourse” which “tends to distract attention away from issues of power” (Holmes 2000: 165). According to Habermas (1984), institutional discourse is by its nature strategic, imbued with, and distinguished by
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.5 (298-315)
The role of humour and laughter in the workplace
asymmetry – it is “power laden” (Thornborrow 2002: 2). However, in today’s institutional environment with its veiled hierarchies and with what Fairclough (1992) has called the ‘conversationalisation’ of institutional discourse, humour provides a way of ‘doing power’ less overtly; in other words, it “can be used to achieve the speaker’s instrumental goal while apparently de-emphasising the power differential” (Holmes 2000: 165). While humour can be used by the institutionally powerful speakers who operate in an environment where “explicit orders” are no longer acceptable (ibid: 175), it can also be used by the subordinates in an organisation to challenge power structures. Thus, the potential of humour to fulfil both positive and negative politeness functions is underlined. To summarise, we can say that (1) as a positive politeness strategy it expresses solidarity or collegiality, as well as self-deprecation (by protecting the speaker’s positive face needs) and (2) as a negative politeness strategy it downtones or hedges a face-threatening act (FTA) or face attack act (Austin 1990) such as a criticism or insult. Holmes and Marra (2002a) explore how humour contributes to workplace culture by helping to create a distinctive identity for the group. Further research by Holmes and Marra (2002b) distinguishes between reinforcing and subversive humour in the workplace. It pivots on the use of subversive humour in the workplace, but also points out that the use of humour can not only reinforce existing solidarity relationships, but also existing power relationships (p. 70). (Reinforcing and subversive humour are discussed in greater detail in Discussion and Results below.) In fact, Holmes and Stubbe (2003: 109–110) consider humour indexical of the power relationships in the workplace, where “humour typically constructs participants as equals”. This implication of humour in issues of power in the workplace is enlightening, and gives the tendency in earlier literature to equate humour and solidarity an added resonance. However, power and solidarity are not necessarily opposed, but mutually entailed (Tannen 2006), and humour seems to be the most regularly used strategy for hierarchically powerful participants in the interaction to downtone this power differential, as well as being a socially acceptable way for the less powerful to contest workplace hierarchies. Mullany (2004) has highlighted the tactical use of humour within the institutional context in relation to power using data from business meetings. Her research includes the additional variable of gender, in an investigation of how meeting Chairs use humour as a device to pursue compliance from the subordinates in the company. To summarise, the idea of the strategic use of humour underlines Holmes and Stubbe’s contention that “humour is a valuable multifunctional resource in workplace interaction [...;] many meetings are punctuated by bursts of humour, which tend to occur at strategic points” (2003: 109, emphasis added). This begs the question: at what point or points do these distractions occur? This issue is explored in the analysis that follows. Previous research contributes substantial evidence of the “fruitful line of investigation” (Mullany 2004: 13) provided by humour. While its
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.6 (315-437)
Elaine Vaughan
solidarity function is uncontested, it is complemented by a potential to be used repressively by those in power, and as a non-threatening means of subversion giving members of an institution an acceptable way of pushing the boundaries of workplace hierarchies. In the case of the teachers in this study, it supplies a safe way of criticising their colleagues, their students and the way the organisation works.
. Data and methodology The participants in this study are all qualified teachers, with a minimum of three years’ experience. From the English language department in México, the first part of the corpus comprises three meetings with an average duration of thirty-five minutes. The teachers are all native speakers of English, although there are a variety of nationalities – American, English, Scottish, Irish, Ugandan, Jamaican and Canadian. There are also three meetings in the corpus from the private language school in Ireland, however the length of the meetings varies fairly dramatically as is reflected in the number of words in each meeting (see Table 1 below). The majority of the staff at this location are speakers of Southern Irish English. As the data was collected with the co-operation of teachers from meetings which took place in very different countries, under very different working conditions, there is little divergence in the type of topics the teachers discuss. This is perhaps unremarkable, but the first reading of the transcripts was completed with the aim of ensuring that there was a sufficient similarity in the concerns of the meetings to justify a professional comparison between the two sites. The general topics that are discussed in these meetings are presented in Table 2 below. Obviously, this context of language use is extremely sensitive, and consent was obtained in the first instance from the Director of Studies (DoS) to approach the Table 1. Description of the Corpus of the Meetings of English Language Teachers (C-MELT)
Subcorpus 1 Public university, México Subcorpus 2 Private language school, Ireland
Meeting
No. of participants
No. of tokens
C-MELT 1 C-MELT 2
10 11
12,784 7,163
C-MELT 3
10
9,876
C-MELT 4 C-MELT 5
9 11
3,707 3,604
C-MELT 6
14
2,841 C-MELT Total
Total
29, 823
10,152 39,975
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.7 (437-472)
The role of humour and laughter in the workplace
Table 2. Overview of topics discussed in the meetings Topics Administration Assessment/examination Classroom issues Extra-curricular activities Future meetings Pilot courses Placing students Professional development Student attendance Student motivation
teachers for recordings, and then from the teachers themselves. The participants in this study have complete access to the transcripts, and in the latter stages of the study have had the opportunity to expunge any utterances or sequences they do not want included. However, none of the participants wished to do so, and so the transcripts provide a complete audio representation of each of the meetings. All names and any references which could identify the exact locations of the data have naturally been removed. In all, approximately 3.5 hours of data, or just under 40,000 words, was transcribed and analysed, and stored together as the Corpus of Meetings of English Language Teachers (C-MELT). When comparison between the two locations in the corpus was necessary, C-MELT was broken down into its two constituent sub-corpora. . Transcription issues The definitional dichotomy which exists in the literature on humour and laughter, and the disciplinary diversity of the literature on humour itself, can pose an initial problem. Does laughter in the corpus indicate humour? Or indeed, does humour always provoke laughter? As previously mentioned, some progress can be attempted in this vexed issue by consideration of the context in which the humour occurs. Adelswärd and Öberg (1998: 412) point out that identifying laughter and humour can have its drawbacks, as although laughter can be associated with mirth, it can “just as well accompany feelings of embarrassment or expressions of maliciousness and spite”; in fact, laughter can also accomplish “complex interactional goals” (see Poyatos 1993 for a detailed treatment of these goals). Therefore, identifying humour in the data is not a simple case of finding instances of laughter, and of assuming that this not unambiguous indication of amusement signifies (a) the intention of the speaker to elicit laughter or (b) the interpretation of an utterance by the listeners as intended to provoke laughter. In Jefferson’s (1985) treatment of
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.8 (472-517)
Elaine Vaughan
the transcription and analysis of the phenomenon of laughter itself, it is pointed out that, in creating a transcript, how, or whether, we “get things right” (an obscure concept in itself) is heavily dependent on what it is we wish to attend to (p. 25). Jefferson’s focus is the pervasive nature of laughter in talk, or the phenomenon of what Goffman (1961) calls “flooding out”: the way in which the word or utterance is “invaded” by laughter. As previously mentioned, the corpus-driven method by which the insight that laughter and, as a corollary, humour, should become a focus of study raised the issue of ‘retrospective’ coding for the transcripts. Although laughter had been selected as an extralinguistic feature to be tagged, there had been no specific attention given to how variations (type of laughter, whether it was part of an individual speaker’s utterance or whether one or more participants responded to an utterance with laughter, etc.) would be marked. The methodological issue of identifying humour in the corpus was addressed as follows. The meetings of which the corpus is comprised were analysed for the phenomenon of laughter; where it was identified, the cause for the laughter was isolated, as shown in Table 3. If only one speaker laughed, the tenor of the laughter was analysed to get at its actual meaning. Group laughter was taken to indicate that the group interpreted an utterance or sequence as humorous, and this was the most obvious starting point. Other cues which helped to identify whether or not an utterance was intended humorously included “smile voice” (Crystal 1969). Ultimately, the phenomenon was tagged in three different ways: –
– –
as part of the utterance if a speaker laughed during or at the beginning/end of an utterance (<$E> laughs <\$E>), or during an utterance (<$E> laughing <\$E>) separately, if the other participants responded with laughter to an utterance (<$E> laughter <\$E>) if the laughter was weak or prolonged (or in some other way marked), this too was noted (e.g. <$E> prolonged laughter <\$E>).
All this seems to ignore Adelswärd and Öberg’s admonition regarding the interpretation of laughter, and its inherent ambiguity. However, I would argue that the context in which the humour is essayed, the workplace, increases the likelihood that it be supported by laughter. That is to say, the preferred response to an utterance which is intended as humorous is laughter, and in the interests of solidarity, and basic politeness (rather than linguistic politeness), colleagues are likely to provide this response. Indeed, Norrick (1993) claims that joking and laughter are an adjacency pair. Hay (2001) discusses a number of ways in which humour can be supported, focussing on the way in which participants contribute more humour, play along with the gag, use echo or overlap, offer sympathy and contradict selfdeprecating humour. Some of these supporting functions, such as contributing more humour, are found in the data presented in this chapter, but the overriding
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.9 (517-617)
The role of humour and laughter in the workplace
trend is to recognise and support humour with laughter, hence its significance in frequency counts and the focus on it in this chapter.
. Discussion and results As laughter was used as an indicator of humour, this was how humorous episodes were initially identified in the transcripts. During a more detailed examination, instances of failed humour were omitted from the final count of these episodes, and in total 73 humorous episodes were isolated, as shown in Table 3. In each case, the speaker who initiates the humour was noted, however no particular pattern emerged until the status of the speaker within the organisation was given attention. Even given the fact that three of the meetings, C-MELT 2, 4 and 6 respectively, were not attended by the Director of Studies (DoS) or Head of Department (HoD), of those 73 episodes, 60 were initiated by the teachers and only 13 by the DoS/HoD. Even taking into consideration the fact that the DoS/HoD is not present at half of the meetings in the corpus, the proportion of humour initiated by the teachers is still significantly higher. It should be noted here that the counts referred to initiation of humorous episodes rather than initiation in general. From this perspective, the DoS/HoD holds more discursive power than the teachers in their ability to initiate, direct and change topics. Richards’ (2006) study of language and professional identity focuses in one chapter on varieties of humour in collaborative talk. The study is particularly interesting because it is based on the staffroom interaction of English language teachers, though, unlike the present one, it deals with a variety of different types of interaction, rather than meetings in particular. Richards points out not only where humour occurs, but also where it does not occur, that is in the discussion of arrangements, agreeing procedures and similar business (p. 103). This is also the trend in C-MELT where there is a marked absence of humour in weekly meetings (usually fairly brief, the longest is twenty minutes) that are dedicated to student placement (see C-MELT 4, 5 and 6 above), deciding which students will stay in Table 3. Initiation of humorous sequences
C-MELT 1 C-MELT 2 C-MELT 3 C-MELT 4 C-MELT 5 C-MELT 6 Total
Initiated by DoS/HoD
Initiated by teachers
Total
5 – 7 – 1 – 13
17 17 5 2 8 11 60
22 17 12 2 9 11 73
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.10 (617-713)
Elaine Vaughan
the classes they have been assigned to, or if they will be moved up or down a level. Richards highlights many functions of humour in his data, but two that stand out as highly applicable to C-MELT are the use of humour to deflect professional concerns, and the tendency to use outside agents, whether individual or institutional, as the butt of jokes. Rogerson-Revell (2006) points out the use of humour to mark a shift in style from formality to informality within a meeting. Within CMELT, humour occurs at topic transition phases, but it also appears to maintain the flow of the meetings (this is also noted by Kushner; cited in Rogerson-Revell 2006), and is used, for example, to relieve tension. Plotting the distribution of the laugh tag (and each of its lemmas) showed a marked tendency for laughter to occur during the opening and closings of the meetings, a finding also reported in Holmes (2000: 179). When humour occurred at other points, it was invariably in response to problematic issues. Although humour, as has been previously stated, is multifunctional, Holmes and Marra’s (2002a: 70) distinction between (1) reinforcing humour and (2) subversive humour is particularly useful here as conceptual prisms through which to view and interpret the data in the following analysis. As mentioned, reinforcing humour maintains or reinforces the status quo and can be divided into two sub-categories: (a) humour which reinforces existing solidarity relationships and (b) humour which reinforces existing power relationships. In general, the main focus of reinforcing humour is on solidarity and the maintenance of friendly collegial relations. Subversive humour, on the other hand, is used to challenge the existing status quo, and can be viewed as a subtle strategy available to those who are not in power. Using humour subversively does not necessarily require the speaker to mount an explicit challenge, however, so it is a relatively risk-free tactic. The teachers tend to use it, for example, as a way of criticising each other and undermining the decisions of the group. The people in the meetings with the most hierarchical power (DoS/HoD) sometimes had to enforce decisions taken at institutional or departmental level, or criticise the actions of the teachers. In order to do this, the preferred strategy is to use reinforcing humour. Table 4 below illustrates the most common functions of humour in the meetings, and the strategies used to realise them. In some cases in this data, the teachers’ use of humour can be viewed as both reinforcing and, at the same time, subversive, as can be seen in extract (1). . Subversive humour In the extract below, the senior administrator, Rachel, interrupts a meeting where student placement in being discussed. When students pay for a course in the school, they are also required to purchase the textbook used by the class; however, in some cases students are tardy about doing this. The administrator has decided that students are not allowed to attend classes until they have paid for the book.
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.11 (713-751)
The role of humour and laughter in the workplace
Table 4. Functions of humour/laughter in C-MELT
Subversive
Reinforcing (power/solidarity)
(1) Rachel:
Tom: Daniel:
Used by teachers to:
Used by DoS/HoD to:
– – – – –
resist each other’s ideas criticise/mock students criticise the decisions of the group criticise/mock institution express jocular abuse (mocking/ double entendre) – allude to money
– criticise the institution – acknowledge how unpopular and ineffective meetings are
solidarity – highlight commonality of experiences (problem students/classes etc.) – problem-solve – relieve tension
power – issue a directive – make criticism of teachers – implement decisions – move meetings forward solidarity – downtone power – create harmony in group (thus moving meeting forward)
They need to come down to buy the book. That’s how you solve the problem of making sure everybody buys it. So like it or not that’s how we’re going to deal with it. Em. Tom and John. Oooh. <$E> laughter <\$E> Scary yeah.
As Rachel’s unequivocal So like it or not that’s how we’re going to deal with it is an obvious FTA, one of the teachers responds to it by making fun of the tone in which it is delivered, and the laughter that follows is both supportive of Tom’s joking reaction to this, and also serves to break the tension that is created both by the interruption and the FTA. The exchanges that follow this extract highlight the teachers’ reaction to the decision taken at an administrative level. While the decision is unpopular, it is not explicitly resisted. The humour in this extract, and its support by the other teachers present, shows that the teachers use humour as a strategy to signal their resistance. It is also interesting to note that while Rachel attempts to make this statement and move the discourse on by addressing two of the teachers (Em. Tom and John) in order to tell them something else, she is prevented from doing so by the laughter that follows Tom’s Oooh. In one respect, this humour can be classed as subversive, in that there is an implicit resistance to the hard line taken by the administrator – however, it can also be argued that the teachers ‘close ranks’ when they support Tom’s utterance, and it reinforces a feeling of solidarity among them.
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.12 (751-793)
Elaine Vaughan
While in extract (1) the resistance is against the strictures of the institution, extract (2) illustrates subversive humour being used by a member of staff as a comment on decisions taken at a departmental level. In this meeting, the teachers are discussing student attendance, which at the university in México was often quite low. This caused countless administrative headaches as although their attendance was compulsory and, as mentioned above, a prerequisite for graduation, students frequently had classes scheduled at the same time as English or did not attend as their workload for other courses was prohibitive. At the beginning of the semester, it was decided that students would only be permitted to take the end of semester exam, which would allow progression to the next level, if they attended a certain number of classes. Those who did not would have to pay a nominal fee to take the exam. This system did not work, but as the decision was made, it would now have to be enforced. Whether or not this would encourage attendance was disputed. (2) Barry: Julia: Peter: Barry:
Kate:
What are we going to do? Threaten them? Bully them? No. Yeah it gets kind of <$E> laughing <\$E>. <$E> some laughter <\$E> Are you supposed to go down on your hands and knees? Bribe them? Give them back their two pesos? <$E> laughter <\$E> What what well we’re powerless really.
While Julia responds seriously to Barry’s joking suggestion that they threaten or bully the students, Peter’s laughing support of Barry’s utterance provokes some laughter, as does the exaggeration that Barry follows up with, warming to his theme, that they bribe them, or give them back their two pesos. This last reference to the decision that has proved so unpopular provokes more laughter than his first utterance. Barry is clearly critical of this decision to make students pay, and the real “butt” of his criticism is the department of which he is also a member. It is not unusual for subversive humour to be directed by a member of the group against the group itself. Moreover, the criticism, couched as it is humorously, is accepted implicitly in the laughter that supports it and the exchange concludes with Kate’s well we’re powerless really – an explicit acknowledgement of the futility of any such decisions. The instance of subversive humour in (2) is directed by a member of the group against the group itself, but subversive humour can also be directed at an individual, or the organisation as a whole. In extract (3), the agenda of the meeting has been established by the Chair, however, a different member of the group suggests that they proceed to the main business of the meeting.
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.13 (793-856)
The role of humour and laughter in the workplace
(3) Rose: Barry:
Shall we start? <$E> laughter <\$E> Shall we?
Here the humour takes the form of jocular abuse. Rose’s utterance provokes laughter because of its incongruous formality; Barry repeats the Shall we? in a highpitched voice and an impersonation of Rose’s accent. Within the data, there are several examples of this type of jocular abuse between Rose and Barry, which belies the very positive working relationship they had. There is yet another element to his contribution as Barry is Irish, and shall is marked in Hiberno-English as evidenced in this exchange from Clancy’s (2000) corpus of family discourse where the family are discussing an acquaintance who is ridiculed for being ‘posh’: (4) Susan: Steve: Susan:
What d’you call him that talks in ‘shall I shall we say this’. He’d be great for it. ‘Shall we go sailing or shall we do this or shall we’. What do they call him? <$E> said in a posh voice <\$E> Bernard? <$E> laughter <\$E> ‘Shall we shall we say am we’ll go sailing at around two. Twoish’.
Humour which is directed against the organisation in this data is indicative of a ‘them versus us’ dynamic and in this case constructs both the other departments in the university and the students as out-groups. What is interesting is that, on the one hand, as we can see in (1), if the administration makes a decision that could affect their students negatively, the teachers close ranks against the administration, and yet, during their meetings, the teachers also frequently direct humour against the students. In some ways, it is a safe method of framing negative evaluations, but it is also a safe way of presenting professional problems or failures. In (5), the teacher is talking about a student who has just joined the class and has clearly been placed in the wrong class. Having the student in her class that morning has been very frustrating, and she is venting some of that frustration as part of the weekly placement meeting. (5) Niamh:
He’s [he has] no comprehension. He doesn’t understand what I ask him <$E> ironically <\$E> he understands because I’m standing in front of him that I’m asking him if he understands. <$E> laughter <\$E> You know. Ciarán: Okay. Siobhán: Oh God <$E> laughs <\$E>. Michaela: He has no idea. Siobhán: Bad scene.
Ciarán as the acting DoS accepts her evaluation of the student’s comprehension but offers no other support whereas the other teachers in the exchange support not
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.14 (856-907)
Elaine Vaughan
only her humorous retelling of the situation in the class but also offer sympathy by echoing what she has said he has no idea, and by summarising the situation as a bad scene. It is in no way unusual for the students to be the butt of humour as (6) and (7) further show, and there is a sense in which these situations are accepted, through the laughter which support them, as occupational hazards that all the teachers recognise. (6) Jessica: Ciarán: Jessica: Ciarán: Jessica:
What’s her name? Fugit? Fergit yeah. Yep. She’s a bit difficult. She’s a bit cross lookin’. <$E> laughter <\$E> She just sat in the class looking <$E> makes a face <\$E>. <$E> laughter <\$E>
The way that Jessica talks about her student is completely acceptable in the meeting, and these sorts of characterisations are common in the corpus. Another example occurs when two of the teachers, Ciara and Emma, team-teach a class and are talking about two new students who have just started. One of them, Mariana, has been placed adequately, but is not a positive addition to the class: (7) Ciara: Emma:
And just Mariana is a bit moody. Yeah <$E> laughs <\$E> a bit limited as far as personality goes. <$E> laughter <\$E>
That is not to say that the teachers only report negatively on students as (8) shows: (8) Ciarán: Ciara: Emma: Ciara:
The new one? Aimée? She’s grand she’s eh very good actually. Yeah. She’s only sixteen or something she was saying but her comprehension is great. She’s just a little shy on her first day. But she’ll be fine.
The meetings provide a forum for the teachers to talk about students in a way that helps them to go about the business of running effective classes (see also Vaughan 2007), and also provides them with an opportunity to vent some of the inevitable frustrations engendered in working with people. Humour is clearly an effective means of doing this, and there is no point at which it is not supported with knowing laughter, or a sympathetic remark. The following exchange highlights the complete acceptability of making humorously negative comments, and the purpose they serve in the meetings:
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.15 (907-974)
The role of humour and laughter in the workplace
(9) Ciarán: So anyone to go down? Michaela: No but it would cheer us up a lot if you could tell us when Juan is leaving. <$E> laughter <\$E> Siobhán: That’s exactly what I wanted to know. Ciarán: <$E> laughing <\$E> I’ll check that out for ye. Siobhán: Please do <$E> laughs <\$E>. Michaela: It would make it worth the time. Siobhán: Oh he’s unbearable. He’s unbearable. Ciarán: I think actually he wants you for two hours private. I was only joking. Siobhán: <$E> laughs <\$E> I would say oh well depends how much <$E> laughs <\$E> maybe all it would take would be a bribe <$E> laughs <\$E>. <$E> laughter <\$E>
Part of the humour near the end of the exchange is based on Ciarán’s quip that not only does Siobhán have to put up with an ‘unbearable’ student in class, but may also have to teach him one-to-one two hours a week. Siobhán retorts that maybe all it would take would be a bribe to convince her to teach him. This allusion to money is a source of humour – perhaps because English language teaching is seen as more a vocational ‘labour of love’, than a career that is well paid. A similar allusion to money is also evident in the highly subversive humour derived from the following comment made by a teacher at the end of what has been a difficult meeting. As previously mentioned, in México student attendance is an intractable problem in the English department, and the teachers have been discussing at great length how they can motivate the students to attend classes, as well as the sort of excuses they will accept for absences. The meeting has reached something of an impasse in this discussion of acceptable excuses for absences: (10) Jack: Sam: Gillian:
They have to get their hair done or go to the dentist. <$E> laughter <\$E> It’s a grey area. But we get paid anyway. <$E> prolonged laughter <\$E>
The laughter that follows but we get paid anyway goes on for twenty or thirty seconds – far longer than laughter anywhere else in the corpus, presumably because of the extremely subversive nature of the comment. It is interpreted here as subversive due to the fact that salary and remuneration are not considered principal aims in English language teaching. The teachers in the study are all dedicated professionals, and the subversive laughter and humour they produce in the exchanges discussed provide a legitimate means of criticising the institution and one another,
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.16 (974-1027)
Elaine Vaughan
venting their frustrations and, consequently, serve a therapeutic function. The relief provided by this type of humour is often essential to maintaining collegiality, and so even though the comment is subversive, the laughter reinforces solidarity. In the following extract (11), there is a subversive element to the humour, but also a sense in which it functions to relieve frayed tempers at the end of a meeting that, rather than resolving an issue, created more administrative problems. (11) Olivia: Anna: Olivia: Jack:
. . .have we achieved what we came to achieve? Got a date or something? <$E> laughter <\$E> No just confused here. <$E> laughter <\$E> <$E> ironically <\$E> I thought the meeting was about Monday. <$E> Laughter <\$E>
Anna’s playful got a date or something is interpreted both as a comment on the formality of Olivia’s have we achieved what we came to achieve? and as jocular abuse (Do you need to get out of work because you’re meeting your boyfriend?). The laughter that occurs in response leads Olivia to join in and humorously explain that she is trying to summarise what the group has decided because she is just confused here. Jack’s wry I thought the meeting was about Monday confirms that she has every reason to be confused and implicitly criticises the course of the meeting. The overall purpose the laughter serves can be said to be at once subversive, in that the activities of the group are being negatively commented on, as well as reinforcing in that it re-establishes group harmony. Reinforcing humour, discussed in the next section, is similarly dualistic in character, as it may well reinforce solidarity but it can also reinforce power relationships. . Reinforcing humour In México, the HoD’s use of humour is occasionally used as a means of downtoning his power status within the group, as is evidenced in extract (12). He has been working with three other teachers on a pilot course and in this meeting they are reporting back on their progress. He has taught an ESP course and has encountered what he feels are setbacks which are familiar to all teachers. He invokes this commonality of experience during his report: (12) Peter:
And eh basically what I worked with them was the book on teaching computers to students and it’s about as dry as you could probably get. You know it’s very hard to get make oscilloscopes and analogue systems sound very very interesting. <$E> laughter <\$E>
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.17 (1027-1087)
The role of humour and laughter in the workplace
Eh related to kind of typical to eh to academic texts and the language they had problems with invariably was language from ordinary English or everyday English you know. So em I’m glad tis [it is] over. <$E> laughter <\$E> I’m sure they are too.
His humorous retelling is responded to with supportive, knowing laughter from the teachers on his staff, as he is reinforcing his identity as a teacher like the others in the group and deflecting attention from his power status within the organisation. This is also emphasised by his stating that he is sure that the students are glad the course is over too, suggesting that despite his being the ‘boss’ that he has the same classroom problems as any other teacher. Although he uses this type of humour to align himself with the teachers as a group, he also uses reinforcing humour to other ends. Reinforcing humour can also include humour which is used repressively, or to control others, therefore this aspect of it can reasonably be restricted to those who have an interest in maintaining power. This tactic is evident in (13). Peter is discussing the issue of allocating mentors to new staff, as one teacher has just joined the group and two more new teachers are expected the following week. There have been staffing shortages, so the HoD has been taking extra classes. Gaby is the new teacher in the group. (13) Peter: Gaby: Peter:
And did we get anybody for you? You were for me Peter. <$E> laughter <\$E> No I’m supposed to be getting out of work okay right.
The group reacts to Gaby’s comment with laughter as there is an element of double entendre in relation to Peter being ‘for’ Gaby. Although Peter also laughs at this comment, his next comment is quite unambiguous, despite his tone of voice being playful. This is further strengthened by the presence of the boundary markers okay right, indicating that he is exercising his right to move the interaction on and have no more discussion on whether or not he will be acting as mentor to Gaby. There is a surface sense of self-deprecation here also. Generally speaking, a boss would not be expected to represent himself as wanting to ‘get out’ of work, but earlier in this sequence Peter referred to the fact that he has been covering classes whilst awaiting the new teachers, so the implication of this ostensibly self-deprecatory comment is that Peter has more important work to do, thus distancing himself from the lower-status teachers whose work it is. This in effect reinforces the existing power relationships in the department. This is also evident in the strategies he employs to give directives, illustrated by extract (14):
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.18 (1087-1149)
Elaine Vaughan
(14) Peter:
So just a few quick little avisos as they say in Spanish and then we’ll get down to the main item on the agenda eh where are they oh God. Right eh number one eh one of the eh this is kind of a general issue by the way . . . Eh about people taking days off eh just I want you to know that personally I have absolutely no problem about it taking days off... I kind of feel myself that even though sometimes we can’t always do it that just sending them to the lab isn’t the most productive way of doing it. Now it is a good way of getting it done <$E> ironically <\$E>. <$E> laughter <\$E>
By the humorously ironic way he says Now it is a good way of getting it done (with the emphasis on is), he softens the force of the directive, and its implied criticism. The pill is also sweetened by his opening reassurance that this is kind of a general issue rather than directed against the actual teachers who were taking an inordinate amount of time off. His directive is heavily hedged (kind of a general issue, kind of feel myself ), but the laughter that he invites with the ironic tone of now it is a good way of getting it done breaks the tension, emphasises his understanding of how sometimes best practice is sacrificed to practicality, thus underlining his solidarity, as a teacher himself, with the other teachers. Although reinforcing solidarity is important to the DoS/HoD, they also use humour as a subtle and non-threatening way of highlighting their power in these meetings. Sometimes this is necessary to get meetings back on track where a distraction has occurred, as is clear from extract (15) below: (15) Rachel:
Ciarán:
Oh and Sally I have an e-mail for you in my office whenever you come up down for it. Daniel says hello. <$E> A few speakers say ‘ooh’ then laugh <\$E> Well well well Sally. <$E> laughter <\$E> So. Where were we in this exciting meeting? <$E> laughter <\$E>
Rather than explicitly saying ‘Right. Back to work’, Ciarán, the acting DoS, uses humour to gently steer the meeting. This is arguably a more successful strategy than being more forceful would have been. While some alternative interpretation could be admitted for many of these extracts, what is clear is the multifaceted ways in which humour is used in the meetings. It is highly strategic in a context that, by its nature, requires speakers to be tactical discursively. Humour and laughter can highlight existing solidarity structures in the workplaces described here, but it also underpins that solidarity, as well as allows for its creation.
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.19 (1149-1190)
The role of humour and laughter in the workplace
. Conclusion Richards (2006: 92) claims that “embedded in, arising from and flowing through many of the routines and rituals of professional life, the stream of humour is one of its most distinctive features” and this is certainly borne out in this corpus of meetings. The pervasive and multifunctional nature of the humour in the staff meetings cannot be over-emphasised. In fact, humour and laughter are multifunctional to the point that any one instance of either or both together could reasonably be interpreted in a number of different ways. The Chair, DoS or HoD can use humour, inter alia, to soften criticisms or directives, while the teachers use it to push the boundaries of institutional power, to define themselves as a group, and to identify the out-group, in this case, the students. It is used as a salve for the marked tension between professional imperatives and institutional power (or lack thereof). Tagging and transcribing extralinguistic information within a corpus can be as detailed or as minimal as the researcher chooses, depending on the purpose being served. However, the potential for a corpus to bring into relief unexpected interactional patterns cannot be underestimated. Nor indeed can the benefits of applying a number of different analytical perspectives to any data, and for each of these perspectives to contribute to a much richer understanding of the interaction itself. There is certainly an argument for providing detailed transcription, and this includes the transcription of extralinguistic information. As this chapter shows, though this extra detail may have not been included in the initial transcription, it is possible to tag significant features retrospectively. This chapter has attempted to show how typical humorous exchanges are in the meetings and to present the ways in which teachers use them as a way of defining professional boundaries and resisting institutional power. It is used to create and reinforce solidarity within the group, and is therefore a potentially significant index of group cohesiveness. When this data is presented to teachers, and English language teachers in particular, there is a high level of identification with the issues discussed. There is also, informally at least, identification with the humour that is enacted. Further studies which consider the meetings of English language teachers as professional interaction would be extremely interesting to compare with this one, most obviously from the point of view of convergence and divergence in terms of topics which are discussed and the professional concerns of the teachers. However, it would also be interesting to investigate whether humour and laughter are as prevalent in other settings as in the one reported on in this study. Given the growing canon of research that shows just how significant humour is in workplace interaction, for researchers interested in studying workplace interaction, the transcription and tagging of humour and laughter is clearly and important issue.
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.20 (1190-1298)
Elaine Vaughan
References Adelswärd, V. 1989. Laughter and dialogue: The social significance of laughter in institutional discourse. Nordic Journal of Linguistics 12: 107–136. Adelswärd, V. & Öberg, B.-M. 1998. The function of laughter and joking in negotiation activities. Humor 11(4): 411–429. Attardo, S. 1994. Linguistic Theories of Humour. Berlin: Mouton de Gruyter. Austin, P. 1990. Politeness revisited – The dark side. In New Zealand Ways of Speaking English, A. Bell & J. Holmes (eds), 277–293. Clevedon: Multilingual Matters. Bargiela-Chiappini, F. & Harris, S. 1995. Towards the generic structure of meetings in British and Italian managements. Text 15(4): 531–560. Bargiela-Chiappini, F. & Harris, S. 1997. Managing Language: The discourse of corporate meetings. Amsterdam: John Benjamins. Blum-Kulka, S. 1997. Dinner Talk: Cultural patterns of sociability and socialization in family discourse. Mahwah NJ: Lawrence Erlbaum Associates. Bonaiuto, M., Castellano, E. & Pierro, A. 2003. Arguing and laughing: The use of humour to negotiate in group discussions. Humor 16(2): 183–223. Brown, P. & Levinson, S. 1987. Politeness: Some universals in language usage. Cambridge: CUP. Clancy, B. 2000. A case study of the linguistic features of a Limerick family. MA thesis, University of Limerick. Collinson, D. 1988. Engineering humour: Masculinity, joking and conflict in shop floor relations. Organization Studies 9(2): 181–199. Crystal, D. 1969. Prosodic Systems and Intonation in English. London: CUP. Duncan, W. J. 1982. Humour in management: Prospects for administrative practice and research. Academy of Management Review 7: 136–142. Eelen, G. 2001. A Critique of Politeness Theories. Manchester: St. Jerome’s Press. Fairclough, N. (ed.). 1992. Critical Language Awareness. London: Longman. Fairclough, N. 1995. Critical Discourse Analysis: Papers in the critical study of language. London: Longman. Goffman, E. 1961. Encounters. New York NY: Bobbs Merrill. Graham, E. E., Papa, M. J. Y Brooks, G. P. 1992. Functions of humor in conversation: Conceptualisation and measurement. Western Journal of Communication 55: 161–183. Habermas, J. 1984. Theory of Communicative Action, Vol. 1 (trans. T. McCarthy). London: Heinemann. Hay, J. 2000. Functions of humour in the conversations of men and women. Journal of Pragmatics 32(6): 709–742. Hay, J. 2001. The pragmatics of humor support. Humor 14(1): 55–82. Holmes, J. 2000. Politeness, power and provocation: How humour functions in the workplace. Discourse Studies 2(2): 159–185. Holmes, J. & Marra, M. 2002a. Having a laugh at work: How humour contributes to workplace culture. Journal of Pragmatics 34(12): 1683–1710. Holmes, J. & Marra, M. 2002b. Over the edge? Subversive humour between colleagues and friends. Humor 15(1): 65–87. Holmes, J. & Stubbe, M. 2003. Power and Politeness in the Workplace: A sociolinguistic analysis of talk at work. London: Pearson.
JB[v.20020404] Prn:11/04/2008; 11:11
F: SCL3105.tex / p.21 (1298-1363)
The role of humour and laughter in the workplace
Jefferson, G. 1985. An exercise in the transcription and analysis of laughter. In Handbook of Discourse Analysis, Vol. 3: Discourse and dialogue, T. A. van Dijk (ed.), 25–34. London: Academic Press. Martineau, W. 1972. A model of the social functions of humor. In The Psychology of Humor, J. H. Goldstein & P. E. McGhee (eds), 101–125. London: Academic Press. Mullany, L. 2004. Gender, politeness and institutional power roles: Humour as a tactic to gain compliance in workplace business meetings. Multilingua 23: 13–37. Norrick, N. 1993. Conversational Joking: Humor in everyday talk. Bloomingdale IN: Indiana University Press. Osvaldsson, K. 2004. On laughter and disagreement in multiparty assessment talk. Text 24(4): 517–545. Poyatos, F. 1993. The many voices of laughter: A new audible-visual paralinguistic approach. Semiotica 93: 61–81. Richards, K. 2006. Language and Professional Identity: Aspects of collaborative interaction. New York NY: Palgrave Macmillan. Rogerson-Revell, P. 2006. Humour in business: A double-edged sword. A study of humour and style shifting in intercultural business meetings. Journal of Pragmatics 39(1): 4–28. Scott, M. 1999. Wordsmith Tools version 3. Oxford: OUP. Tannen, D. 2006. Ventriloquizing in family interaction: Rethinking dialogue, framing and power. Plenary presented at Sociolinguistics Symposium 16, 6–8th July, University of Limerick, Limerick, Ireland. Thornborrow, J. 2002. Power Talk: Language and interaction in institutional discourse. Harlow: Pearson. Vaughan, E. 2007. ‘I think we should just accept ... our horrible lowly status’: Analysing teacherteacher talk within the context of community of practice. Language Awareness 16(3): 173– 189. Vinton, K. L. 1989. Humour in the workplace: Is it more than telling jokes? Small Group Behaviour 20(2): 151–166. Zajdman, A. 1995. Humorous face-threatening acts: Humor as strategy. Journal of Pragmatics 23(3): 325–339. Zijderveld, A. 1983. The sociology of humour and laughter. Current Sociology 31(3): 1–100.
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.1 (47-111)
Determining discourse-based moves in professional reports Lynne Flowerdew Hong Kong University of Science and Technology
This chapter reports on a study which combines corpus-based and genrebased approaches to the analysis of a 225,000-word corpus of 60 environmental recommendation-based reports. I first describe the discourse-based coding system, which draws on the concept of genre move structure analysis, accounting for three different, but inter-related levels of text: (i) macrostructure; (ii) genre structure, and (iii) textual patterning, i.e. elements of the Problem-Solution pattern. I then describe the keyword analysis for the corpus as a whole and the key-key word analysis for each individual report. These keyword analyses provide internal linguistic evidence for classifying the reports as Problem-Solution based. An analysis of selected words (problem / problems and impact / impacts) reveals that their collocational behavior and involvement in certain causative phrases are related to specific discourse-based move structures.
.
Introduction
Corpus linguistic techniques have been criticized for encouraging a more bottomup rather than top-down processing of text (Swales 2002). This point is certainly true for those analyses in which truncated concordance lines are examined without recourse to the overall discourse, such as is the case where 2000-word samples of texts are used for compiling the corpus. However, a more top-down approach can be achieved by integrating corpus-based and genre-based approaches to analysis of whole texts (Flowerdew 1998, 2005). See Biber, Connor & Upton (2007) for corpus-based research which employs both top-down and bottom-up approaches in the exploration of discourse structure. In the last few years, some very fruitful genre-based corpus studies have been carried out based on Swales’ (1990, 2004) ‘move structure’ analysis. Such studies are mostly carried out on specialised corpora displaying fairly prototypical move structure patterning of a particular genre, e.g. introductions to guest speakers (Henry & Roseberry 2001), non-profit philanthropic letters (Upton 2002), job application letters (Upton & Connor 2001), and
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.2 (111-167)
Lynne Flowerdew
law cases (Bhatia et al. 2004). However, it has to be borne in mind that it is only a restricted range of fairly conventionalized genres which are amenable to the coding of move structures, and even then difficulties may arise as one may find embedding of move structures and possible genre-mixing (Flowerdew 2004). This chapter reports on a study which combines corpus linguistics and genreanalysis techniques in the investigation of recommendation-based reports, another type of professional documentation which is conventionalized to a large extent, and therefore lends itself to the identification and coding of prototypical move structures. The chapter first describes the corpus of reports and then presents the discourse-based codes that were devised for this particular corpus. The second part of the chapter presents an analysis of some selected keywords within this discourse-based move framework. The chapter concludes by tentatively answering the question as to whether certain linguistic features can be equated with the type or nature of a move, an issue raised in Swales (2004: 229).
. Corpus and coding . Description of corpus The corpus of professional recommendation-based reports comprises 60 reports totaling approximately 225,000 words. These reports, commissioned by the Hong Kong Environmental Protection Department from various consultancy companies in Hong Kong, document the potential environmental impacts of the construction and operation of proposed buildings and facilities. The reports also contain a section on suggested mitigation measures to alleviate any possible adverse impacts arising from the planned construction. The reports thus clearly follow a ProblemSolution rhetorical pattern which can be uncovered, for the corpus as a whole and for each individual report, via a keyword and key-keyword analysis, respectively. The keyword and key-keyword functions in WordSmith Tools (Scott 1997) identify words of unusually high frequency in the whole corpus or in each report through comparison with a reference corpus, thus providing valuable linguistic evidence for classifying the reports as Problem-Solution based (see Flowerdew 2003; 2008). . Rationale and description of coding system In the field of computational linguistics several types of taggers have been developed which, automatically or semi-automatically, assign part-of-speech tags to text (see, for example, Garside 1996). More recently, in the field of corpus linguistics, text linguists such as Connor et al. (2002) and Upton (2002) have devised coding systems to account for the rhetorical move structures of various genres such
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.3 (167-240)
Determining discourse-based moves in professional reports
Table 1. Framework of Causal Relations for problem (Flowerdew 2003) Causal relation
Example
Reason-result Means-result Grounds-conclusion Means-purpose Condition-consequence
. . . export scheme will create a noise problem. . . . thereby averting an odour problem. . . . and so flooding is not a serious problem. . . . in order to alleviate the problem of . . .. . . . If there is a problem with . . .
as job application letters and direct mail letters, where the moves are fairly clearly delineated. (As the identification of move structures can only be done on a purely manual basis, I shall use the terms ‘labelling’ or ‘coding’ here, rather than ‘tagging’, which involves an automatic or semi-automatic process.) For example, Connor, Precht and Upton (2002) identified and coded various move structures (e.g. applying for the position, indicating desire for an interview etc.) in job application letters written by applicants from Finland, Belgium and the US to ascertain which positive and negative politeness strategies were associated with what particular move structures and whether these were related to cross-cultural factors. Acknowledging the concept of move structures, I have attempted to compile a coding system which can account for three different, but inter-related levels of text: (i) Macrostructure (Introduction, Body, Conclusion); (ii) Genre structure (e.g. background, purpose and scope within the Introduction) and (iii) Textual patterning, i.e. elements of the Problem-Solution patterning inherent in all the reports in the corpus. For the first stage of this analysis, I extracted the keywords (e.g. impact; pollution; treatment) acting as signals for the Problem-Solution pattern, and investigated the various types of causal relations, if any, in which particular keywords occurred (Flowerdew 2003). Examples from the corpus of the various causal relations that problem is involved in are given in Table 1 above. In the next stage of analysis, which is the focus of this chapter, my aim was to ascertain whether the keywords, in both causative and non-causative phrases, could be related to any particular section of the macrostructure (Introduction; Body; Conclusion) and also to any particular genre move structure occurring within these three parts of the macrostructure. Below, I describe the decisions I made for labeling segments of text for achieving this aim. I first examined the macrostructure of each report, where a clear division into an Introduction, Body and Conclusion was evident for each report. Within these three sections, each main heading and sub-heading were examined together with their corresponding content, which were both used as the basis for drawing up the discourse-based coding scheme for delineating textual boundaries, i.e. the discourse-based move structures. This content-oriented approach for the division of textual boundaries has been commented on by Paltridge who, citing the genre analysis research of Bhatia, Swales and Crookes, notes thus:
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.4 (240-316)
Lynne Flowerdew
. . . all draw essentially on categories based on content to determine textual boundaries, rather than on the way the content is expressed linguistically. These divisions, furthermore, are often intuitive; it is by no means always the case that headings and other textual indicators are found to be present. . . .. indicators of boundaries are most clearly seen in terms of content; that is, it is a cognitive rather than linguistic sense that guides our perception of textual division. (Paltridge 1994: 295)
Although I initially assumed that the headings and sub-headings would be accurate indicators of content segments and thus play a pivotal role in devising the coding scheme, this did not always prove to be the case as discussed below. In the Introduction section, the headings and sub-headings did very much reflect the content within that sub-section, although the wording in the subheadings might vary quite widely. For example, the sub-headings ‘description of the scheme’, ‘conceptual design of facilities’, ‘airport plan’, ‘study area and the project’ do not share any terminology in common, but they all relate to the proposed project, and for this reason the content in these sections is coded <prop> in order to identify the proposal. All the headings and sub-headings were coded so that instances of phrases in headings and sub-headings could be distinguished from their instances in the text (see Appendix 1 for details of the coding scheme). An extract of coded text from an introduction is given below. For the most part, the coding of text in the introductions was fairly straightforward except in a few cases where sub-headings had a dual focus, namely ‘backINTRODUCTION Background to the Study The construction of China Light & Power’s large thermal power station at Black Point may affect the operation of the Island East Transfer Station (IETS) marine reception area at Yung Long. It is therefore proposed that a reception jetty be built at Phase 1 of the WENT landfill, with a marine access channel from Yung Long to the WENT landfill to be dredged to facilitate the navigation of the IETS vessels. . . Purpose of the Study The purpose of the assessment is to provide information on the nature and extent of the potential impacts associated with the dredging work. This information will contribute to decisions on: – the acceptability of any adverse environmental consequences that are likely to arise; and – the conditions required for dredging the marine access channel. Description of the development <prop>The proposed fairway will be the permanent marine access for the marine reception facilities for the IETS vessels at the WENT landfill. The 1790m access channel which covers an area of about 27 hectares, extends from . . .
Figure 1. Extract of coded text from an introduction section
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.5 (316-360)
Determining discourse-based moves in professional reports
ground and purpose’, and ‘objectives and scope’. I decided to code both of these as the objective was the main information in each sub-section and the background and scope were only contained within a short phrase embedded in the purpose statement. Another reason for streamlining the coding was that, in other reports, the sub-heading ‘objective’ was used which, in actual fact, also contained a brief reference to the scope of the project. Although one could argue for a more finely-grained coding system which would take account of content with a dual focus, I made the decision to adopt for the introduction sections only codes indicating a single focus as the content in these problematic sections was very much weighted towards one aspect. In contrast to the close correspondence between sub-headings and content in the Introductions, except in the few cases mentioned above, particular subheadings in the Body of the reports did not always accurately reflect the main content of the sub-sections. This was especially noticeable with headings relating to the problem element such as ‘impacts’, ‘environmental impacts’ and ‘findings of the EIA’, which also, in reality, dealt extensively with the possible solutions, i.e. mitigation measures to reduce the environmental impacts. Hence, I used the double coding, <prso>, for stretches of text with multiple information of equal status, although the heading might not have reflected this. Similar problems have been encountered in coding speech events in MICASE (Michigan corpus of Academic Spoken English; Simpson et al. 2002). For example, Simpson-Vlach & Leicher (2006) note that what was labeled as a ‘discussion section’ in the course notes, in fact, resembled more a small lecture. I also made a decision not to code the problem and solution elements separately, unless these were in two distinct sections where all the problems were first mentioned, e.g. ‘Environmental Impacts’, followed by a section which covered all the solutions, e.g. ‘Mitigation Measures’. This was for the following two reasons. First, having the coding <prso> in addition to the coding <prob> and <sol> enabled me to see whether the Problem-Solution pattern is organised on a ‘whole by whole’ basis (i.e. all the problems followed by all the solutions), in which case the coding would be <prob> and <sol> or on a ‘part by part’ basis where we have a succession of problem and solution elements, in which case the text segments would be assigned as a chain of <prso>, <prso>, <prso> labels, if three separate problems and corresponding solutions were recorded. Secondly, in the sections of text organised on a ‘part by part’ basis it was noticed that often the problem and solution elements were contained within the same sentence so this would have necessitated coding the text at a sentence level, which would have been very time consuming. However, a potential disadvantage of this streamlined coding scheme for the interpretation of results is that it is impossible to tell from the code <prso> whether a lexical item is part of the problem or solution element. For example, in Extract 1 below impacts is in a problem statement, whereas impact is part of the
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.6 (360-444)
Lynne Flowerdew
solution element. Where problem statements such as impact are found in solution elements they very often co-occur with verbs such as ‘avoid’. These two-way signaling verbs serve as a linking device between the problem and solution elements (Hoey 2001). Extract 1: Text segment containing impacts and impact under <prso> code Marine sediment <prso> The environmental impacts relating to transport and disposal of marine sediments are primarily due to the resuspension of marine sediments. During transport of marine fill from marine borrow areas to TMPD, loss of material from barges and hoppers may occur, thus increasing suspended solid concentrations in the water column. In order to avoid this potential impact it is recommended that the water quality in the vicinity of dredging and dumping sites have to be strictly monitored and controlled.
Similar to the case of impact in Extract 2, problem can also occur in the solution element, when a solution is proposed to a future hypothetical problem, or when a problem raised previously in the text is being referred to. Extract 2: Examples of problem under <sol> code <sol>If there is a problem with access to the normal monitoring position, an alternative position may be chosen, and a correction to the measurements shall be made. <sol>The entire leachate problem generated from the Tseung Kwan O landfills will be addressed comprehensively in the subsequent Restoration of Tseung Kwan O Landfills Study.
Moreover, as the coding scheme in Appendix 1 shows, the problem and solution elements have only been explicitly coded in the Body section of the report. Problem and solution elements also occur in the Introduction and Conclusion, but it was not necessary to code these separately as I am mainly interested in the textual distribution of various lexical items signaling the Problem-Solution pattern. A concordance for problems with the tag function activated would indicate whether a particular instance occurs in a particular move structure in the introduction or whether it occurs in the conclusion part of the reports, as illustrated in Extract 3 below. Extract 3: Examples of problems under codes used in the introduction and conclusion Introduction Section In addition to the problems of flooding, the watercourses within the study area are among the most polluted of the Territory. Conclusion Section Provided that mitigation measures are properly and fully implemented, it is considered that the works for 43CD and 30CD will not result in any insurmountable environmental problems.
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.7 (444-507)
Determining discourse-based moves in professional reports
The remaining sub-headings in the Body of the reports did not present any coding difficulties as they all accurately reflected the subsequent content. Although there was quite a lot of variation in terminology, there was a close match between the content and its corresponding sub-heading. For example, ‘existing environment/conditions’, ‘site description and operations’ and ‘study area’ all described the situation. Likewise, ‘environmental management system’ and ‘monitoring and auditing’ reflected auditing procedures. Of the three main sections of the reports, it was the Conclusion section that showed the least close match between the content and its sub-heading. Two main headings were mostly used, ‘Conclusion(s)’ and ‘Conclusions and Recommendations’, although the type of information contained under these two headings seemed to be very similar, as shown in the two extracts in Extract 4 below. The codes I adopted reflected this difference in labeling rather than content (see Appendix 1). Extract 4: Conclusion sections of two reports Conclusions The section of Route 5 is a priority project to reduce traffic congestion and environmental impacts in the central parts of Tsuen Wan. The EIA study has considered the magnitude and acceptability of all environmental impacts from the project and has concluded that the key issues will be noise during construction and operation. Mitigation measures have been considered to reduce noise from traffic and road enclosures and barriers have been recommended along parts of the road together with the application of a low noise road surfacing. Conclusions and Recommendations In general, road traffic noise will have the most pronounced impact in the area. Measures capable of satisfactorily mitigating the impact of noise have been developed. Subject to the implementation of these measures the environmental impact of the roads to be constructed under Contract No. TK/39 will not be significant.
The above discussion highlights the complexity of attempting to assign discoursebased labels to text, especially in view of the fact that the headings and subheadings do not always accurately reflect the content in their sections. In order to ensure as much reliability as possible, the student assistant in charge of coding the text manually double-checked the labels against my criteria for assigning the codes before entering them. The section below presents an analysis of the singular and plural forms of two items, discussed within the framework of the discourse-based coding scheme.
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.8 (507-649)
Lynne Flowerdew
. Analysis of selected items for the problem element within discourse-based moves As mentioned previously, in the first stage of this analysis, I extracted the keywords which defined elements of the Problem-Solution pattern. Table 2 presents the keywords occurring in ten or more reports, relating to the problem element. Although problem and problems did not surface as keywords in the corpus, they are worth examining further as they can be regarded as superordinate terms for all of the keyword nouns listed in Table 2. I now describe the textual distribution of problem(s) and the keywords impacts(s) with reference to the discourse-based coding scheme outlined in the previous section. I also consider whether these words co-occur with items marking any type of causal relation and if these occupy any particular move structure slot, as so defined in my coding system. In the following section, I will first examine the singular form of problem, followed by problems, and secondly, the noun impact followed by its plural form impacts. . Analysis of ‘problem’/‘problems’ The frequencies of occurrence for problem and problems are 41 and 51 instances, respectively, as shown in Table 3 on the following page. A good starting point for looking at these items from a more discoursal perspective is to determine where exactly in individual reports these instances occur. As Baker (2006) points out, texts have beginnings, middles and ends and that it may be relevant to know whether a particular word form is more likely to occur at the start or at the end of the text. The number of words per section for problem and problems given in Table 3 below indicates that these different forms of the same lemma have somewhat different distributional tendencies. Problem tends to cluster in the middle sections, with 83%, i.e. 34 out of 41 tokens found here, but only 4 tokens occurring in the beginning section. Problems, meanwhile, displays slightly different tendencies, generally occurring in the beginning as well as in the middle sections. 39%, i.e. 20 out of the 51 tokens for problems are found in the introduction sections, and 57%, i.e. 29 out of the 51 tokens in the body sections. Table 2. Keywords for the Problem element impacts (50) noise (44) waste (20) pollution (10) contaminated (14)
impact (26) traffic (23) dust (20) emissions (10)
sewage (12) sediments (10)
Note: Figures in parentheses denote the number of texts in which the words were found to be key.
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.9 (649-678)
Determining discourse-based moves in professional reports
Table 3. Breakdown of tokens for problem/problems by codes and macrostructure Frequency of problem Introduction <prop> <sit> Sub-total Body <prso> <prob> <sol> Sub-total Conclusion Sub-total Total
Frequency of problems
– – 4 – – 4 (10%)
7 1 4 2 6 20 (39%)
20 3 10 1 34 (83%)
22 1 4 2 29 (57%)
2 1 3 (7%)
2 – 2 (4%)
41 (100%)
51 (100%)
I will now take a closer look at the behaviour of problem and problems found in the various discourse-based move structures of different sub-sections of the reports, especially in relation to the causal relations exemplified in Table 1. Over 90% of the tokens for problem and problems occurred is some type of causative construction, the most commonly being with a causative verb, e.g. the export scheme will create a noise problem, signalling the Reason-Result relation. However, a manual examination of the expanded concordance lines checked against the codes revealed that this type of patterning could not be identified with any particular move structures across the three main sections of the reports. It has already been noted that the majority of the tokens for problem occur in the body, with 20 tokens under <prso>, three under <prob> and 10 coded under <sol>. A comparison of the phrases containing problem coded under <prso> with those under <prob> and <sol> reveals some interesting differences. It was found that problem occurred in a Means-Purpose relation six times incorporating a two-way signalling verb such as ‘ameliorate’, ‘improve’, ‘alleviate’, which provide a linkage between the problem and solution elements, as in (1): (1) In order to alleviate the problem of noise, . . .
All six of these instances were accounted for by the <prso> tag. This data therefore suggests that the Means-Purpose relation has a tendency to be used when reference
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.10 (678-792)
Lynne Flowerdew
is made to a problem in the immediately preceding text (i.e. for text coded <prso>, which reflects the ‘part-by-part’ organisation), but does not tend to occur if the solution aspect is separated from the problem aspect (i.e. for text coded <sol>). Here, the 10 instances of problem coded <sol> were mostly found in phrases referring back to a previously introduced problem, e.g. Because of this potential problem. . . ; . . . as for the analysed noise problem. . . ; Nevertheless, the noise problem generated by additional traffic. . . , before the proposed solution was introduced. Another observation was that problems appears to have a different collocational patterning in the introduction and body sections of the reports. In the introductions, it is found to collocate with a more general adjective such as ‘environmental’, as in Example (2) coded , rather than specific ones (e.g. ‘noise’, ‘traffic’) which are found in the body of the reports under the <prso> tag, as in Example (3): (2) Where potential environmental problems may arise, mitigation measures were recommended . . . (3) Severe traffic noise problems already exist in TMNT. . . . The traffic noise impact on Butterfly Estate can be relieved following the introduction of the Foothills Bypass. . . <prso>
The fact that problems is a significant feature of the introductions, whereas problem has this role in the body sections of the reports, is suggestive of the fact that problems is probably used in a topic-like sentence in the introductions, and then each problem, in turn, is itemized in the body of the reports. Moreover, half of the tokens for problems found in the Means-Purpose relation were also coded <prso>, with none coded <sol>, indicating that in the body sections of the reports the Means-Purpose relation is used for binding at a local level. However, an examination of the tokens for problems in the introduction sections showed that the purpose statement referred to problems expressed in a very general way, or to problems that would be elaborated in the body of the reports, as in Example (4): (4) The EDS identified that highway improvement works would be required to overcome the anticipated traffic problems on Lung Mun Road. . .
Therefore, whether the Means-Purpose relation in which problems is found is used for local or more textual binding seems to be dependent on its positioning in the overall discourse structure. . Analysis of ‘impact’/‘impacts’ A breakdown of the tokens for impact and impacts across the various move structures in each macro-section of the reports is presented in Table 4. Interestingly,
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.11 (792-820)
Determining discourse-based moves in professional reports
Table 4. Breakdown of tokens for impact/impacts by codes and macrostructure Frequency of impact
Frequency of impacts
Introduction <prop> <meth> <scope> Sub-total
16 22 6 8 1 32 14 48 147 (22.5%)
18 30 9 6 4 57 51 30 225 (23.5%)
Body <sit> <prso> <prob> <sol> Sub-total
7 311 69 55 20 462 (71%)
13 422 81 42 45 603 (63%)
Conclusion Sub-total Total
29 14 43 (6.5%)
104 24 128 (13.5%)
652 (100%)
956 (100%)
in percentage terms, the frequencies of impact and impacts are reasonably similar across the three sub-sections. I will now examine some of the key findings of impact and impacts within the different move structures of each macro-section of the report and compare these with their superordinate terms problem and problems discussed in the previous section. As in the previous section, the discussion will focus on their involvement in the Means-Purpose and Reason-Result relations, and their collocational preference for different adjectives in different sections of the reports. Nearly half (311) of the total number of tokens for the singular form impact are found under the <prso> tag in the body. In 51 cases impact is involved in a Means-Purpose relation, and it is to be noted that 28 of these instances fall under the <prso> tag. The remaining 23 Means-Purpose phrases are mostly found under and , with only three phrases found under <prob> and <sol>. This is very similar to the distribution for Means-Purpose phrases with impacts, where 30 out of 66 tokens for impacts in a Means-Purpose relation are coded <prso>. The fact that, as in common with problem and problems, Means-Purpose phrases tend to cluster under the <prso> tag lends further weight to the argument
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.12 (820-880)
Lynne Flowerdew
that they act as a signal of local coherence in a part-by-part organizational structure rather than a signal of global coherence associated with a whole-by-whole organisational structure for the Problem-Solution pattern in the body section of these reports. One key aspect which sorting impact under the codes throws up is its inclusion in the fixed phrase ‘Environmental Impact Assessment/Study’, only found in the discourse-based moves in the introduction sections. For example, of the 22 phrases for impact coded , indicating background to the study, 19 are examples of this fixed phrase. This term is also very obvious in those phrases coded where 39 out of the 48 instances of impact are found in this fixed phrase, relating either intertextually to a previous study, as in Example (5), or to the purpose of the report in question, as in Example (6): (5) An Environmental Impact Assessment (EIA) Study has been undertaken. . . (6) The purpose of this Environmental Impact Assessment is to provide information on. . .
The sorting by codes also throws up differences in premodifying adjectives between the collocations of impact coded under <prob> and those under <prso>. It was found that 8 of the tokens for impact under <prob> were preceded by ‘direct’ and 10 tokens by ‘indirect’, although these adjectives were found to occur only once each with impact under <prso>. In contrast, a more specific noun modifier such as ‘noise’ was found to collocate with impact 36 times in those phrases coded <prso>, but only three times with those phrases coded <prob>. These adjectival collocations revealed by the sorting according to codes are thus suggestive of the two different organisational structures for encapsulating the Problem element in the body of the reports: in the ‘part-by-part’ structure (coded <prso>) one would expect more specific noun classifiers such as ‘noise’, ‘air’ to be used, as in Example (7), whereas in the ‘whole-by-whole’ structure (coded <prob>) it is not surprising to find more general adjectives such as ‘environmental’, ‘direct’, ‘indirect’ for describing the problems, as shown in (8). (7) Dust emission sources on site include site preparation/formation works and storage/handling of loose aggregates. On-site mitigation measures to reduce dust impacts to acceptable levels include:. . . (8) Indirect impact would result from disruption or loss of amenity to adjoining land uses including: disruption to traffic, restriction on access, noise pollution. . .
Turning now to impacts, it is interesting to note that 45% of the tokens for impacts (422 tokens out of the total of 956) fall under the <prso> tag in the Body section, which is a very similar distribution pattern to the tokens for impact. Moreover, it is in this move structure where certain patterns associated with particular causal
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.13 (880-949)
Determining discourse-based moves in professional reports
relations have a tendency to occur. For example, out of the 82 phrases with impacts signalling Means-Purpose, as in (9), 32 were coded <prso>. (9) . . .mitigation measures are recommended to reduce impacts on water quality.
Of the 50 remaining purpose phrases 22 occur in various move structures in the introduction, with 13 under . Another 10 are coded , 10 coded and only two tokens are found under <prob> and six under <sol>. What is striking about these Means-Purpose phrases are the differences in lexis across the different move structures. For example, the verbs ‘assess’, ‘evaluate’ and ‘mitigate’ are the most common in phrases coded , as in (10), whereas the verbs ‘reduce’ and ‘minimise’ are found in phrases coded <prso>, as in (11). (10) . . .to assess in detail the environmental impacts of the Construction Proposals (11) . . .in order to reduce potential noise impacts.
Causative verbs signalling Reason-Result were concentrated in the move structures in the body section of the reports. The non-causation phrases with impacts, on the other hand, were mostly found in all the move structures of the introduction section and in those move structures associated with the monitoring and assessment aspect of the impacts. This kind of information is contained in the phrases coded (which are not purpose-related) and in the phrases coded dealing with assessment, as in (12). (12) . . .the EIA assessment has identified adverse impacts in relation to the following.
Given that the reports usually discuss several types of environmental impacts, it is not surprising to find that in the conclusion sections there are more occurrences of impacts than impact, with many of the phrases embodying the company’s overall assessment of the situation, as in (13). (13) In view of the low potential for environmental impacts from the operation of the GRS, no environmental monitoring is considered necessary. . .
. Conclusion Swales (2004: 229) reviews various genre studies in which grammatical features have been found to indicate the type or nature of a move. Citing Gledhill (2000), Swales notes that Gledhill found that “was to” was used to signal the onset of an introduction’s third move, i.e. outlining the purposes of the present study. Likewise, lexical signals can also be indicative of move structures, the most obvious of which would be lexis indicating section headings, such as method and results, e.g. The results are shown in Table 1.
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.14 (949-1019)
Lynne Flowerdew
The research reported on in this paper has suggested that in addition to individual grammatical or lexical items, a consideration of lexical items together with their involvement in causative phrases and collocational behaviour could also signal move structures. This point has been especially borne out by an analysis of problem/problems and impact/impacts, which were found to be involved in a Means-Purpose relation under the sections of the reports coded <prso>. Moreover, these purpose clauses were often accompanied by verbs such as ‘reduce’ and ‘minimise’ linking the Problem and Solution elements. Although this is quite a small-scale study, and the results are suggestive rather than definitive, nevertheless, the coding scheme outlined in this paper has proved useful for shedding light on corpus data from a more discourse-based perspective. This research has also shown that the field of corpus linguistics can make use of genre-based methodologies.
References Baker, P. 2006. Using Corpora in Discourse Analysis. London: Continuum. Bhatia, V. K., Langton, N. & Lung, J. 2004. Legal discourse: Opportunities and threats for corpus linguists. In U. Connor & T. Upton (eds), 203–231. Biber, D., Connor, U. & Upton, T. 2007. Discourse on the move: Using corpus analysis to describe discourse structure. Amsterdam: John Benjamins. Connor, U. & Upton, T. (eds). 2004. Discourse in the Professions: Perspectives from corpus linguistics. Amsterdam: John Benjamins. Connor, U., Precht, K. & Upton, T. 2002. Business English: Learner data from Belgium, Finland and the U.S. In Computer Learner Corpora, Second Language Acquisition and Foreign Language Teaching, S. Granger, J. Hung & S. Petch-Tyson (eds), 175–194. Amsterdam: John Benjamins. Flowerdew, L. 1998. Corpus linguistic techniques applied to textlinguistics. System 26 (4): 541– 552. Flowerdew, L. 2003. A combined corpus and systemic-functional analysis of the ProblemSolution pattern in a student and professional corpus of technical writing. TESOL Quarterly 37(3): 489–511. Flowerdew, L. 2004. The argument for using English specialised corpora to understand academic and professional language. In Discourse in the Professions: Perspectives from corpus linguistics, U. Connor & T. Upton (eds), 11–33 Amsterdam: John Benjamins. Flowerdew, L. 2005. An integration of corpus-based and genre-based approaches to text analysis in EAP/ESP: Countering criticisms against corpus-based methodologies. English for Specific Purposes 24: 321–332. Flowerdew, L. 2008. Corpus-based analyses of the problem-solution pattern: A phraseological approach. Amsterdam: John Benjamins. Garside, R. 1996. The robust tagging of unrestricted text: The BNC experience. In Using Corpora for Language Research, J. Thomas & M. Short (eds), 167–180. London: Longman. Gledhill, C. 2000. The discourse function of collocation in research article introductions. English for Specific Purposes 19(2): 115–135.
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.15 (1019-1118)
Determining discourse-based moves in professional reports
Henry, A. & Roseberry, R. L. 2001. Using a small corpus to obtain data for teaching a genre. In Small Corpus Studies and ELT, M. Ghadessy, A. Henry & R. L. Roseberry (eds), 93–133. Amsterdam: John Benjamins. Hoey, M. P. 2001. Textual Interaction. London: Routledge. Paltridge, B. 1994. Genre analysis and the identification of textual boundaries. Applied Linguistics 15(3): 288–299. Scott, M. 1997. WordSmith Tools [Computer software]. Oxford: OUP. Simpson-Vlach, R. & Leicher, S. 2006. The MICASE Handbook: A resource for users of the Michigan Corpus of Academic Spoken English. Ann Arbor MI: The University of Michigan Press. Simpson R. C., Briggs, S. L, Ovens, J. & Swales, J. M. 2002. The Michigan Corpus of Academic Spoken English. Ann Arbor MI: The Regents of the University of Michigan. Swales, J. M. 1990. Genre Analysis: English in academic and research settings. Cambridge: CUP. Swales, J. M. 2002. Integrated and fragmented worlds: EAP materials and corpus linguistics. In Academic Discourse, J. Flowerdew (ed.), 150–164. London: Longman. Swales, J. M. 2004. Research Genres. Cambridge: CUP. Upton, T. 2002. Understanding direct mail letters as a genre. International Journal of Corpus Linguistics 7(1): 65–85. Upton, T. & Connor, U. 2001. Using computerized corpus analysis to investigate the textlinguistic discourse moves of a genre. English for Specific Purposes 20(4): 313–329.
JB[v.20020404] Prn:27/03/2008; 9:07
F: SCL3106.tex / p.16 (1118-1165)
Lynne Flowerdew
Appendix 1: Coding scheme for professional reports Code Specification
Code Name
Report heading/sub-heading
Headings: main headings sub-headings
Introduction: introduction
introduction introduction and objectives
foreword scope
<scope>
foreword scope
background
background study context
objectives
objectives objectives and key issues key issues to be addressed purpose of the study background and purpose objectives and scope
structure