(developments In English As A Lingua Franca) Beyza Björkman-english As An Academic Lingua Franca-de Gruyter Mouton (2013).pdf

English as an Academic Lingua Franca

Developments in English as a Lingua Franca 3

Editors

Jennifer Jenkins Will Baker

De Gruyter Mouton

English as an Academic Lingua Franca An Investigation of Form and Communicative Effectiveness

By

Beyza Björkman

De Gruyter Mouton

ISBN 978-3-11-027914-6 e-ISBN 978-3-11-027954-2 ISSN 2192-8177 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.dnb.de. ” 2013 Walter de Gruyter, Inc., Boston/Berlin Printing: Hubert & Co. GmbH & Co. KG, Göttingen 앝 Printed on acid-free paper 앪 Printed in Germany www.degruyter.com

To Ella and Mikael, my wonderful family

Acknowledgments

This monograph takes as its basis my PhD project that I completed in June 2010. During the production of this manuscript, I have received invaluable help from many colleagues, friends, and family. First of all, I would like to thank De Gruyter Mouton for seeing the importance of the topic and investing in this monograph. Birgit Sievert, Julie Miess and Angelika Hermann patiently answered every question I had. The editors Professor Jennifer Jenkins and Dr. Will Baker have helped me shape this manuscript to the monograph you have before you today. Needless to say, any remaining errors are mine. I am especially grateful to Jennifer Jenkins for being so supportive and helpful from the very early stages of this project, when this book was only an idea. The research project that forms the basis of this book was almost exclusively financed by the language unit at the Technical Institute despite all the financial hardships, and I am eternally grateful for this support. I will not be able to mention any names here for purposes of anonymity, but I am indebted to my colleagues for their hard work that generated the resources for this project as well as their never-ending encouragement. I am also grateful to the English Department at Stockholm University where I was a doctoral student for letting me work on a topic that I knew was very important and for providing financial support for the final stages of the project. I would also like to thank my colleagues at Roskilde University for the feedback and help during my post-doc stay in Denmark. I have benefited greatly from discussions with my PhD supervisor Professor Philip Shaw, who has been extremely generous with truly inspiring discussions and invaluable advice. Special thanks go to Professor NilsLennart Johannesson for creating special characters so I could be true to my Early Modern English source. Thanks also go to Christina Alm-Arvius, Britt Erman, Rebecca Hincks, Nils-Lennart Johannesson, Elizabeth Keller, Andy Kirkpatrick, Maria Kuteeva, Margareta Lewis, Magnus Ljung, Hans Malmström, Gunnel Melchers, Alan McMillion, David Minugh, Jan Pedersen, Erik Smitterberg and Annelie Ädel. I have learned so much from them. There are other colleagues who have helped me with feedback and encouragement. I cannot help but mention John Airey, Anne Kari Bjørge, Gibson Ferguson, Gregory Garretson, Rachel Giora, Hartmut Haberland, Spencer Hazel, Glenn Ole Hellekjær, Jennifer Jenkins, Anna Mauranen, Janus Mortensen, Diane Pecorari, Elina Ranta, John M. Swales and Hedda Söder-

viii

Acknowledgments

lundh. I must thank Jennifer Jenkins, Anna Mauranen and Barbara Seidlhofer separately for paving the way for numerous studies on ELF, for being so supportive to us new ELF researchers, and for being so generous with their expertise. And finally my family whom I love tremendously. It is to you that I owe everything. I have produced this monograph while taking care of my baby girl Ella. I am deeply grateful to my husband Mikael for his sunny smile, never-ending patience and encouragement, and for shouldering so much of the household responsibilities. Without his help, I would not have been able to complete this book. And our sunshine Ella… Thanks for coming into my life and showing me what really matters. I dedicate this book to you both.

Earlier versions of some of the text in this monograph, and some of my thoughts and arguments, have appeared in the following papers that I have produced. I am grateful to the publishers and editors below for permitting me to include them in this book: 2012 2012

2012 2011

2011

2010 2010

2009

Questions in academic ELF interaction. Journal of English as a Lingua Franca 1 (1): 93–119. Investigating English as a lingua franca in applied science education: Aims, methods and norms. In (Re-)Conceptualising LSP Research: Methods and Aims, Linguistic Insights, Studies in Language and Communication, Margrethe Petersen and Jan Engberg (eds.), 163– 187. Bern: Peter Lang. The grammar of English as a lingua franca. The Encyclopedia of Applied Linguistics, (ed.). Oxford, UK: Wiley Blackwell. English as a lingua franca in higher education: Implications for EAP. Ibérica 22: 79–100. Special Issue on EAP in Parallel language and ELF Settings. Pragmatic strategies in English as an academic lingua franca: Ways of achieving communicative effectiveness? Journal of Pragmatics 43(4): 950–964. So you think you can ELF: English as a lingua franca as the medium of Instruction. Hermes 45: 77–99. Spoken lingua franca English at a Swedish technical university: An investigation of form and communicative effectiveness. Unpublished PhD thesis. Stockholm University, Department of English. From code to discourse in spoken ELF. In English as a Lingua Franca: Studies and findings, Anna Mauranen, and Elina Ranta (eds.), 225–254. Cambridge: Cambridge Scholars Press.

Acknowledgments 2009

2008

2008

ix

English as a lingua franca at a Swedish technical university: An effective medium? In Proceedings of the Annual BALEAP Conference: 'EAP in a Globalising World: English as an Academic Lingua Franca‘. Whong, Melinda (Ed.), 11–19. Reading: Garnet Education. English as the lingua franca of engineering: The morphosyntax of academic speech events. Nordic Journal of English Studies 7(3): 103–122. 'So where we are': Spoken lingua franca English at a Swedish technical university”. English Today 24 (2): 11–17. Reprinted with permission from Cambridge University Press, Cambridge Journals Online.

Contents Chapter 1 ......................................................................................................1! Introduction ...............................................................................................1! 1.1. English today ..................................................................................3! 1.2. English as the language of science and technology .......................6! 1.3. Globalization and English in higher education ............................13! 1.3.1. Ideological responses to globalization ..................................17! 1.3.2. Globalization and multilingualism in continental Europe ....18! 1.3.3. Globalization and multilingualism in Sweden ......................23! 1.3.4. The answer: English as a lingua franca .................................28! Chapter 2 ....................................................................................................31! Previous research on ELF .......................................................................31! 2.1. The relevance of normative elements in ELF research ................32! 2.2. Work without normative elements: Pragmatics ...........................33! 2.3. Work including normative elements: Form .................................43! 2.4. When form and pragmatics meet .................................................52! 2.5. Attitude and irritation studies .......................................................54! Chapter 3 ....................................................................................................59! Exploring an academic ELF setting in Sweden: The site .....................59! 3.1. An international university: Student and teacher body ................60! 3.2. Research questions and some important terms ............................62! 3.3. Research methodology and design ...............................................65! Chapter 4 ....................................................................................................82! Operating in a Swedish ELF site .............................................................82! 4.1. Dimension 1: Form ......................................................................82! 4.1.1. Commonalities of usage ........................................................82! 4.1.2. Non-standard usage vs. standard usage .................................94! 4.2. Dimension 2: Communicativeness ...............................................98! 4.2.1. Investigating overt disturbance .............................................98! 4.2.2. Analyses at discourse level: Pragmatic strategies ...............123! 4.3. Perceived communicativeness and attitudes: A survey of student attitudes .............................................................................................138! 4.4. Summary of results ....................................................................141! Chapter 5 ..................................................................................................147! Theoretical and practical implications ..................................................147! 5.1. Discussion of the findings ..........................................................147! 5.2. Theoretical implications .............................................................160! 5.2.1. The status of ELF ................................................................160! 5.2.2. Norms and standards for speech .........................................173!

xii

Contents

5.2.3. Good English .......................................................................176! 5.3. Practical implications .................................................................178! 5.3.1. Comprehension-facilitating lecturing behavior ...................179! 5.3.2. Support for lecturers and students who need to operate in ELF settings ..................................................................................185! 5.3.3. Pedagogical applications: Issues for the language classroom .......................................................................................................189! Chapter 6 ..................................................................................................197! Looking ahead .......................................................................................197! 6.1. Summary ....................................................................................197! 6.2. Notes for native speakers ...........................................................201! 6.3. The internationalization of higher education and language policy practices .............................................................................................205! 6.4. ELF: Final remarks ....................................................................210! Appendices ............................................................................................212! Appendix 1: Sample lecture transcription .........................................212! Appendix 2: Sample group-work transcription .................................219! Appendix 3: Sentences used in the questionnaire .............................232! Appendix 4: The observation protocol ..............................................234! References ................................................................................................235! Index .........................................................................................................261!

Abbreviations ARS BASE BNC CANCODE CL CLIL ELF ELFA ELT ENL ERASMUS IELTS IA/P L1, L2 LFC LLC MICASE NNS NonS NP NS PL QLs QRLs QSs SOV SVA SVO TESOL TOEFL TSE VOICE VSO VP WO WrELFA

Audience Response System British Academic Spoken English British National Corpus Cambridge-Nottingham Corpus of English Clause Level Content and Language Integrated Learning English as a Lingua Franca Corpus of English as a Lingua Franca in Academic settings English Language Teaching English as a Native Language European Region Action Scheme for the Mobility of University Students International English Language Testing System Interrogative Adverb/Pronoun First language, Second language Lingua Franca Core London-Lund Corpus of Spoken English The Michigan Corpus of Academic Spoken English Non-native Speaker Non-standard Noun Phrase Native Speaker Phrase Level Questions asked by Lecturers Rhetorical Questions asked by Lecturers Questions asked by Students Subject-Object-Verb Subject-Verb Agreement Subject-Verb-Object Teaching English to Speakers of Other Languages Test of English as a Foreign Language Test of Spoken English Vienna Oxford International Corpus of English Verb-Subject-Object Verb Phrase Word Order Written academic ELF

Chapter 1 Introduction

This book is about the widest use of English in the world today: English as a lingua franca (ELF). ELF is defined as “any use of English among speakers of different first languages for whom English is the communicative medium of choice, and often the only option” (Seidlhofer 2011: 7). English today is a lingua franca which brings millions together in a wide range of communicative situations in numerous settings for a broad spectrum of purposes. As you are reading these lines, a very large number of people with different first languages are communicating through English as a lingua franca in business meetings, in conferences and other academic discussions, or sports activities, to name a few. Businessmen are busy trying to land deals, academics are giving lectures or having research meetings, university students are working out the details in their new institutions, and all of this, they do through English as a lingua franca. English, in this sense, has reached truly global dimensions no other language has come near before. It is used in a very large number of domains, spoken by millions of people for different purposes. This is not to say that there are no other lingua francas. Other languages are used as lingua francas centralized in particular regions in the world, such as Russian and Spanish; however, “it is English and English alone that can reasonably claim to have become a global lingua franca” (Van Parijs 2011: 11). In the present context, ‘lingua franca’ is used in a different sense from the original meaning of the term. The original term ‘lingua franca’ refers to the oldest pidgin for which there is a reasonable amount of data for investigation (Parkvall 2005). It has been suggested that the meaning of the term comes from Arabic and Greek. Before the Crusades and during the Middle Ages, Western Europeans were referred to as ‘Franks’ in Arabic and ‘Phrankoi’ in Greek during the times of the late Eastern Roman Empire. So lingua franca was the language of the Franks, and it was a mixed trade language used by the language communities around the Mediterranean to communicate with others, as these comunities did not share a common language. It consisted mainly of Italian mixed with Turkish, French, Greek, Arabic, Portuguese and Spanish. It had limited vocabulary and grammar, and it lacked verb tenses and case– it did however develop a past and a future tense around the seventeenth century during its golden age (Corré

2 Introduction 2005). Pidginists have maintained that the earliest text in lingua franca goes back to 1353, and there are traces of its use from the twentieth century. This long period of time suggests that it may well be “the most long-lived pidgin language we know of” (Parkvall 2005). The original lingua franca had the same purpose with today’s lingua franca English: It was used by speakers from different first language backgrounds as a vehicular language. Those involved in trade had to sell and buy goods through a common language, and with the Mediterranean lingua franca, they were able to do so. There are important differences, however, between the original lingua franca and today’s lingua franca English. Today’s lingua franca is obviously not a mix of languages, nor does it have limited vocabulary or syntax the way the original lingua franca did. The original lingua franca was mostly spoken, and not so often written, as it was a contact language. When it was written, it was generally in early opera libretti and ballads, and this was done generally to include exotic elements in these works (Corré 2005). Otherwise, the original lingua franca was merely a practical language and not a literary medium. This is unlike today’s lingua franca English, which is used in several domains, both in spoken and written form (see e.g. the WrELFA corpus project). Perhaps most importantly, the original lingua franca was not expanded or nativized anywhere (Parkvall 2005) unlike English, which is the native language of a number of countries. English is the only language in history to have countries where it is the native language and to have become a truly global lingua franca. This is surely a fascinating linguistic phenomenon. It is, however, not only a linguistic phenomenon. English gained the lingua franca status as a result of a series of political events and other significant historical developments, becoming the language of several domains, such as higher education. In this introductory chapter, I will consider these developments, alongside the reactions and ideological responses raised to one language gaining such a powerful status. We will start by a brief review of some historical developments and then move on to how English has become the language of science and technology. Turning to reactions and ideological responses, we will go through some of the arguments raised by concerned academics who have been arguing for over a decade that ELF is simply about hegemony, Anglification and the spreading of market economics, and that minor languages are losing one domain after another against English. The chapter will end with a brief introduction of English as a lingua franca, which is

1.1 English today

3

another perspective one could adopt with regard to the widespread use of English in Europe today.

1.1. English today “…although the global spread of English as a lingua franca belongs very much to the present, it needs to be put in perspective by reference to the past”. (Seidlhofer 2011: xi)

To be able to understand the phenomenon of English as a lingua franca, we need to refer to the past and see the events that have led to English gaining such a global status. To many, the nineteenth century was the time when English first reached a strong global position. By the end of the nineteenth century, Britain, with its many communities of English speakers settling around the world along with trade, enabled English to become the world’s lingua franca. Consequently, French declined outside its borders, and different varieties of English around the world emerged and were “partially standardized” (Graddol 1997: 7). It was, however, the rise of the US as a superpower in the world that bolstered the dominance of English. The US quickly became the most powerful industrialized country because of its natural and human resources, and the fact that it was not destroyed by war unlike the countries in mainland Europe that had been war zones. As the world’s third largest country with reference to population, it is the country that has the largest group of native speakers of English in the world (Graddol 1997: 8). While the expansion of English to so many domains in the world is generally regarded as a result of the expansion of Great Britain with its colonies and the dominance of the US after World War II, there are other factors to consider. It is true that there are political developments that contributed to the growth of English; however, the need to communicate knowledge was a bigger factor. The growing needs of sharing and disseminating knowledge in the twentieth century required a language for all communication to take place in. This was true for scientific and technical knowledge dissemination as well as for other areas, such as for commerce and consumer culture (Graddol 1997: 14). English in the twenty-first century is used predominantly by three main groups of speakers: Those who speak it as their native language, those who speak it as a second (or additional) language and those who have learned it as a foreign language. A classical view of these groups was expressed by

4 Introduction Kachru as “inner circle”, “outer circle” and “expanding circles”, well known to linguists (Kachru 1985). In the inner circle are countries where English is the native language, i.e. the USA, the UK, Ireland, Canada, Australia and New Zealand. The outer circle countries are those where English has some historical significance and is spoken as the language of some or all of the country’s institutions, e.g. Singapore, India, Nigeria, South Africa, the Philippines. These two circles, however, do not represent the major groups of speakers of English in the twenty-first century. The largest group is the expanding circle countries, where English is a foreign language, e.g. China, Japan, Poland etc. These countries do not have a history of being colonized by any of the inner-circle countries, and English does not have much intra-national function.

Inner circle, 320–380 Outer circle, 150–300 Expanding circle, 100–1000

Figure 1.1. Approximate numbers of speakers in millions in inner, outer and expanding circle countries (Kachru 1985; numbers from Graddol 1997).

Kachru uses the term “norm-providing” for inner-circle countries, “normdeveloping” for outer-circle countries and “norm-dependent” for the expanding-circle countries. When the outer and expanding circles are merged, we have before us the largest group of users of the English language in the world. In a way, the numerical balance has shifted tremendously from the inner circle countries to these two groups of countries. Today, English is used predominantly by its non-native speakers as a lingua franca. Much happened demographically in what is now almost three decades as a result of globalization. Kachru’s description of the speakers in each circle no longer reflects the reality of the linguistic situation in the world. The Three Circles Model has undoubtedly been helpful in addressing the different groups of speakers of English who use English for a variety of

1.1 English today

5

functions. It provided a description of different types of users, thereby making an important contribution to showing the changing balance of speakers in the world by its inclusion of outer and expanding circles. However, the model is now dated, and it falls short of accurately accounting for the true heterogeneity of English today. It is very much a debated issue whether the inner circle speakers should be norm-providing for the speakers in the other two circles. The model, therefore, has been criticized for failing to show the true dynamics of the usage of English today (Jenkins 2009). World Englishes and ELF researchers together point to a need for pluricentralism (Bruthiaux 2003; Seidlhofer 2003). The applicability of the model has even been contested on the grounds that the model “perpetuates the very inequalities it otherwise aims to combat, such as the distinction between native and non-native speakers” (Park and Wee 2009: 390). Kachru however, never claimed these circles would stand against time and against changes that take place as speakers move around the world for a number of reasons and use English as a vehicular language. He himself acknowledged this, that the Three Circles Model may be somewhat simplistic and that it shows less awareness of the grey areas (Kachru 1985; Rajadurai 2003: 113). On the scale we are witnessing today, English is being used as the working language of many international domains. The twelve major international domains Graddol listed in 1997 have continued to use English increasingly as their working language (Graddol 1997: 8): 1. Working language of international organizations and conferences 2. Scientific publication 3. International banking, economic affairs and trade 4. Advertising for global brands 5. Audio-visual cultural products, e.g. TV, popular music 6. International tourism 7. Tertiary education 8. International safety 9. International law 10. In interpretation and translation as a relay language 11. Technology transfer 12. Internet communication

6 Introduction A significance of Graddol’s list above is its inclusion of at least eight domains that are so-called élite domains (1, 2, 3, 7, 8, 9, 10, 11). As the number of people using English in the above domains keeps expanding, English is likely to maintain its position as the most dominant foreign language among the world’s languages. The spread of English as the working language to so many of the world’s élite domains has evoked a variety of reactions. These reactions range from fascination to combative rejection. English as the most dominant language has been investigated and debated widely (sections 1.3.1, 1.3.2 and 1.3.3); however, these investigations and debates have had political issues rather than linguistic ones as their foci. The linguistic community has been relatively slow in dealing with the unprecedented growth of English (Mauranen 2003: 513). Although much work has been done in the last decade or so, detailed linguistic investigations are scarce. The extremely dynamic nature of the speakers of English today, i.e. that they travel and use English for a variety of purposes, makes it complicated to carry out investigations. With the situation today, where English has no boundaries, we need to explore the use and usage of English, we need more information on how it is used in the domains where it serves as the working language, but perhaps most importantly, we need to be able to cater for the needs of all those from different language backgrounds who use it as a vehicular language, as a lingua franca (ELF), so that we can help them compete on equal terms. The vast number of its users and the domains in which it is used bring a legitimate need for knowledge of the use of English as a lingua franca. The study used as the basis of the present monograph is an attempt to contribute to the existing knowledge on the use of English today in one of the aforementioned domains: Tertiary education. It investigates English in an academic engineering setting as spoken by lecturers and students, by those who use it as an academic lingua franca.

1.2. English as the language of science and technology The presence of English in Europe today can be observed in many domains. Scientific and technical domains are two such domains where English dominates over other languages (James 2000), and it is English only that is now the dominant lingua franca of science and technology in the world. Science is and has always been a global enterprise, and academic

1.2 English in science and technology

7

communities are international by nature. A useful preliminary here is a historical perspective. This inherently international nature of science and technology is reflected in the number of languages that have been the lingua francas of science and technology in history. The development has been somewhat different for instruction and publication. When we look at the academy, we see that different languages have been used in instruction and publication. With regard to the language of instruction, up to the sixteenth century, Latin was used. As commonly known, Latin was the first international language of the ‘learned’ and retained its position until the seventeenth century (Lindberg 1984). The attitude to Latin in the eighteenth century is clear from the quote that follows (Johnson 1706). Latin was the language of the learned, and replacing it with another language was seen as extremely unlikely, if at all possible:

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hat Johnson said wasthought virtuallywas impossible was abandoned What Johnson virtuallyhappened: impossibleLatin happened: Latin was aboutabandoned three hundred years ago. Historically, there have been other about three hundred years ago, leaving its place tolanguages other lanin theguages, position English has today. Not surprisingly, when English started and then to English, which is now “the universal language” for to the be used for education, it caused considerable controversy. If we look at the dissemination of knowledge. debates Historically, back in the there middle of been the seventeenth century, see three focal have other languages in thewe position English has points: whether to learn Latin or the vernacular, inEnglish this case English, today. For this reason, perhaps notLatin surprisingly, when started to be first, used whether to acknowledge the advantages of using whether for education, it caused and still seems to beEnglish causingand considerable to usecontroversy. Latin to make English a legitimate choice (Mitchell, 2006:475). LecTurning to the debates in the middle of the seventeenth centuturingry,inwethe forbidden in the Middle and seevernacular three focalwas points: WhetherintoEurope learn Latin or the Latin Ages vernacular, through the Renaissance (Nastansky, 2004:49). Some scholars advocated the use of the vernacular English, saying that it would be a very good skill to have in the future (Aickin, 1693:28). The aim in doing so, ironically, was to enable the public to have access to knowledge and obliterate the social division generated by Latin between the common public and the élite.

8

1 Introduction

in this case English, first, whether to acknowledge the advantages of using English and whether to use Latin to make English a legitimate choice (Mitchell 2006: 475). Lecturing in the vernacular was forbidden in Europe in the Middle Ages and through the Renaissance (Nastansky 2004: 49). There were scholars who advocated the use of the vernacular English, for the reason that it would be a useful skill to have in the future (Aickin 1967: 28). The aim in doing so was to enable the public to have access to knowledge and obliterate the social division generated by Latin between the common public and the élite. Subsequently, other languages did indeed come into use, gradually making room for even more languages. When it comes to the language of scientific research, up to the seventeenth century Latin dominated, slowly leaving space for other vernaculars. This, however, happened much more slowly than it did for instruction. Thereafter, both Latin and French were used. In the nineteenth century, French, English and German were all used for science and technology in international publication. The situation changed during the mid twentieth century, and the language of international publication became mainly English with the increasing importance of international publication in the twentieth and twenty-first centuries. These changes can be observed in the philosophical transactions of the Royal Society between the years 1665 and 1990 (Figure 1.2, produced using the information given in Allen et al.). 100 90 80 70 60 50 40 30 20 10 0

English

French

German

Latin

Other

Figure 1.2. The language of cited materials in the Royal Society 1665–1990.

1.2 English in science and technology

9

The Royal Society has been the centre of scientific activity in England since it was officially founded by twelve philosophers in 1660 to promote physico-mathematical experimental learning. Among its many activities are supporting scientists and engineers, influencing science policy, debating scientific issues etc. As I mentioned above, after World War II, most of the world’s scientific potential became localized in the US, mainly because of the fact that the country was intact unlike the countries in Europe, which were all badly damaged by the war (Graddol 1997: 8). World Wars I and II depended heavily on science and technology, which resulted in increased scientific activity during the war years. The US was able to preserve its scientific foundation and structure since none of the battlefields were located in the US, which in turn ensured its leadership in science and technology (Kaplan 2001: 11). In addition, more resources were available in the US, which made it a popular destination for a large number of students and scientists (Kaplan 2001: 10). Consequently, the US became the leading country in scientific and technical publishing. Later on the “design, production and dissemination” of scientific and technical knowledge was globalized; nevertheless, the US has managed to keep its place in the center (Truchot 2002: 10). However, there were additional reasons for the spread of English. It was reasonable to ask people to learn three languages when German was dominant in science, with English being important economically, and French being influential in cultural spheres (Shaw 2005, 2008). Japanese scientists did indeed learn three languages. With actual globalization however, which is what is happening today, there are Chinese, Japanese, Russian, Indian readers and more. With readers from a large number of countries, it became more difficult to insist that everybody should learn three languages. This being so, the extension of the international research community beyond Europe would have required a common language to operate in even if there had not been other external influences. Since World War II, many scientific journals have adopted English as their language instead of writing and publishing in their national language (Graddol 1997: 9). Over 90 per cent of the articles in the major SCI database are taken from journals that are published in English. European databases also have this as common practice instead of leaving room for articles written in European languages other than English (Truchot 2002: 10). Most journals have switched to English for purposes of broad readership. Already in 1997, Swales reported on the change of language from Swedish to

10 1 Introduction English-only in the last medical serial in Sweden, Läkartidningen (Swales 1997: 379). In short, as these examples show, for scientists today, publishing findings automatically suggests writing the articles in English since most journals, even in non-English speaking countries, require articles in English (Murray and Dingwall 2001: 86). French-authored articles in English are cited much more frequently than the articles in French: “The 1978 English-language papers received about 57,600 citations from 1978–1982, yielding a five-year impact rating—based on an SCI1 calculation — of 6.5 for the average paper. The French-language articles received 15,650, or a five-year impact of 1.9” (Garfield 1989: 1). One of the consequences of such practices is that publications in English are valued more than those in other languages. This is true for Dutch and Scandinavian languages; readers seem to value research articles that are in English more than those written in Dutch or Scandinavian languages (Ammon 2001). This reported trend certainly continues. In a recent study of bibliometric indicators for research evaluation in Italy, Gazzola discusses using bibliometrics as a performance indicator and states that doing so encourages Italian researchers to publish in English (Gazzola 2012). According to the three-year research evaluation VTR (Valutazione Triennale della Ricerca) carried out between 2001 and 2003, English dominates in hard sciences with 90 per cent and economics with 80 per cent. 2The numbers also show that the output for the VTR was in English in three fourths of all cases (Gazzola 2012: 141). Gazzola argues that using bibliometrics in this way functions as an implicit language policy tool (see section 6.3 in this monograph for a discussion of language policies). However, we should note that the situation described above is for publications aiming at an international readership. For national readerships, there still may not be much reason to write in English. This distinction is made clear in an investigation of a biliterate environment in a Danish business school. The study found that applied disciplines need to communicate nationally as well as internationally, which brings “a more complex mix of genres and languages” (Petersen and Shaw 2002: 372). In other words, if there is a strong reason to communicate scientific activity nationally, naturally the choice is the local language.

1. Science Citation Index of the Institute for Scientific Information in Philadelphia (SCI) is an example of the most influential databases on science and technology. 2. The figures show that English is used much less in political science (30 per cent), humanities (25 per cent) and law (10 per cent) (Gazzola 2012: 141).

1.2 English in science and technology 11

Subsequent to the establishment of English as the language of publication, practices in scientific activity followed the same trend: English became the main language to access scientific information, to take part in discussion, symposia and congresses and to collaborate with other scholars in the field. The use of English is promoted in academia and publications, networks, programs and institutions. The scientific programs in the EU, for example, operate entirely in English (Truchot 2002: 11). Certainly, there have been reactions to the use of English as the only international language. These reactions come from a wide range of countries. Ammon groups them as countries where the local languages have been lingua francas of science and technology (e.g. French and German) with the aim to achieve wide use of their L1s as international languages of science and technology (e.g. Japanese) where the L1s have had only local use and do not/cannot aim for international usage (e.g. Swedish) where the L1s have not yet been modernized or have been modernized to a very limited extent (e.g. Haus(s)a) (Ammon 2001: 348) Among these four groups, the first three are of relevance to the present monograph. The strongest reactions naturally come from the first group, including the countries that have used their L1s as the international languages of science and technology. A group of full-time researchers in France admitted to using English in a variety of situations, e.g. giving presentations in English at international conferences, but they expressed their concern regarding the use of English in laboratories and research centers (Truchot 2001: 321). Unlike other countries where international researchers speak English with each other in a lab situation, this group was in favor of preserving French as the language of operation. Another area where they aimed to keep French alive was doctoral work. They preferred postgraduate studies to take place entirely in French and theses to be written in French. This study, which took place in 1984, was followed by a plan to achieve language pluralism in France with the aim of performing high quality science and yet keep it available to the citizens. This did not reach great success however, perhaps not so strangely because 75 per cent of

12 1 Introduction doctoral students in France came from other countries at the time (Truchot 2001: 322). Another, and a stronger approach was the adoption of the law on the use of the French language in 1994 by the Parliament against the dominance of other languages, specifically against English. Although there were public protests, the law was maintained in the education domain. Researchers were given the right to give conference presentations in French, were obligated to provide a copy of the presentation document in French, and all other documents had to have a French abstract. This was, nevertheless, quite marginal in comparison with the conference work in English. It was also decided that French was to be the sole language of education, exams and theses. However, despite these concerted efforts, the debate seemed less alive in France already about a decade ago than it was in the 80s and 90s (Truchot 2001: 327). These efforts were criticized in a more recent work (Wright 2006). Wright lists the three misconceptions in the arguments used in the campaign to promote and protect French as: “(1) The intrinsic qualities of a language are factors in the promotion of a language as a lingua franca; (2) language policy making at national level can affect language practices in international contexts; and (3) language diversity is served by the promotion of another prestige lingua franca”. (Wright 2006: 35). (See Wright’s discussion for an elaboration of these misconceptions). German is another language that once was a lingua franca of science and technology. The situation of German today differs greatly from the times it had the lingua franca position. According to Ammon, after World War I, German was “banned from all international conferences”, and it never gained its previous position back (Ammon 2001: 345). German scholars have also had problems with publications in general when they wanted to publish in German internationally (Ammon 2001: 345). (See Darquennes and Nerde 2006 for the state of German as a lingua franca, from the past to present). There is work from countries outside the first group, as classified by Ammon above. In the second group, where the countries might aim for international use, Wu et al. reported from China, and Inoue from Japan (Inoue 2001; Wu et al. 2001), where the situation is rather different. These countries have never been colonized like some of the other countries in Asia, which might be the main reason for wider use of L1 in China and Japan. In China, English seems predominantly to be a language of writing and not used widely for spoken communication. In Japan, there is a long tradition of translating foreign technical terms into Japanese, which has allowed Japanese researchers to do science in their native language without

1.2 English in science and technology 13

relying on another language, strengthened by the availability of textbooks and other reference materials in their language. This situation is expected to change as higher education is becoming increasingly international; there is an increasing intake of foreign students into Japanese universities (Inoue 2001: 468). (See Kawai 2007 for a detailed description of the position English has in Japan today and implications of this for teaching). The third group in Ammon’s classification includes the countries where the L1s have always stayed at a national level, and among them is Sweden. Swedish is one of the languages that have never reached outside its borders for wide use within science and technology. Although it was used as the language of higher education from some time in the middle of the eighteenth century (Gunnarsson 2001: 293) until education became globalized about a decade ago, it was never a language that research could be published in internationally. Naturally, Sweden being a relatively small country and Swedish a small language, Swedish started to be challenged (see Chapter 3 for information on Swedish higher education). This is construed by some as leading to diglossia and Swedish being on its way to dying a natural death as an academic language, and measures are suggested if Swedish is to be resuscitated as an academic language (Gunnarsson 2001: 312–313). These views and worries, however, are not shared by others (Berg et al. 2001; Josephson 2004; Murray and Dingwall 2001: 106;). Already a decade ago, more researchers reported from Switzerland (Dürmüller 2001), Finland (Haarman and Holman 2001), Hungary (Medgyes and László 2001), Sweden and Switzerland compared (Murray and Dingwall, 2001), Russia (Kryuchkova 2001) and Belgium (Willemyns 2001) among others, with only slightly varying results. A decade or so later, English is increasingly being used in international publication, international conferences, general academic activity and higher education within science and technology.

1.3. Globalization and English in higher education If we go back to the list of the twelve major domains of English (section 1.1), place seven is taken by tertiary education (Graddol 1997: 8). The reason more and more countries are choosing English-medium higher education seems obvious: The scene described above and the globalization of studies.

14 1 Introduction A considerable number of changes have already taken place, specifically but not only, in Europe within tertiary education, and English is being used increasingly often. Student exchange programs within the EU result in changes especially at the Master’s level: A growing number of programs are offered in English to allow students to receive education in countries other than that of their origin. The development of additional programs in English is reported to be under way in several countries in continental Europe, allowing students from all over the world to participate. This expansion of use of the English language undoubtedly has advantages; student and staff exchanges are much easier, collaboration between universities is livelier than ever, and job opportunities are plenty. English is the language of science, academia and the professions. There is a growing trend of using English in general in European tertiary education (Cenoz 2006). Tertiary education in science and technology is, naturally, following this general trend. There is an additional reason for science and technology to adopt English as the medium of instruction in a large number of programs. As discussed in the previous sections, English is also the language of scientific publications and activity. Consequently, technical universities and institutes are responding to demands from students and industry by introducing English in tertiary education as the medium of instruction. Among the aims of the Bologna Process is “to make European Higher Education more compatible and comparable, more competitive and more attractive for Europeans and for students and scholars from other continents” (European commission), and the main aim of the Bologna process is stated as “to establish a common European area for higher education by 2010” (Swedish Agency for networks and Cooperation in Higher Education). Today, in 2012, we can see that this aim seems to have been reached to a great extent with student exchange programs. Creating a ‘common area’ requires a common language, and since English is the most widely studied language, it is the best-known second language. Consequently, the ‘common’ European area for higher education has evolved in such a way that English has become the ‘common language’. Although it is not stated anywhere that this common language has to be ‘English’, it is the natural choice. There are obvious advantages in making English the medium of instruction: Mobility, employability and competitiveness/attractiveness, which are all among the objectives of the Bologna Declaration (European Commission 2009).

1.3 Globalization and English in higher education 15

Universities today are more than ever advertising their multilingual programs and courses. They have several main reasons for doing so. First, they want to recruit more students. Secondly, it improves their public image and chances of competition in the education market. There are idealistic reasons as well, such as promoting multilingualism, creating world-citizens and strengthening internationalization locally (van Leeuwen and Wilkinson 2003: 11). Finally, there may be educational reasons like offering new degrees. Among these, it is likely that institutional survival, as Wilkinson calls it, plays a pivotal role. If the local market is too small, if the income that can be generated from international students will constitute a substantial income for institutions, learning and teaching through an additional language becomes an attractive option. Many universities worldwide are introducing programs where English is the medium of instruction. The number of students going abroad to study for one year or extended periods is increasing so dramatically that some countries have been reportedly considering making changes to their laws to stimulate this movement even further (Kruseman 2003: 7). Changing the medium of instruction from the local language to another language makes a number of changes necessary to the curriculum, the assessment and the general organization of education. One such change is the arrival of Content and Language Integrated Learning (CLIL). CLIL entails the teaching of content and language in connection with each other instead of as separate components (van Leeuwen 2003). Language learning, therefore, is not an add-on, rather a part of the teaching of content. In this approach, students acquire language and content together. Naturally, CLIL requires very close collaboration of content and language teachers. A wide range of benefits are likely to be gained by CLIL programs, namely, building intercultural knowledge and understanding, developing intercultural communication skills, improving language competence and oral communication skills, developing multilingual interests and attitudes, providing opportunities to study content through different perspectives, allowing learners more contact with the target language, not requiring extra teaching hours, complementing other subjects rather than compete with them, diversifying methods and forms of classroom practice and increasing learners’ motivation and confidence in both the language and the subject being taught” (European Commission 2012)

The impact of globalization on higher education institutions throughout the world has been much discussed (sections 1.3.1 and 1.3.2). Some see the globalization of higher education as creating a level playing field where

16 1 Introduction scholars and students from different countries are functioning together, sharing and disseminating knowledge, allowing higher education institutes to reach the aims mentioned above. On the other hand, there are those who see it as a basis of inequality (sections 1.3.2 and 1.3.3, e.g. Ljosland 2007; Phillipson 1998). Major higher institutions have a history of being highly dominant with regard to production and dissemination of knowledge, unlike smaller institutions with smaller budgets and fewer resources, which try to co-exist with them. Important academic institutions are, consequently, in richer countries and not in developing countries. Funding, libraries, access to major networks and qualified scholars are typically available in these larger institutions (Altbach 2004: 7). In this sense, globalization might be seen as automatically posing a disadvantage for developing countries (Altbach 2004: 7; Rodrik 1997 and 1999; Stiglitz 2002). At the same time, it should be noted that globalization does not necessarily create inequality. Countries like China, India or Brazil have benefited considerably from globalization financially, e.g. outsourcing, call centers etc. Having said that, apart from economic matters, there are also linguistic ones. Major higher education institutions and leading international journals operate in English-speaking countries, such as the US, the UK, Australia and Canada. The role of the language surely affects higher education policy and the work atmosphere of scholars and students. This kind of globalization is certainly not unprecedented in academia. Although it has been suggested that the emergence of a world language is a “wholly new phenomenon” (Coleman 2006: 1), this is not entirely true. Universities have always been international. Students have always travelled to prestigious institutions to get the best education. Similarly, scholars have always been able to work in foreign countries where there was expertise in their fields or have provided expertise where they have moved. As I have touched upon in the previous section (section 1.2), historically, we know that there have been other languages in the position English has today. The situation English is in now is unprecedented since this time it is on a global scale; however, communication and information technology are also on a global scale. In an age of such global communication and collaboration, higher education institutions worldwide cannot distance themselves from scientific collaboration. For collaboration and mobility to be possible, a common language is required. English is used internationally in academia for the same purpose: To bring scholars and students together so they can create, share and disseminate knowledge. Those who are critical of the

1.3 Globalization and English in higher education 17

growth of English claim that although English is supposed to bring people together, it does the opposite by creating gaps and inequality in today’s academia. Graddol said more than a decade ago that the higher education sector would become “increasingly complex” and refered to credit transfer, accreditation and hybrid courses such as “engineering through English”, which would result in new practices between institutions in the world (Graddol 1997: 45). This has certainly been happening in continental Europe and elsewhere. These practices and developments are likely to continue, and they will continue to enable even a greater number of students to study in other countries. As long as English remains as the leading language in technology and science, the situation is likely to put pressure on tertiary education to adjust itself. A distinction needs to be drawn at this point between the language of publication and the language of instruction. There have always been languages of publication, but only when a language becomes the vernacular and is used in instruction, do matters get complicated. English has been the lingua franca of science and education in publication for some time now (section 1.2). Concerns started to be raised only when it started being used in instruction. There are surely advantages of using English in instruction in higher education. However, English, being both the language of publication and now, increasingly, the language of instruction, has gained a much more powerful position. On this, there are two main concerns: If English is used in instruction instead of the local language, the local language might be threatened and if students cannot study in their native language but in English, they might not be able to learn as effectively as they would in their L1. These concerns have been expressed in numerous studies (section 1.3.2); a number of scholars from Europe have focused exclusively on whether this unprecedented growth of English is threatening the languages of Europe or not. I will cover these concerns in the following sections.

1.3.1. Ideological responses to globalization One of the words that is used very frequently in discussions of globalization is ‘dominance’. There are two ways of interpreting the term ‘dominance’. As Ammon points out, dominance could simply mean the widespread use and acceptance of a language with contrast to other languages

18 1 Introduction (Ammon 2001:v). It could also be perceived as the superiority of all the values and cultural attachments that can accompany a language. The latter interpretation of the term has been the center of some of the ideological responses to globalization. Central to our interest in the present monograph are reactions from scholars in continental Europe and Sweden, which will be covered in the next two sub-sections (sections 1.3.2 and 1.3.3).

1.3.2. Globalization and multilingualism in continental Europe Two main issues are raised in studies on the wide use of English instead of the local language. One of these has been the possible effects of the use of English on national languages and multilingualism, which has been discussed in a number of research projects as well as concerted efforts in the form of conferences. The second question these studies raise is whether the use of English has a negative effect on learning, and if so, whether it is avoidable. I will now consider these two issues. The main discussion in the first question has been threefold: Is English really a neutral lingua franca or a dominant language that takes its toll on multilingualism, or both? Whether or not English affects multilingualism negatively has been addressed amply in the literature. There are scholars who warn about the negative effects the widespread use of English might have on local languages and the diglossia it might cause. One of the leading scholars who have questioned the widespread use of English as a lingua franca is Phillipson. In numerous publications, he has argued that English is anything but a neutral lingua franca (Phillipson 2008). Behind the spread of English, he claims, is deliberate ‘Englishization’ and linguistic imperialism (Phillipson 1999). This he gets from Bourdieu, for whom Englishization is synonymous with “symbolic imperialism and linguistic hegemony” (Phillipson 2006b: 3). For Phillipson, preference for English as the medium of education in higher education solely because the textbooks are in English is a “recipe for diglossia” (Phillipson 1998: 274). His work draws attention to the linguistic diversity Europe has had in history and how English affects a number of areas, such as science, culture and education. He calls for linguistic equality for all citizens of the EU if the EU is to be a union that provides equal political and democratic opportunities for all. He sees the

1.3 Globalization and multilingualism in Europe 19

product of English as a lingua franca, or ‘lingua frankensteinia’ as he terms it, as a phenomenon that is linked to economic systems and the US, supported by certain ideologies among other things (Phillipson 2008: 261). In an earlier paper, he warns that English might be threatening the other languages in Europe (Phillipson 2006a: 1). Phillipson is also critical of the objectives of the Bologna Declaration, where languages do not come up once and that there is not one mention of bilingualism or multilingualism in the document (Phillipson 2006a: 16). He adds that ‘internationalization’ stands for ‘English-medium education’ in the Bologna process. This kind of perspective finds some support in research. Perhaps not so surprisingly, some of this support comes from Germany and Norway. German, as I discussed briefly in section 1.2, was once the holder of a strong position, and Norwegian, being a relatively small language, is not traditionally used in many domains. According to Mühleisen, the advantages of using English as the dominant language in academia and scientific activity are negligible compared to its serious consequences: Global diglossia and possible loss of languages together with the knowledge that is present in works in these languages. Two major concerns in this perspective are making knowledge available to all and continuing scientific activity in other languages (Mühleisen 2003: 117). The suggested solution is translation. Warnings about diglossia come from Norway as well. Some have argued that the dominant position of English is taking Norway to diglossia. Concerns were raised by Norwegian scholars, and consequently the Nordic Council of Ministers investigated the possible effects of English and whether a diglossic situation was under way for Norway (Tislevoll 2001). The report of this investigation was judged to be “inconclusive” in a more recent study (Ljosland 2007: 395). The fact that Norwegian is only used marginally in some domains in higher education has been seen as leading to ‘subtractive learning’ (Brock-Utne 2001: 221). The increase of borrowed words in Norwegian from English in academia, bureaucratic settings and technology, the sale of literature in English instead of Norwegian, recruiting staff who cannot speak Norwegian, the increasing number of Master’s courses in English and the prestige and financial benefits of publishing internationally in English have been reported as some of the factors leading to diglossia and eventually, possibly, to the loss of Norwegian in a large number of domains (Brock-Utne 2001: 121). It is argued that, if a language is abandoned, it will no longer be able to cater for the needs of its speakers in the domains where it has been abandoned (Brock-Utne 2001: 231).

20 1 Introduction More recent studies that investigated the use of English in Norwegian academia has been even more critical towards English and reported Norway to be on its way to a definite diglossia (Ljosland 2007, 2008, 2011). In these studies, Ljosland mentions doctoral work, and reports that English is seen as more prestigious than Norwegian in PhD thesis writing. Researchers in Norway see the advantages of writing their theses in English: Easier employability and establishing for themselves a position in their fields. The majority of the researchers that took part in the study had not even reflected on the choice of language since English was the natural choice. The results of the study show clearly that English is favored over Norwegian in scientific publications as the medium of instruction, and that it finds political support. English is the high-status (H) language in Norway today, leaving Norwegian as the low-status (L) language in some respects. Since PhD theses are high status publications, the preferred language for them is the H variety, which is English. Although there are still fields where writing a PhD thesis in Norwegian is acceptable, English dominates in the majority of fields. Academic communities are seen as different from other social uses of the English language in the sense that the former are easier to regulate by law. If it were necessary, the government could introduce laws to promote parallel usage of English and Norwegian in academia (Simonsen 2005: 249). With all being said, this is not fully realistic concerning the internationalization of higher education. Ljosland’s results are not unexpected, either, since PhD theses in Norway do not have a tradition of being written in Norwegian, especially not in natural sciences. It is clear from the reactions covered in the previous two paragraphs that in this kind of perspective, the focus is on resisting English, thereby Englishization as it is seen in the mentioned studies, and not necessarily on the needs of those that need a common language to operate in, e.g., scholarly activity (Mauranen 2005: 271). Some of these criticisms are made regardless of what kind of English might be the one used for international communication (Mauranen 2005: 271). One such recent work is Van Parijs’ work on linguistic justice (Van Parijs 2011). 3 Although Van Parijs is in favor of (English as) a lingua franca for the sake of egalitarian global justice, he argues that one lingua franca creates injustice, simply due to the fact that those who have the lingua franca, in this case English, get it for free unlike those who need to acquire it with

3. See Björkman 2012b for a review of Van Parijs 2011.

1.3 Globalization and multilingualism in Europe 21

much effort, time and money. In his monograph, Van Parijs offers a number of policies to combat this type of linguistic injustice, including a linguistic tax on Anglophone countries in order to subsidize the learning of it by nonnative speakers and the banning of dubbing altogether. Surely, the most provocative measure he proposes is the idea of imposing a linguistic tax on native speakers. He sees this necessary to achieve linguistic justice and to compensate those who have had to learn English as a foreign language. He further argues that non-Anglophone countries need to be protected by the linguistic territoriality principle (Van Parijs 2011: 133), which refers to native language groups in multilingual countries having the right to give their language the official language status in their territory in domains such as public administration, education etc. Van Parijs sees this principle central to linguistic justice. Various studies, however, have challenged the claims that English poses a serious threat to local languages. An early survey carried out on the potential language choice of the citizens of the EU states investigated the role of English and questioned whether it could be the lingua franca of the EU (Labrie and Quell 1997). The study was a large-scale European survey and contacted members of the younger and older generations. The results showed a general increase in the use of English and its dominance in general, but altogether, an English-only Europe was seen as very unlikely (Labrie and Quell 1997: 22). This did not automatically suggest that multilingualism would be on the increase; however, there was not enough evidence to say that the local languages were under threat. In a more recent study, the same assumption was challenged. House argued against the so-called language loss due to the omni-presence of English (House 2003a). She made a distinction between ‘languages for communication’ and ‘languages for identification’. House argued that those who speak English in lingua franca situations are very unlikely to regard it as a language of identification (see however section 5.2 in the present monograph for quotes that show that ELF can also be used by non-native speakers to signal their non-native speaker identity) (see section 6.3 for a brief discussion of ELF being a tool to express culture and identity). English does have a very strong position in the EU, but this is by no means a formal and official condition. According to House’s distinction of languages for identification and communication, English would not be a threat (House 2003a: 562). She supports this by presenting empirical research findings from three different research projects carried out in Hamburg. The results of these three studies suggest that English could not be threatening the local language: Native norms were

22 1 Introduction followed in translation, ELF interactions showed L1 presence, and L1 was included in English-medium instruction (House 2003a: 571). House suggests that instead of fighting the already powerful position of English or welcoming it without looking into its possible effects, we should accept hybridity and using English creatively (House 2003a: 574). Whether we welcome English without considering its effects or not, as long as the internationalization of higher education continues, and as long as English remains the most well-known second language, it is here to stay. More studies have argued that English does not pose a particular threat. In a study of the case of English in Switzerland, no such effects of English were found to be under way (Demont-Heinrich 2005). Switzerland is an interesting setting for such studies since it is a multilingual setting. Multilingualism in Switzerland has survived despite the increasingly strong position English has in the country. To a great extent, this is seen as a consequence of the protection and establishment of multilingualism as one of the basic components of the Swiss national identity by the government (Demont-Heinrich 2005: 66). This indicates that these speakers can speak English when needed and at the same time keep their multilingual national identity. Now let us turn to the second question I mentioned in the beginning of this section: Does the use of English have a negative effect on learning in the university-instruction domain, and if so, is this avoidable? Concerns are raised about domain loss and diglossia and the teaching and learning of content when students are taught in a foreign language. This is a question raised in a smaller number of studies compared to the studies on the first question. There has been relatively little work on this matter, some of which come from outside Europe (Gerber et al. 2005; Neville-Barton and Barton 2005), and some from the Netherlands (Klaassen 2001; Vinke 1995). The general findings of these studies point to negative effects of changing the medium of instruction from the local language to English (Gerber et al. 2005; Klaassen 2001; Neville-Barton and Barton 2005; Vinke 1995) with the exception of one study of longitudinal nature which found that the negative effects of learning in a foreign language actually disappeared within the same year (Klaassen 2001). This is certainly an important finding, as most higher education programs are spread through at least two years.

1.3 Globalization and multilingualism in Sweden 23

1.3.3. Globalization and multilingualism in Sweden Some of the studies on the effects of globalization focused on the situation in Sweden. The same type of questions have been raised in these studies: Whether English has a negative effect on the local language Swedish, leading to diglossia and domain loss, and whether the use of English affects the teaching and learning negatively. It is quite natural that voices are raised about the preservation of Swedish since Swedish is a small language with about nine million speakers only. On this, there are two views. According to the first view, Swedish is either on its way to being lost because of its increasingly limited use, losing to English. According to the second view, there might be some restricted loss of Swedish, but this is not likely to change the position Swedish has whereever it is used. Among those who questioned the dominance of English in Sweden is Hyltenstam. He argues in an early work that the dominance of English in higher status domains, e.g. academia, threatens Swedish and it might even lead to a social split in society where English is used in high-status domains and Swedish in low-status ones (Hyltenstam 1999). Hyltenstam fears that English might take over in Sweden in government affairs as a consequence of its position in the EU, leading to the death of Swedish. Skutnabb-Kangas and Phillipson are two other scholars who have similar concerns. They discuss the neutrality of English as a vehicular language and the equality of the other 11 official languages4 of the EU and refer to this neutrality as a myth (Skutnabb-Kangas and Phillipson 2000). In their perspective, similar to what Phillipson covered in other work (section 1.3.1.), English is by no means an equal language for people from all backgrounds. It may be opening doors to some, but it certainly closes them to many others (SkutnabbKangas and Phillipson 2000: 71). The term ‘language spread’, they argue, is not an accurate reflection of what is happening to English. To SkutnabbKangas and Phillipson, it is a misleading term since it suggests no doers of the action. Calling it a world language is equally misleading and is inappropriate practice, they say, since it ignores all the citizens of the world who do not speak English. An essential preliminary in any discussion of ‘language spread’ is Widdowson’s very useful dichotomy of ‘language spread’ and ‘language

4. At the time of their publication.

24 1 Introduction distribution’ (Widdowson 1997). Language spread is defined as “the virtual language being spread and in the process variously actualized” whereas “the distribution of the actual language implies adoption and conformity” (Widdowson 1997: 139). The two processes, Widdowson explains, differ greatly from each other with respect to their nature. Language spread welcomes adaptation (not conformity) whereas, as noted here, distributing a language suggests adopting a set norm and conforming to it. In this sense, the two processes cannot co-exist (Widdowson 1997: 139). What Skutnabb-Kangas and Phillipson are concerned about, in this respect, is language distribution and not language spread. In another study carried out after Hyltenstam and Skutnabb-Kangas, the two possible consequences of the dominance of English in Sweden are given as domain loss in Sweden and ‘genre death’ (Gunnarsson 2001). English seems to be used not only in academia but in the streets of Sweden as well. Hult reported that English is appropriated in daily conversation in the streets of Sweden, so it is used in public and interpersonal domains as well as high status domains (Hult 2003: 60, 2007). The assumption that English is a threat to Swedish does not find support from all. Among those who have investigated the use of English in Sweden are scholars who have analyzed the use of English in different domains. A study that investigated how and to what extent English is used in Sweden’s élite domains showed that English is not used exclusively; it is actually far from being fully dominant functionally, and it is not likely to affect language ideology in the élite domains it appears in (Berg et al. 2001). Therefore, diglossia is not the most appropriate term for what Sweden is going towards linguistically (Berg et al. 2001: 315). If we consider the influence of English in Swedish, we see three main areas: Borrowing words from English, the phenomenon of code-switching and domain loss (Josephson 2004: 7). The present monograph is not directly concerned with any of these areas although it looks at a domain said to have been lost. Among the three areas mentioned, domain loss is possibly the most serious one (Josephson 2004: 12), which makes a discussion of it useful here. Domain loss takes place when a language can no longer be used in a specific domain. It is important to note that although it is the language that does not get to be used in a certain domain, it is the speakers of that language that suffer from its serious consequences (Josephson 2004: 13). Therefore, it is suggested that it is not the efforts to “save and keep a language intact” that should take precedence but the social situation of the individuals who find themselves in complex situations where they cannot

1.3 Globalization and multilingualism in Sweden 25

use the domain language effectively. Josephson draws attention to those situations where speakers cannot fully perform because they cannot use their mother tongue in that domain (Josephson 2004: 14). In a way, this resembles the social division generated by Latin between common public and the élite (section 1.2). This social dimension needs to be handled with caution. The term domain loss, I believe, should be used with caution. There are domains Swedish has never had; such a domain is Information Technology. Since Swedish, or any language, cannot have lost a domain it has never had, those who are discussing domain loss (e.g. Gunnarsson 2011) must be concerned about the inability to extend Swedish to a certain domain. In this sense, domain loss is a misleading term. While warning us about domain loss, Josephson sees little reason to be concerned about the Swedish language in the foreseeable future. Among the reasons for this are the high literacy rate in the country, relatively large numbers of Swedish speakers in and outside Sweden as well as comprehensive descriptions of its rules for spelling, stable syntax and sentence formulations (Josephson 2004: 14). Josephson, referring to the five stages of language loss from a ‘domain’ point of view, says that Sweden is in stage 1 and possibly 2, but also that there is little reason to think that this will continue: Stage 1: English becomes the dominant language in a number of domains, e.g. in natural sciences. Stage 2: The position and dominance of English in some domains become stabilized. It becomes common knowledge that English is the language that is used in these domains instead of the mother tongue. Stage 3: English becomes the dominant language in many or all domains, which causes a large population of people to be bilingual. Stage 4: English with a Swedish influence becomes the norm and this becomes a variety. Stage 5: This variety divides itself to variations of it, which makes the Swedish English even more established as the standard language. Josephson says Swedish has slowly started to move into the territory of Stage 2; however, this information is from 2004. It is hard to gauge degrees

26 1 Introduction of such stabilization, but eight years later, especially looking at higher education, e.g. in engineering, we see that Swedish is in Stage 2.5 None of the studies I have mentioned in this section so far have been carried out recently, with the exception of Hult’s dissertation work from 2007 (Hult 2007). Why this is so is unclear, and to my knowledge, has not been discussed elsewhere. However, there are possible reasons. First of all, most concerns are about the use of English in élite domains. There is now empirical research that shows that using English does not necessarily threaten multilingualism or the local language in certain settings, e.g. higher education settings. Using English as a medium in a university program does not automatically mean there is no room for the native languages of the lecturers and students in that setting. Söderlundh, in her ethnographic study of English-medium programs in Sweden, reports that lecturers and students do use English as a de facto lingua franca in their interactions, but they use other languages as well, depending on their needs (Söderlundh 2012: 87). In fact, this is her main finding. Her observations show that participants choose the language depending on their local or social needs, choosing the “path of least linguistic resistance”, in turn displaying substantial variation in their use of languages (Söderlundh 2012: 105). She, in a way, shows how global polices are adapted locally by the participants in that setting, which she highlights in the title of her study6. Söderlundh’s paper is from a journal special issue that has as its topic the language choice in the international university.7 Other papers in the same special issue (e.g. Risager 2012), either explicitly or implicitly, show the multilingual nature of English-medium universities. While the editors Haberland and Mortensen are critical towards “an unquestioning acceptance of English as the ‘natural choice’ of language in a world with increased transnational contacts”, they end their introduction to this special issue by saying

5. Bolton and Kuteeva’s survey results from Stockholm University show clearly that the use of English in the Sciences “is a pragmatic reality for both teachers and students alike” (Bolton and Kuteeva 2012: 444). 6. Global policies and local norms: Sociolinguistic awareness and language choice at an international university. See References for further details. 7. This special issue is an outcome of the 2008 conference on transnational student mobility, organized by the CALPIU (Cultural and Linguistic Practices in the International University) research center. The conference led way to two other publications: An edited volume (Preisler et al. 2011) and a journal special issue (Björkman 2011).

1.3 Globalization and multilingualism in Sweden 27

that “the situation in academia in the present “cannot be described as English swamping all the other languages, marginalizing them or making them redundant” (Haberland and Mortensen 2012: 5). In this sense, it does not seem like there is much reason to be concerned when it comes to the linguistic diversity of international university settings, as long as there is a wide spectrum of languages in play. The reason for fewer studies on the perceived threat of English could also be that those concerned have shifted their focus to language policy documents, some of which are rather ‘L1 protective’. Language policy issues are not central to the present monograph; however, they are highly relevant in that the results of research on ELF from university settings should be factored in the design and implementation of language policies. I will discuss these issues in Chapter 6. Turning to the second question, whether the use of English as the medium of instruction affects learning negatively, has been discussed but researched to a rather limited extent. If we look slightly outside the Swedish frame at the studies carried out in Scandinavia, we see some work concerning higher education in the form of surveys which deal with the extent to which a second language is used in educational situations or have focused on the language learning effects of such teaching (Hellekjær and Westergaard 2003; Tella, Räsänen and Vähäpassi 1999; Wilson 2002). When it comes to Sweden, there has been some work on how students learn in English as opposed to learning in their mother tongue Swedish. In Airey’s thesis on the teaching and learning of physics in Swedish and English, the subjects reported no significant difficulty with English as the medium of instruction. Nor did they think that the choice of language was important. Despite what they said, Airey’s investigation showed that they were less able to explain disciplinary concepts in English, to interact in lectures as they asked and answered fewer questions in lectures given in English, or to take notes and follow the lecture simultaneously (Airey 2009). Apart from Airey’s study, there are two smaller studies on the effects of language on learning (Karlgren and Hansen 2003; Söderlundh 2004). Both these studies have looked into written work. Their conclusions are very similar: Students are greatly disadvantaged by having to read in another language. Söderlundh showed that students who received tuition in Swedish reached a deeper understanding of the content compared to the ones who were subjected to English (Söderlundh 2005).

28 1 Introduction 1.3.4. The answer: English as a lingua franca As I mentioned in the introductory notes to this chapter, there is another perspective one could adopt with regard to the widespread use of English in Europe today. Using English in a number of international domains does not have to mean Englishization or Americanization nor does it have to be linked to linguistic imperialism, provided that efforts are made in validating and legitimizing its use by its non-native speakers. This brings the need to describe its use by its non-native speakers from different L1 backgrounds in international settings. This area is defined as English as a Lingua Franca (ELF), when English is used by those who need it as a vehicular language to communicate with each other. ELF usage must be investigated for a number of reasons, the first of which is theoretical (Mauranen 2005: 273). ELF has the same purpose with pidgins and creoles (see the beginning of this chapter), which is the important purpose of enabling communication for speakers of different L1 backgrounds. In this sense, for this theoretical reason, it deserves to be studied just like pidgins and creoles. We have already seen from the ELF research available so far that the results and benefits of ELF research fall nothing short of the interesting results of research on pidgins and creoles (Chapter 2). ELF speakers come from highly diverse backgrounds, and the number of speakers who use English as a lingua franca will only continue to increase. This creates a complex linguistic scene, which is worthy of detailed descriptions. ELF does have the same purpose with pidgin and creoles; at the same time, there are important differences between the two kinds of contact languages (see section 1.1). ELF differs also from the use of English as a native language (ENL). In this sense, it needs to be studied for linguistic descriptions to be available (Mauranen 2005: 273) and to be able to observe these differences. We need to see the kind of variations from ENL usage, adaptations to it and creativity in ELF usage. With these variations, adaptations and creative usage, ELF has already challenged dated norms, such as native speaker as the ideal target. ELF settings are by nature heterogeneous. Accordingly, in today’s heterogeneous world, we need heterogeneous linguistic models (Canagarajah 2006: 211). With linguistic descriptions of ELF available, it already has been to some extent, and will be even more so in the future, possible to understand how communicative effectiveness can be achieved and what the norm should be for ELF settings.

1.3 Globalization and multilingualism in Sweden 29

With more information available, it will be possible to equip those who need to operate in ELF settings with the skills they need to achieve communicative effectiveness. In academic settings, which is the focus in the present monograph, there is a great need for descriptions of the type of communication that takes place in international university settings, to be able to provide the support needed for students and lecturers. Academia lends itself nicely to such detailed linguistic investigations, as it is and has always been international by nature (Mauranen 2012:1). Most academic settings are of a high-stakes nature where individuals from different L1 backgrounds need a well-functioning common language to meet the demands of academia. Most academics do research, solve problems and devise solutions, present their research in international conferences, publish their findings in international journals and network in a number of ways with other researchers, and teach. Students in higher education settings attend lectures, work in groups with other students on tasks that have been assigned to them, and they present their work in spoken or written form. The demands on academics and students are high when performing these acts and interacting with each other while doing so, and all these acts can be performed by means of a common language that functions as a lingua franca among these individuals. Such dynamic use of the language for such important purposes further make academic settings ideal for linguistic investigations. Any university with aspirations to become truly international will need to understand the dynamics of international communication, in this case, communication among people of different L1 backgrounds. With proper descriptions of ELF, it will be possible to do this in a number of ways. One way is EAP/ESP instruction for students and teachers. ELF research will provide information on what to prioritize, and deprioritize, in the EAP/ESP classroom. Another way is through better decisions made at higher levels, presupposing that there is awareness of the importance of language issues and interest in bettering the every day lives of lecturers, students and other personnel who need to operate through the medium of English. Decision makers at higher levels will be provided with more information regarding the type of communication that takes place in international university settings, resulting in better decisions, provided that this information is communicated to them. Such important decisions mainly deal with providing support for staff and students, setting university English language requirements for admission, and producing language policy documents.

30 1 Introduction With this monograph, I provide a detailed description of ELF usage in an academic setting and explore the ways in which speakers achieve communicative effectiveness, thus adding to existing knowledge on ELF.

Chapter 2 Previous research on ELF

In the preceding chapter, I gave a brief review of the journey English has had in becoming the language of science and technology after Latin and other languages. I also discussed the notion of globalization briefly along with ideological reactions to globalization in the world and in Sweden, the setting of the present monograph. I turned, at the end of the chapter, to the English as a lingua franca (ELF) perspective and argued that globalization does not necessarily mean Englishization or Americanization. Clearly, when saying so, I do not mean to deny the colonial past or capitalist present that English has (see Seidlhofer 2011: xi). When speakers of different L1 backgrounds communicate through ELF however, they do not seem to be under much influence of its native cultures; they communicate without “becoming anglicized–if anything, they are ‘de-anglicizing’ their English’” (Seidlhofer 2011: xi). I hope to show this clearly with the several accounts of ELF interactions I present in Chapters 4 and 5. Most ELF speakers in high-stakes situations are focused on the result rather than the process; they simply want to get the job done. The use of English by its non-native speakers must be legitimized and validated, and ELF usage needs to be described thoroughly for several reasons I mentioned in the previous chapter. In this chapter, I will provide an overview of some of the most significant ELF studies from recent years. My aim here is not to provide an exhaustive overview including all ELF studies to date from all levels of language, which would be close to impossible because of the growing output of publications on ELF, especially in the past decade. Instead, I will focus on the dimensions, modes, settings and domains that are of immediate relevance to the present monograph. Since the appearance of the first ELF studies (Jenkins 2000; Mauranen 2003; Seidlhofer 2001), the proliferation of investigations that have been carried out on different aspects of it has been dramatic. The increasing use of English as a lingua franca in a wide range of settings opens up possibilities for carrying out studies of a varying nature. Documenting these different types of studies is, therefore, no easy matter. The literature could be reviewed by taking any of these aspects as organizational criteria: Dimension (form, pragmatics etc.), mode (speaking, writing etc.), countries (dif-

32

2 Previous research on ELF

ferent L1s), domains (business, technology etc.), levels of proficiency, media of communication (phone conversations, e-mails etc.), purposes, topic type or method of data collection/analysis. The work that has been done in the field so far has focused generally on more than one of these aspects; therefore, it is hard to draw sharp distinctions. The review I will provide here will thus be organized with reference to the normative elements. After a brief discussion on what normative elements refer to, our focus will first be on work without normative elements, i.e. studies on pragmatics, and then we will shift to works including normative elements, i.e. studies on form. Later on, I will discuss studies that have combined pragmatics and form, followed by attitude and irritation studies.

2.1. The relevance of normative elements in ELF research Those who speak English as a lingua franca throughout the world are in a constant state of flux, the number of first languages is enormous, and there is frequent contact in English among speakers of other L1s. For optimal communicative effectiveness, different ELF situations need to be described, outlining general tendencies and commonalities of usage, providing information for those who need to function in such settings on a daily basis. We do know that speakers from the outer circle speak their own variety of English (Kachru 1982), but the ones that are in the expanding circle display a large variety of usage, and the way these speakers use English too needs to be investigated in its own right. Such an investigation within speech can be carried out either contrastively or descriptively. If we consider the different dimensions mentioned in the previous section, it is quite natural that work on the morphosyntactic dimension needs to be done contrastively. Most studies investigating different elements within morphosyntax have taken the native speaker production as the yardstick, which is not a waterproof approach. The native speaker yardstick does not always lend itself to reliable comparisons, because there are forms used by native speakers that are considered non-standard according to prescriptive grammars. Examples of such forms are Heads and Tails a.k.a. Pre- and Post-dislocation (section 4.1). These forms are perfectly common forms within speech but are not parts of language instruction since they are not standard forms in written discourse (Mauranen 2007). So in cases where the yardstick is not the native speaker, simply because the non-native speaker group is extremely heterogeneous or because there is no

2.1 The relevance of normative elements 33

comparable corpus, comparisons have been made with the written standard (Carter and McCarthy 2006: 164; Simpson et al. 2002). Applying the written standard to speech is undoubtedly highly questionable (section 5.2), but it seems to be the practice when no other standard is available. Another question regarding norms is how unitary ELF usage is. Research results from different settings document commonalities, but differences are observed in different dimensions. It is unlikely that an informal conversation at a train station somewhere in Europe (James 2000) would be similar to what speakers would do in an academic setting in Scandinavia, such as the current setting here. Therefore, the requirements for successful communication should be rethought with reference to different settings, levels of proficiency, activities etc.

2.2. Work without normative elements: Pragmatics Most of the research into ELF has been descriptive. A part of this descriptive work has been carried out on the level of pragmatics. Some of the earlier works on ELF actually were within pragmatics, arising from the need to understand how non-native speakers of English communicate with each other. These descriptive studies were carried out mostly without normative or comparative elements. However, some studies have made references to how native speakers communicate in similar situations. It is important to point out that these studies included comparative native corpora to observe what native speakers do in similar situations and in principle not because what native speakers do should be the norm for ELF speakers to adhere to. In this section, we will examine ELF research within pragmatics from early days to present day. The first studies that focused on ELF pragmatics were carried out differently compared to the more recent empirical ELF studies (see a discussion on this in Jenkins, Cogo, and Dewey 2011). Among the earliest of these pragmatic studies were by Firth (Firth 1990, 1996) and House (1999). In Firth’s study, Danish export managers’ and their clients’ business telephone conversations were investigated through conversation analysis. The conversations are strictly work-related, focusing on the buying and selling of food and micro-electronics. Firth’s general findings showed that these business lingua franca conversations were perfectly normal, meaningful and ordinary. The speakers in the audio-recorded data have two main aims: To sell the goods through successful talk and meanwhile make the talk

34

2 Previous research on ELF

normal and ordinary despite the occurrence of grammatical infelicities such as unidiomatic clause constructions along with prosodic and pronunciation variants (Firth 1996: 242). It is of particular interest that a misunderstanding occurs in one of the conversations between two speakers and that neither of these speakers seems to reveal any awareness of it, possibly because their main aim is the selling and buying of the goods. The main difference between this study and empirical ELF work today is the approach to ELF usage: Firth investigated NNS communication and demonstrated that despite the non-standard usage by low-proficiency export managers and their clients and despite their non-native speaker status, they were able to communicate with each other effectively. Firth’s take on ELF is thus the use of English by low-proficiency speakers and how they use English successfully as a lingua franca despite their non-native usage, taking native usage as the ideal target (see Jenkins 2011: 928; Jenkins, Cogo and Dewey 2011: 286 for criticisms of this take on ELF). This is quite a difference from more recent empirical ELF studies, where the native/non-native dichotomy has been shown to be an unreliable yardstick (e.g. Björkman 2011). Firth’s study showed, however, a number of strategies which the speakers in this ELF business situation made use of when they communicated with each other. The first of these is termed the Let-it-pass strategy, where speakers let an unclear word or utterance pass. They choose to avoid a potentially problematic situation and prioritize building a common ground before they ask for any clarification. The second strategy is the Make it normal strategy. Such a case can be identified by the hearer treating the speaker’s non-standard usage as normal. The hearer focuses on the content instead and produces reformulations of the other’s opaque usage. There were very few cases of what is known as other repair, which would mean focusing on the form and not the message the speaker wants to convey in the conversation. On the basis of the observations from the material, it is suggested that these interactions have a quality of interactional robustness, i.e. this kind of interaction seems to be robust and can withstand nonstandard production. The interactants do “interactional work” as Firth terms it and focus on communicative effectiveness (Firth 1996: 256). A study that was carried out after Firth’s first work (1990) also analyzed business phone conversations between speakers of different European L1s (Gramkow Andresen 1993)4. The findings reported similar cooperativeness between speakers, focusing on the message and not the form. The speakers in ELF conversations seemed to adopt their own style.

2.2 Work without normative elements 35

Firth’s work was followed by House’s study on misunderstandings in intercultural communication (1999). House focused on a classroom setting to study ELF communication; however, the setting was a simulated one set up for research purposes and did not lend itself to authentic communication. The participants did not really feel the need to communicate most likely because of the nature of the setting where there was nothing at stake. They did not engage in serious conversation or use the strategies that we now know are used frequently in spoken ELF communication to achieve effectiveness (see e.g. Björkman 2011; Kaur 2009; Mauranen 2006b). House suggested the concept of pragmatic fluency in this study and stated that pragmatic fluency did not require conforming to ENL norms and that ELF speakers can be pragmatically fluent in their own ways without following ENL patterns. The five different performance criteria for pragmatic fluency were suggested to be (House 1999: 151): 1. Appropriate use of routine pragmatic phenomena 2. Ability to initiate topics and topic change, making use of appropriate routines 3. Ability to “carry weight” in a conversation 4. Ability to show turn-taking, replying/responding 5. Appropriate rate of speech, types of filled and unfilled pauses, frequency and function of repairs Another study that made reference to native usage was Meierkord’s (2000). Meierkord analyzed small talk conversations among non-native speakers of English who all used English as a lingua franca (Meierkord 2000). The corpus comprised conversations of a total of 13.5 hours in a student hall of residence in Great Britain, with speakers from 17 different first language backgrounds. The findings show differences on the level of pragmatics from the native speaker varieties BrE and AmE with reference to discourse structure and politeness phenomena. This work is not comparative, i.e. the investigation is not of two comparable corpora, but references were made to general native-speaker usage. Lingua franca speakers did not seem to link their opening phrases to the “core phase” of the conversations. Pauses instead were rather frequent, especially before the conversations end, to make the transition clear (Meierkord 2000: 7). The participants’ preference for safe topics is also striking, e.g. life in the hostel, meals, jobs, or their university classes. Most topics were changed after less than 10 turns. Another interesting finding was the occurrence of overlap,

36

2 Previous research on ELF

which varied greatly according to the speakers. One aspect of the participants’ speech that was reminiscent of that of native speakers was backchanneling behavior, albeit with the addition of supportive laughter. The specific characteristics of ELF conversations in this setting then can be summarized as the lack of linking closing and opening phases of conversations to the main part of the conversation, preference for safe topics, the presence of long pauses within and in-between turns, the use of politeness phenomena and backchanneling similar to NS usage. Meierkord concludes by saying that the linguistic behavior of the participants in her study seems to be shaped by two main principles, namely participants wishing to save face and participants supporting each other with supportive discursive behavior in the form of backchanneling, supportive laughter and excessive use of cajolers (Meierkord 2000: 10). This appears to indicate that ELF small-talk communication can be characterized by cooperativeness rather than by ineffectiveness or misunderstandings, which Firth’s, House’s and Meierkord’s studies have in common. Further investigation of pragmatic fluency investigated along with the ‘habitat’ factor (Bourdieu 1991) in ELF interactions and cultural identity show that where the interaction takes place is decisive on the significance of culture for communication (Pölzl and Seidlhofer 2006). Pölzl and Seidlhofer define habitat as “the setting which interlocutors recognize as their own” (Pölzl and Seidlhofer 2006: 155). They investigated the setting with specific reference to House’s first and fourth performance criteria mentioned earlier here. The setting investigated was the Hashemite Kingdom of Jordan, and the participants were Arabic speakers at the Department of Modern Languages, interacting in their own habitat in the form of spontaneous talk. The study is somewhat atypical as regards its subjects and setting: Arabic speakers speaking English in Jordan. The subjects transferred their native language communicative norms to their ELF-like interaction. They displayed a high rate of speech and frequent overlap and codeswitching, which was not perceived as irritating or inappropriate by the rest of the participants. The overall conclusion is that although ELF is the use of English by speakers of different L1s, local pragmatic norms apply in interactive situations where speakers are mainly from one culture. In the case of this study, all the speakers were from the same culture, and the interactions took place in their homeland. Pölzl and Seidlhofer refer to this situation and suggest that the speaker can make himself/herself more at home in the foreign language if the setting is the home territory. They point to the habitat factor as an important socio-psychological effect of a speech

2.2 Work without normative elements 37

event (Pölzl and Seidlhofer 2006: 172). Their findings are in accordance with earlier findings mentioned here: Speakers in ELF situations do not need to conform to ENL culture, and due to the heterogeneous nature of these situations, they generally cannot rely on their own culture either, but they do if they can (Pölzl and Seidlhofer 2006: 173). From this type of study, one could perhaps extrapolate that the interactional patterns adopted by mainly or only Europeans might be difficult for other cultures. It seems like speakers keep their cultural identities by the type of linguistic behavior described above. More recent studies that have followed report similar results (Cogo 2010; Klimpfinger 2009). Especially code-switching as a strategy has been reported to be a way of signaling cultural identity (see Jenkins, Cogo, and Dewey 2011). In addition to these cultural studies, some recent work has focused on how speakers from different backgrounds communicate using ELF (Kaur 2011; Pullin Stark 2009). These studies have focused on the speakers’ turn-taking behavior (Kaur 2011) and observed supportive behavior, willingness to achieve common ground and interactional support that enables effective communication (Kaur 2011; Pullin Stark 2009). Another strategy mentioned among those employed by ELF speakers is repetition (Cogo 2009; Lichtkoppler 2007; Mauranen 2006b). Mauranen mentions repetition along with clarification and self-repair as proactive strategies that ELF speakers use. Lichtkoppler identifies three types of repetition with reference to the scale of fixity: Exact repetition (a.k.a. verbatim or full repetition), repetition with variation (a.k.a. non-exact repetition or partial repetition) and paraphrasing (Lichtkoppler 2007: 43–44). In this investigation of conversations recorded at the office of a student exchange organization, a number of types of repetition were observed and grouped according to their functions, i.e. repetition for time-gaining, utterance developing, repetition for prominence, repetition that ensures accuracy, repetition that signals listenership and establishes cohesion. All these types of repetition are reported to have three main functions, namely helping the participants with language production, supporting mutual understanding and enabling participants to show their attitudes and opinion (Lichtkoppler 2007: 59). Similarly, Kaur shows how the university students in her Malaysian corpus employ repetition (along with paraphrasing) as a frequently used strategy to communicate effectively, especially after long silences, short responses and overlaps (2009). Björkman’s university corpus reveals similar results (2011); engineering students use repetition often to make themselves understood and to ensure that the message has been received. It

38

2 Previous research on ELF

can be suggested on the basis of these findings then that repetition is a strategy ELF speakers employ frequently to make successful conversations despite differences in culture and language backgrounds. A plethora of reports on the communicative effectiveness of ELF speakers have been produced over the last fifteen years. An investigation of international students’ interactions in simulated meeting situations showed that the students in these meetings were able to communicate in a pattern which at first seemed chaotic but then proved to be quite systematic (Lesznyák 2002). The students seemed to develop a dynamic topic management skill, which enabled them, without much trouble, to find common ground and make decisions. A critical question to address is obviously the issue of understanding and the resolution of non-understanding in ELF settings. One might expect that interaction in a language which is non-native for all participants would be ineffective or at least inefficient. However, the findings from the literature reveal otherwise. A study providing examples from a business setting indicated that ELF speech in business settings seemed to be characterized by interactional and pragmatic effectiveness (Pitzl 2005). Pitzl defines understanding as “not a passive ability, but an interactive and jointly constructed process which is dynamic and cooperative and which all participants of a conversation continuously engage in” (Pitzl 2005: 52). So understanding is constructed by the joint efforts of the speaker and the listener. As much as this is true for any type of spoken interaction, it has been reported as a typical characteristic of ELF interaction (Firth 1996; Meierkord 1996, 2002; Seidlhofer 2001). In this study, the speakers had a relatively successful way of asking for clarification when non-understanding occurred and they did so without interrupting the communication. The ones who provided clarification also seemed to be able communicators and gave precisely the right kind of information needed for the interaction to continue (Pitzl 2005: 14). The communicative effectiveness of speakers in ELF situations finds support in research based on ELFA data (www.eng.helsinki.fi/elfa). Misunderstandings overall were not as common as one would perhaps expect them to be in ELF situations (Mauranen 2006b). As mentioned briefly earlier, what appeared to be characteristic of ELF interactions was the effort put into preventing misunderstanding through self-repairs, clarifications and repetitions. However, when misunderstandings did occur, the clearest signals that indicated them were direct questions, repetitions of problematic items and other, more indirect signals of misunderstanding, i.e. minimal

2.2 Work without normative elements 39

incomprehension signal “mhm?” repetition of a single item and some questions. Misunderstandings were resolved by the frequent occurrence of selfrepairs, co-construction, repetition and clarifications, all of which are linguistic signs of the effort ELF speakers put into the situation (Mauranen 2006b: 146). It is interesting to note that there were no cases of interactive grammatical correction; however, there were self-repairs that were reformulations of grammar. This seemed to be a difference between ELFA data and the data from the MICASE corpus including mostly native-speaker speech that some of the findings were compared to (Simpson et al., 2002). The MICASE data included no syntactic reformulations but rather paraphrases of longer statements. Based on these findings from ELFA data in general, it is maintained that ELF speakers spend considerable effort to ensure mutual understanding (Mauranen 2006b: 146). Data from the ELFA corpus demonstrate further that both economy and creativity are qualities of lingua franca English along with explicitness through (self)-rephrasing and topic negotiation (Mauranen 2007). When the results were compared with MICASE, it became clear that rephrasing markers occurred many times more often than in NS speech. ELF speakers used fewer rephrasing markers, e.g. I mean, but they used them much more often than the speakers in MICASE data (Mauranen 2007: 49). Another difference lay in the choice of expression for rephrasing. The ELF speakers’ choice was different from that of the MICASE speakers, and this was not the most written-discourse-like one as might be expected but a very everyday one. It is interesting to note that most of this rephrasing deals with form and not meaning. When an utterance is rephrased, the chances that the utterance will be understood increase, and ELF speakers use rephrasing in order to achieve clarity and comprehensibility (Mauranen 2007: 257). According to Mauranen, another way to increase explicitness was topic negotiation, usually arising from the need to achieve successful interaction. Finally, discourse reflexivity, or metadiscourse (discourse about discourse), helps speakers achieve organization within their utterances as well as helping the speakers to achieve clarity and explicitness. Mauranen’s findings show that discourse reflexivity occurs together with hedging expressions in ELFA data, which corresponds to what happens in native speaker conversation in MICASE data (Mauranen 2007: 258). All in all, in the studies we have looked at so far here, we see that speakers in ELF settings use a variety of strategies to communicate effectively. They have been reported to do so even through phonological tools. A study on experimental ELF interactions has shown that participants pay

40

2 Previous research on ELF

attention to pitch cues both as a signal of potential “trouble source” and as a means to indicate that negotiation or repair has taken place (Pickering 2009: 235). The study, revisited recently with naturally-occurring ELF data (Pickering and Litzenberg 2011), provides further support that speakers in ELF interactions use intonational signals as a resource to communicate effectively in interaction. Although these two studies did not have questions as their focal point, they provide evidence that intonation can be an important resource in ELF interactions. Some studies in ELF pragmatics differ from the ones mentioned above in terms of their results. A small number of studies reported on elements of ineffectiveness in communication. A study reported from nine simulated sales negotiations performed by ten international business students and eighteen professional negotiators from different countries, and investigated interactional safe talk and personal pronouns (Planken 2005). Interactional safe talk is known as an important element of creating and maintaining rapport (Aston 1993: 226). The student negotiators had much fewer instances of initiated safe talk unlike the professional negotiators who initiated safe talk frequently. Moreover, the professional negotiators initiated safe talk in all three stages of the negotiation, i.e. opening, bargaining and closing whereas the student negotiators did so only in the opening and closing stages if at all (Planken 2005: 389). Another difference between the two groups of negotiators has to do with the highly personalized style the student negotiators adopted (Planken 2005: 399). The professional negotiators had occasional instances of personalized style in the opening and closing stages of the negotiations; they otherwise maintained their professional style. However, the data in this study came from simulated business negotiations performed by students, and it is therefore not surprising that students who are engaged in an activity had a more personal style unlike the professionals’ in similar situations. Because the data came from a simulated situation that was set up for research purposes and not from an authentic communication situation, it is best to interpret the results accordingly. In addition, these results tell us something about the speakers’ negotiation skills and not necessarily about ELF usage. Another setting where communicative ineffectiveness was observed was internal communication from two companies, Swedish and Finnish, two corporate mergers with Swedish and Finnish as their native languages and English as the lingua franca for communication with each other (Salminen, Charles, and Kankaanranta 2005). The initial aim of this study was to explore how the employees were coping with cultural and linguistic challeng-

2.2 Work without normative elements 41

es. The employees’ daily discourse was also investigated in meetings and e-mail messages. Terming the discourse BELF, Business English Lingua Franca, Salminen et al. point to its complex nature. It was suggested by some earlier research that ELF is or can be cultureless (e.g. House 1999), and the ‘culture irrelevance hypothesis’ by House suggested that national and native language and culture adherence are eclipsed in ELF interaction (House 1999). The approach adopted in Salminen, Charles and Kankaanranta’s study is, however, that speakers in BELF settings do have a cultural background they bring to the communicative situation, as also discussed by Meierkord (Meierkord 2002). A survey was conducted in both companies with reference to the communicative environment, followed by interviews. To investigate whether the phenomena reported by the questionnaire results and the interviews were apparent in their discourse, four video-recorded meetings of a total of nine hours were studied. The BELF in this study helped communication in some cases because it has no native or non-native speakers, or learners; however, it did not ensure communicative effectiveness since there were other elements at play. ELF was not found to be free from culture or ownership as suggested by some in ELF literature (House 1999), and differences of cultural perceptions were observed in the discourse. This is similar to the arguments made by Pölzl and Seidlhofer, that culture does play a role, but its role depends on the setting in which the interaction actually takes place (Pölzl and Seidlhofer 2006). Some of the differences mentioned by the participants in the questionnaire and interviews were observed particularly in meetings. E-mail correspondence, however, displayed similarities across the two language/culture groups rather than differences (Salminen, Charles, and Kankaanranta 2005: 418). Another study on negotiations, carried out at a German university setting where English is used as a lingua franca, reports similar results (Knapp 2011). On the surface, participants appear to be effective communicators of the subject matter through ELF, but as soon as a range of speech acts are involved, complications may arise. Negotiations and resolving issues resulting from ineffective negotiations constitute examples of such speech acts. The contradictory results within the pragmatics of ELF settings might be related to the nature of the settings in which these studies were carried out. In addition to all the studies mentioned above, recent journal special issues have been dedicated to ELF pragmatics (House 2009; Björkman 2011). Two interesting articles from House’s special issue deal with the discourse marker you know (House 2009) and chunking in ELF (Mauranen

42

2 Previous research on ELF

2009). House shows that ELF speakers use the discourse marker you know differently from native speakers: Instead of using it to relate to other speakers in conversation, they use it to highlight a certain topic, to introduce a new topic and to announce what they are going to move on to next. In other words, they use it as a discourse-organizing tool. Mauranen’s work in the special issue focuses on the notion of chunking, creating phraseological patterns used to manage interaction. The results show that speakers in the ELF setting described approximate their usage of chunks to conventional forms, but they also use these chunks, especially the longer variable units, in unconventional forms. These unconventional forms however are never so variable that comprehensibility cannot be ensured. In addition, the regularity in the usage of these unconventional forms is interesting, showing that the speakers do not use these non-standard chunks randomly. The second journal special issue (2011) on ELF pragmatics differs from the first one in terms of its special focus: It focuses only on ELF in the international university. Compared to social settings, in instructional discourse, the demands on communicators are considerably higher. The highstakes nature of instructional settings is a key issue here. In particular, using another language than one’s own for high-stakes communication requires heavy investment in the communication process. Any lack of communication or steering away from the topic would lead to the incompletion of the task, and this would adversely affect the performance of the students and/or the teachers involved. Therefore, the aim in real high-stakes interaction is to communicate in a practical and functional fashion and achieve the desired outcome. In such settings, one needs to acquire an appropriate pragmatic competence to achieve effectiveness in communication. The authors in the special issue focus on several important notions, such as the definition of the successful English user (as ‘skilled user’ in Jenkins 2011 and as ‘effective communicator’ in Björkman 2011), what being an ‘international’ university should entail at different organizational levels (Björkman 2011, Haberland 2011, Jenkins 2011, and Knapp 2011), ownership issues (Haberland, Jenkins), the ways in which speakers in academic ELF settings strive to communicate successfully (Hynninen 2011 through ‘mediation’, Björkman 2011, through a variety of pragmatic strategies) as well as concerns regarding the importance of cultural and linguistic elements in successful ELF communication (Knapp 2011, Ljosland 2011). What the articles have in common is the definition or description of the ‘effective speaker’: Not someone whose spoken production is similar to or near a native speaker’s but someone who is pragmatically competent (Björkman

2.2 Work without normative elements 43

2011, Jenkins 2011). Some of the ways of achieving such pragmatic competence are described as using a variety of strategies (Björkman 2011) and mediation (Hynninen 2011). The strong message that arises from the special issue is that calling a university ‘international’ should have implications for the language policy documents in such settings, expressed explicitly in Jenkins’s article (see also section 6.3 on language policy).

2.3. Work including normative elements: Form In comparison to the great interest in a large number of studies on ELF pragmatics, a relatively smaller number of studies have been carried out on describing ELF as a language form, and a small part of the research has been text-oriented. Earlier studies have pointed to a need for research on the efficiency of ELF communication (House 2003a), the need to describe salient features and to develop descriptive work in general in ELF (Jenkins et al. 2002; Seidlhofer 2004) as well as the need for description and codification and how ELF is used in European educational settings (House 2003a; Seidlhofer 2004). Significant developments have taken place in the last five years, and systematic efforts are now made to record what is going on linguistically in ELF situations, especially as a result of the corpus work (e.g. ELFA and VOICE) and the consequent PhD projects. Work on form differs from work on pragmatics with regard to the use of norms. The yardstick for non-native speakers of English has almost exclusively been the native speaker ideal when it comes to phonological and morphosyntactic features. What is interesting about ELF usage is its divergence from the varieties previously described. It is precisely the native speaker as the ideal for international settings that is challenged by ELF research, calling for a need for prescription based on descriptive norms that are appropriate for international use. Some of the work on ELF linguistic forms has focused on phonology. Jenkins’s groundbreaking work on phonology is now well known to those both in and outside the field (Jenkins 2000, 2002). Phonological problems constitute an area known to create a great number of intelligibility problems. Jenkins’s description of L2 speech and assessment of which phonological features cause intelligibility problems and which do not has provided us with the core and non-core areas of English phonology, respectively, the set of sounds one needs to produce correctly in order to achieve intelligible speech, and the sounds that are not essential for intelligibility. The

44

2 Previous research on ELF

aim behind Jenkins’s study is pedagogical and twofold. First, the norm for speakers of English had always been native-speaker pronunciation, which is not an attainable goal for all speakers of English. The norm needed to be based on empirical evidence from appropriate speakers. The native speaker as the ideal target is unrealistic for many international speakers of English. The second issue concerns the standpoint in the study, which is one of the most critical parts of this work: Non-native-likeness is described as unproblematic and variation is acceptable as long as the speaker produces the core sounds correctly. According to the results of the study, based partly on recordings and partly field observation, the core areas are (Jenkins 2000: 159): • The consonantal inventory with the following conditions: Rhotic [!] [used] rather than other varieties of /r/ [i.e. rhotic, and using an approximant realization of /r/] Intervocalic /t/ rather than [!] Variations of [!], [!] and [!] allowed Approximations to core consonant sounds permissible (provided that the pronunciation does not cross the boundaries to another phoneme, allophonic variations within phonemes are allowed e.g. Spanish /v/ vs. [!] in word initial position) • Other phonetic requirements: Aspiration of the fortis plosives /p/, /t/ and /k/ Fortis/lenis differential effect on preceding vowel • Consonant clusters: No omission of sounds in word initial consonant clusters. Omission permissible in middle and final clusters (only if they comply with native English rules of syllable structure) • Vowel sounds: Maintenance of distinction between long and short vowels L2 regional variation acceptable (provided that it is consistent), [!!] to be kept • Tonic/Nuclear stress: Nuclear stress production and placement used appropriately to signal meaning The non-core areas, namely the areas that do not impede communication in English as an International Language (EIL) settings, are: • • • •

The consonant sounds [!] and [!] and the dark allophone [!] Vowel quality provided that the usage is consistent Weak forms Assimilation (in two consecutive words, the assimilation of the final sound in the first word to another sound due to the initial sound in the second word)

2.3 Work including normative elements: Form 45

• The direction of pitch movements regardless of what it signals, e.g. grammatical meaning or attitudes • Placement of word stress in general • Stress-timed rhythm The analyses of sounds that cause intelligibility problems surprisingly did not include areas that people regard as specifically ‘English’, e.g. the sounds [!] and [!] and the dark allophone [!]. Substitutes for these sounds, such as /f, v/ or /s, z/ or /t, d/ for /[!], [!]/ were permissible, not to mention that such variations are present in native varieties of English (Seidlhofer 2002: 26). Although Jenkins said herself that the LFC was not intended as a language production model but was a set of guidelines for effective use, this work created considerable debate. The debate was taken even further by questions regarding teaching pronunciation without native-speaker models, the impression a heavy accent might give and the effects it may create, whether the Lingua Franca Core (LFC) is merely an excuse for mediocre language use, why imposing LFC on students would be right if presenting them with Received Pronunciation (RP) or General American (GA) is unjustified, whether one could or should try to teach an accent which is nobody’s native accent, and finally, whether there are any other alternatives to achieving a model that is suitable to those who speak English in international settings. These questions have been central to the debates among those who have expressed concerns against a common core (Sobkowiak 2005; Szpyra-Kozlowska 2003) and those who see it as a useful tool (Björkman 2012a; Dalton-Puffer et al. 1997; McKay 2002; Walker 2001a, 2001b, 2005, 2010). Those who have criticized the LFC have not done so on empirical grounds but rather with native English language ideology while those who have replicated Jenkins’s study have found support for what is outlined in the LFC (Deterding and Kirkpatrick 2006; Osimk 2009; Rajadurai 2007) with the exception of Walker, who has identified a few areas where the LFC needs to be modified only slightly (Walker 2010). Listener factors are as important as speaker factors that other work has focused on (e.g. Jenkins 2000). The responsibility for successful interaction is naturally not solely with the speaker since intelligibility is not the speaker’s or the listener’s task; it is interactional between the two (Smith and Nelson 1985b: 333). A study on phonology within the international use of English also focuses on intelligibility issues and outlines the variables affecting intelligibility and comprehensibility, which are speaker and listener

46

2 Previous research on ELF

factors (Pickering 2006). Among the speaker factors are phonology and accentedness, while the listener factors are experience with variation in the phonology of the target language and listener attitudes. It is suggested that intelligibility of certain forms is directly proportional to familiarity with those forms (Pickering 2006: 224). Moreover, a listener who is willing to understand a speaker will find it easier to achieve intelligibility and will find that speaker intelligible (Pickering 2006: 226). Some other listenerrelated factors are familiarity with the accent, willingness, familiarity with the context and topic or simply tiredness or background noise. In the review of all these factors, the use of a NS model for pronunciation is somewhat dated since it has limited relevance in a world where English is used much more widely among its NNSs. As noted earlier in this chapter, in NNS-NNS interaction, speakers employ a variety of communication and accommodation strategies that aid the conversation, and they adjust their language to the situation, in pursuit of optimal communication. According to Pickering, these strategies are unique to ELF settings, and they may clash with the way native speakers negotiate meaning (Pickering 2006: 227). Pickering’s findings accord with other studies that investigate accents of English in specific ELF settings in different countries. A study on accents in Finnish textbooks emphasizes the benefits of a more realistic model for speakers in ELF situations and draws attention to the importance of familiarizing Finnish learners of English with a variety of accents since they speak English mostly with other NNSs (Kivistö 2005). According to the investigation, there are few NNS accents in the textbooks studied, and that textbook writers seem ambivalent about including any NNS accents at all in their work. Another study that investigated phonological features in one of its parts reports on the phonological features of German students’ English and maintains that their variations do not impede comprehensibility (Erling and Bartlett 2006). Phonology plays a prominent role in comprehensibility and has been described as the “greatest single barrier to successful communication” (Jenkins 2000: 83). It lends itself to such investigation rather readily since it is a closed system, unlike syntax, where possibilities of production are virtually infinite. Within investigations of ELF form, there has been some work on phraseology and idiomaticity. Idiomaticity has been reported not to have much relevance for the ELF speaker (Jenkins 2000: 220). It has been defined as a characteristic of L1 English, i.e. a feature that is present in native speakers’

2.3 Work including normative elements: Form 47

speech, and also a feature where native and non-native Englishes diverge (Seidlhofer 2001). The term ‘unilateral idiomaticity’, coined by Seidlhofer, refers to the phenomenon where the interlocutor, an L1 speaker, uses an opaque idiomatic expression that the non-native speaker fails to understand (Seidlhofer 2001: 136). It has been suggested that native speakers, as a part of ELF situations, may need to make adjustments to their idiomaticity to be understood by the non-native speakers they interact with (Prodromou 2003: 47). Advanced non-native speakers too have a certain degree of idiomaticity that they make use of, sometimes with a certain divergence from the original idiom. This divergence may be essential for their cultural identities and therefore perfectly legitimate. Idiomaticity becomes a problematic issue only if the set norm is L1 idiomatic usage (Prodromou 2007a: 23). The same author investigated idiomaticity in another study and pointed to the relationship between ENL and ELF (Prodromou 2007b). The data came from a corpus of naturally-occurring second language interaction where the investigation focused on idiomaticity and the effects it has for L1 and L2 speakers. The study suggests that idiomaticity depends on the goal the speaker has set for himself/herself. Most ELF speakers avoid opaque usage with which they are not familiar. Those who speak English in lingua franca situations generally have a good sense of judgment of where their boundaries are when it comes to employing idiomaticity. They do not use phrases that might lead to pragmatic failure (Prodromou 2007c: 38). Although the findings of these earlier studies on idiomaticity are in agreement with more recent studies, the difference in approach is most striking. These studies, unlike the more recent studies, have approached ELF as a type of usage that needs native-speaker English as a yardstick. This is not an appropriate approach to describe ELF usage for two reasons. First of all, most speakers in ELF settings do not speak it to communicate with native speakers. Therefore, their first agenda is not to be understood by native speakers of English but to be understood by a wide range of speakers from different L1s. Second, ELF deserves to be studied and described in its own right independent of ENL norms. Speakers in ELF settings employ idiomaticity creatively in ways that are clear and transparent to speakers from different L1 backgrounds. This was shown successfully in a more recent study (Pitzl 2009). Pitzl shows, first of all, that idiomatic expressions are used very differently in ELF usage compared to ENL usage. While in ENL, idiomatic expressions are used as fixed idioms, in ELF settings, speakers use them creatively and innovatively, e.g.

48

2 Previous research on ELF

We should not wake up any dogs instead of the conventional Let sleeping dogs lie. As we can see in this example, functionality in the ELF form of this idiomatic usage has not been affected. The meaning is preserved while the form is flexible. Some studies investigated other lexical phenomena within ELF form. Lexical vagueness is one of these phenomena, investigated in a study based on ELFA data (Metsä-Ketelä 2006). The expression more or less was first analyzed qualitatively and then quantitatively through statistical analyses to compare the results with MICASE data. The results showed that the expression was quite frequent in the ELFA data, and it occurred more in monologic speech events than in dialogic ones. The comparative part of the study showed that non-native speakers made more use of this expression. It also seemed to have three main functions, namely “minimizing, comparing similarities and approximating quantities” (Metsä-Ketelä 2006: 141). The earlier scarcity of research on ELF form can perhaps be explained by the lack of large corpora. The release of two important corpora which are free for use by researchers has changed this situation tremendously and has already provided researchers with substantial data, resulting in studies of ELF form that have focused on syntax and lexis, the two other elements that are important for optimal communication. A useful and important preliminary here, before moving on to studies on morphosyntax, is a review of these corpora that provided data for large-scale investigations on form. The Vienna Oxford International Corpus of English (VOICE; www.univie.ac.at/voice) was the first corpus of ELF to be started, led by Seidlhofer with recordings of about 1 million words of mainly English produced by speakers of continental European languages from professional, educational and informal settings. Projects based on this corpus have revealed valuable information on ELF settings, predominantly with reference to lexico-grammar. At the time of writing, VOICE includes recordings of an estimated 1250 ELF speakers with approximately 50 different first languages, and comprises a wide spectrum of speech events from different domains, functions and participant roles, i.e. interviews, press conferences, service encounters, seminar discussions, working group discussions, workshop discussions, meetings, panels, question-answer sessions and conversations. VOICE corpus became an open corpus, available online in May 2010 for all researchers. The English as a Lingua Franca in Academic Settings corpus (ELFA; www.eng.helsinki.fi/elfa) from Finland was started two years after the VOICE corpus but was the first one to be finished. It is the largest corpus

2.3 Work including normative elements: Form 49

work on ELF usage in academic settings. It supplies authentic data from naturally-occurring speech events, and these are crucially high-stakes academic events. ELFA contains 1 million words of transcribed speech of approximately 131 hours of recorded speech and has 650 speakers from 50 different first languages. The high-stakes nature of the recordings is a key issue, because it ensures that speakers actually want to and need to achieve the communicative goal set for the situation and that not being able to do so would be consequential for them. VOICE and ELFA corpus projects bear striking similarities with each other. They have only naturally-occurring speech and both have led the way to a number of doctoral projects that have already started contributing to knowledge of the usage of English in ELF settings. They are of similar sizes and are naturally both results of teamwork. The main difference, as mentioned briefly above, is that the ELFA corpus is a corpus of academic encounters only and does not include social types of talk, in contrast to the VOICE corpus. It comprises data from a range of disciplinary domains, the distribution of which are 29 per cent social sciences, 19 per cent technology, 17 per cent humanities, 13 per cent natural sciences, 10 per cent medicine, 7 per cent behavioral sciences and finally, 5 per cent economics and administration. The VOICE corpus on the other hand comprises speech events from professional and leisure domains as well as educational domains. So in this sense, VOICE is a broader project with more speakers from different domains. It has led to a number of projects in lexicogrammar, but it does not offer data from a large number of educational domains like the ELFA corpus. The most recent corpus project on ELF is ACE: The Asian Corpus of English. The compilation of this corpus started in Hong Kong with Andy Kirkpatrick as the team leader. The corpus is currently being compiled across East and Southeast Asia (Kirkpatrick 2010, 2012). The addition of this corpus to VOICE and ELFA enable comparisons of ELF usage between European settings (VOICE, ELFA) and Asian settings (ACE). Some of the earlier work based on VOICE data reported observations of ELF usage (Seidlhofer 2004). Although these findings are not based on quantitative data and were “hypotheses of usage” (Jenkins, Cogo, and Dewey 2011: 289), they constituted the set of ELF features as candidates for commonalities known to many today (Seidlhofer 2004). Among these commonalities are (Seidlhofer 2004: 240): 8

8. See Section 5.1, Table 5.1 for a brief discussion of these commonalities.

50

2 Previous research on ELF

Dropping the third person -s The interchangeable use of the relative pronouns which and who The non-standard usage of articles in general Invariable tag questions, e.g. isn’t it? No? Redundant prepositions, e.g. study about High frequency of some verbs of “high semantic generality” e.g. do, have, make Using that clauses instead of infinitive clauses, e.g. I want that... Increased explicitness, e.g. black color Seidlhofer’s study did not offer a list of features that resulted from empirical research; however, it served another very important purpose: It presented non-standard usage as variants instead of deviant forms of usage. Some of the features that were proposed as commonalities above found support from studies that are based on empirical data. To begin with, in an investigation of the third person -s in group discussions of representatives of the EU and national agencies of higher education, there were 29 cases of the third person singular -s and 15 cases of superfluous –s (Breiteneder 2005). However, almost 80 per cent of all cases followed standard norms. The study suggests that the variation in the use of the third person -s might be explained by the relation between some linguistic and extra-linguistic features. The linguistic features causing the omission of third person -s or the overgeneralization of it are summarized as collective head nouns (e.g. “ministry decide”), coordination (e.g. “the institutions and the network thinks”) and an indefinite expression (e.g. “everybody talk about this”). With reference to the extra-linguistic trigger, Breiteneder mentions the speakers’ focus on interaction and the content rather than on form. This kind of usage, she adds, shows parallelisms with New Englishes, some pidgins and creoles. Breiteneder, however, does not mention varietal differences in her discussion of the linguistic features, e.g. the way collective nouns are dealt with in BrE and AmE. Another item in the list of commonalities is the invariable question tag. This item, along with some other non-standard usage, is investigated in another study where the data are made up of four hours of naturallyoccurring ERASMUS student conversations of two different types: Casual talk and advisory talk (Hülmbauer 2007). The results are quite similar to the ones in the previously mentioned study, namely that correctness and effectiveness do not go hand in hand, and such variations of usage in ELF situations are natural developments that are based on effectiveness (Hülmbauer 2007: 29).

2.3 Work including normative elements: Form 51

Extended use of the progressive aspect, though not one of the features listed by Seidlhofer, was reported to be one of the commonalities in ELF usage (Björkman 2008a, 2008b; Ranta 2006). Speakers in ELF situations often seem to opt for the progressive form instead of the simple form even when the situation requires the simple form, e.g. referring to scientific and engineering phenomena where the results are always the same provided the conditions are kept (Björkman 2008a, 2008b). The results of both studies indicate that this feature is employed by speakers of different L1s. Comparative work based on the ELFA and MICASE corpora showed that speakers in ELF situations used the progressive form in a larger variety of contexts. However, it was actually used in a “standard-like” way in 87 per cent of all the cases (Ranta 2006: 111). Another study related to tense usage reported from a German university where students display a variation of tense and aspect usage (Erling 2002); however, the reason, as suggested in the results, was not because the students lack that particular usage in their L1 but because it simply was not essential for communication (Erling 2002: 11). This suggests that the students in Erling’s study dropped the features that were not necessary for communication irrespective of their absence or presence in their L1. In a broader study carried out in the same setting, Erling and Bartlett examined attitudes and motives of university students studying through the medium of English to features of phonology, lexico-grammar, i.e. article use, time, tense and aspect, expressing condition, adverbs and prepositions (Erling and Bartlett 2006). In the investigation made through interviews, student essays and questionnaires, they find a wide range of non-standard features. The variation in the features investigated was dealt with similarly: Nonstandard norms should be accepted on the basis of communicative effectiveness, which relates to the phenomenon of new Europeanism. This new Europeanism brings within the need to accept non-standard forms based on comprehensibility (Erling and Bartlett 2006: 22). Another study that focuses on the syntactic variation in the use of English in international settings analyzed 22 hours of naturally-occurring interactions and reported both unsystematic and systematic grammatical choices that the speakers made (Meierkord 2004). The results of this study indicated a surprising 9 per cent only of divergent usage, i.e. usage that did not conform to the prescriptive norms of English. This study differs from the others mentioned in this section in terms of its approach to the source of non-standardness in that the divergent usage was investigated in terms of L1 backgrounds. The finding that is of highest relevance to the present

52

2 Previous research on ELF

study is the set of underlying patterns in usage that coincide with patterns in New Englishes, e.g. patterns such as not marking plurality (Meierkord 2004: 113). Meierkord describes informal ELF as a heterogeneous language form that is heavily shaped by speakers’ L1 backgrounds. In addition, she claims transfer phenomena, developmental patterns and nativized norms as well as simplification, leveling and regularization strategies characterize ELF speech (Meierkord 2004: 129). In another study, Meierkord analyzed 24 hours of informal student conversations, half of which came from a student hall of residence in Great Britain and the other half from interactions at the University of Erfurt. A total of 74 speakers took part in the study, who came from a wide range of linguistic backgrounds with varying levels of competence in English, and English was not the first or the dominant language for any of them. The study analyzed the speakers’ lexicon, and the results showed that the speakers had few phrasal verbs or idioms in their speech, and the lexicon presented itself as more culturally neutral (Meierkord 2005: 101). The lexicon also seemed unstable and heterogeneous, varying according to the individuals having the conversations. Meierkord suggests that for the lexicon to show stability, the individuals would have to have had regular interactions in a stable group of participants (Meierkord 2005: 101). It is certainly striking that at least three studies, Breiteneder (2005), Ranta (2006) and Meierkord (2004), report fairly low percentages of nonstandard syntax. Whether this is generally true or not is unclear since few studies have provided information on frequency. To address this issue, the present study includes a section where standard and non-standard syntactic features are weighed against each other systematically (section 4.1).

2.4. When form and pragmatics meet The ELF work that started with pragmatics moved towards issues on form, and the relationship between form and pragmatic issues became a favorite topic for analysis. It has been argued that the two need not and cannot be separated. It has been suggested that there is an inherent link between lexico-grammar and pragmatics and that users’ lexico-grammatical choices are determined by pragmatic motives (Cogo and Dewey 2006). Investigations of the variation in lexico-grammar caused by pragmatic needs and the interrelationship between the two are significant elements in understanding the dynamics of ELF interaction. In other words, it was the area between lexi-

2.4 When form and pragmatics meet 53

co-grammatical variation and communicative effectiveness that called for further investigation. This was highlighted by Seidlhofer (2009a) as the need and challenge to move from a description of common ELF features to determining what functions the features fulfill in communication. A study carried out on the micro-scale has looked at dyadic conversations, group discussions and seminar presentations at King’s College London in terms of both pragmatic and lexico-grammatical issues, using data from two studies (Cogo and Dewey 2006). The two studies were merged, and the results presented valuable insights from two micro corpora with regard to how ELF speakers in the relevant setting communicated, i.e. what kind of pragmatic strategies they employed and what changes they made to the language to achieve optimal communication. Their conclusions suggest that speakers’ pragmatic motives, added prominence, reinforcement of position, increased explicitness and exploiting of redundancy, often caused changes in their choice of lexico-grammar (Cogo and Dewey 2006: 87). The lexico-grammatical features earlier presented by Seidlhofer (2004) (section 2.3) occurred in these two micro corpora with the exception of one feature only, which was using that clauses where the verb is expected to be in the infinitive, e.g. I want that-. From the analyses of both pragmatic and lexico-grammatical features, Cogo and Dewey conclude that the two are interconnected and the changes made to lexico-grammar arise from pragmatic communicative needs, which they refer to as efficiency of communication, added prominence, reinforcement of proposition, increased explicitness and exploiting redundancy (Cogo and Dewey 2006; Dewey 2007). The authors emphasize further that speakers in ELF settings make adjustments to form and use strategies to “enhance prominence” and “increase explicitness”, and they do so by adapting the language to the communicative situation (Dewey 2007). In their monograph, Cogo and Dewey (2011) focus on additional high frequency features in ELF lexico-grammar. The most common features include collocations, articles and prepositions. These studies show clearly that one of the aims of ELF research has been to show that form and function are inseparable (Cogo 2008: 60), and to be able to understand the relationship between form and function, it is necessary to investigate both areas thoroughly first. Cogo and Dewey restated this key point in their more recent study (2011). It is not the features themselves that are key but for what functions the features are employed.

54 2 Previous research on ELF 2.5. Attitude and irritation studies Another question of relevance to ELF research is how non-standard language patterns are perceived by those that need to operate in ELF settings. There has been much research about what speakers in ELF settings do when they interact with each other, but work is still scarce when it comes to speaker/listener relations and reception and receptive skills in general. As noted earlier, the responsibility for successful interaction rests with the listener as much as with the speaker. The importance of speaker reactions was recognized by much earlier research with the aim of assessing the direct effect of non-standard language, based on native speaker ratings of non-native errors (Albrechtsen et al. 1980; Burt 1975; Hultfors 1986; Johansson 1975; Ludwig 1982; Olsson 1977). These studies, naturally (due to date), were not carried out in ELF settings; however, they form the foundation of attitude and irritation work, which is why they have been included here. The aim in such earlier work was to find out what types of grammatical errors were more serious and irritating than others. This type of research treated non-standard production by non-native speakers as ‘interlanguage’9 or ‘learner language’ (Selinker 1972, 1992), and started with studies that investigated comprehensibility of learner language, which were studies on ‘error gravity’ and ‘tolerance tests’. A problem with such investigations was the inherent subjective nature of attitude studies. Attitudes are abstract and cannot be observed easily, which leaves the researcher with one main way of obtaining data, namely asking the interlocutors to judge comprehensibility and irritation. This undoubtedly generates subjective data and puts the validity of such investigations into question, which is a problem that was recognized in earlier studies. When it comes to meaning, interlocutors can only make guesses as to what the speaker intended to say, and concerning language form, they might choose to avoid reactions they feel would be unfavorable (Albrechtsen et al. 1980: 367). Subjective methods should, therefore, be supported with objective methods that show participant comprehension as opposed to what they think they understand. The studies that investigated the direct communicative effect of learner language have done so by either basing the study on the researcher’s analysis of the code, the interlocutors’

9. The term ‘learner language’ will be used throughout in the present study for its transparency.

2.5 Attitude and irritation studies 55

subjective description of reactions to different levels of production, or the researcher’s objective description of the interlocutors’ subjective reaction (Albrechtsen et al. 1980: 367). The subject, a non-native speaker, produces an utterance and the interlocutor, a native speaker, reacts to it. See Table 2.1 for a summary of the direct communicative effect of learner language (Albrechtsen et al. 1980: 367): Table 2.1. Analyses of the direct communicative effect of learner language, table taken from Albrecthsen et al. (1980: 367). A Researcher’s objective description of the code

B Interlocutor’s description of

C Researcher’ s objective description of the interlocutor’s

1 Pronunciation 2 Grammar 3 Lexis 4 Discourse

Reaction to 1 Code 2 Message 3 Personality

Reaction to 1 Code 2 Message 3 Personality

Comprehensibility of 4 Pronunciation 5 Grammar 6 Lexis 7 Discourse

Comprehensibility of 4 Pronunciation 5 Grammar 6 Lexis 7 Discourse

Going back to the literature, we see that Burt (1975) studied A2, Johansson (1975) A2, B1 and C1, Olsson (1977) A2 and A3 as well as C5 and C6. Albrechtsen et al. (1980) investigated the correlations between B1, B2 and B3.10 According to these studies, the direct effect of non-standard production on interlocutors could be that the message does not get decoded if the interlocutors are distracted by the non-standard nature of the utterance. This could be due to irritation (Johansson 1975). The results of these studies were somewhat contradictory. Some suggested it was the number of errors that was important rather than type of errors (Burt 1975; Galloway 1980; Olsson 1977) whereas some judged

10. The present study deals with A2, A3 and A4 and B1, B5 and B6.

56 2 Previous research on ELF lexical errors in general as more serious (Johansson 1978). Studies at a later date raised concerns at the accent level, that it is the speaker’s accent and the stereotype of his/her accent that is decisive as to whether the errors in that speaker’s speech “enhance or hinder second language oral communication” (Delamare 1996: 292). Regardless of their varying results, such studies led to other investigations that focused on the effects of learner language on the interlocutor, always conceived of as a native speaker of English (Albrechtsen et al. 1980: 366). Error analysis studies of the 1980s measured native speakers’ reactions to non-native speakers’ non-standard production with the exception of a few studies that included non-native reactions to non-native speech (e.g. Fayer and Krasinski 1987). Even some much more recent work looked at native speakers’ reactions toward non-native production (Munro and Derwing 1999). Today’s ELF situation requires a new approach. Most speakers of English today are non-native speakers, and they use English to interact with other non-native speakers much more frequently than with native speakers. This is certainly true when it comes to many domains in continental Europe. It would thus be of interest to learn how speakers in ELF settings feel about non-standard production. What communicates well and what kind of attitudes do they have towards non-standard production? Few investigations so far have focused on reactions to non-standard production by NNSs. Similarly, the investigations towards the use of English in continental Europe have been scarce. One of these few studies reported from Germany on students’ attitudes towards the increasing use and role of English in Germany (Erling 2004). The attitudes in this study were not towards linguistic forms since the focus was on macro attitudes which are more political than purely linguistic. Nevertheless, the investigation was carried out in an ELF setting and documented micro attitudes as well, i.e. students’ attitudes. The results showed that the macro fears and worries indicated by the media were not reciprocated by the students in a university setting. Students’ attitudes to varieties of English revealed their preference for native varieties as models, but there was noticeable tolerance for what is termed ‘regional variety’, i.e. English in the mouths of German speakers. Erling and Bartlett (2006) studied the same setting and examined these students’ attitudes and motives with reference to English. The findings of the study showed three clear groups of students: The pro-US group, the pro-British group and the pro-lingua franca group. The pro-lingua franca group was subjected to further investigation. Among the linguistic behavior

2.5 Attitude and irritation studies 57

of this group are making adjustments to the language depending on the purposes they have, showing their identities through English and claiming ownership of the language, seeing themselves as legitimate users of the language. Erling and Bartlett (2006) bring up the native/non-native dichotomy in connection with this notion of legitimacy, originating from Bakhtin (1981) and Bourdieu’s work (1991). Speakers of non-native varieties of English lack symbolic capital. First, their habitus, i.e. way of being, is influenced by the local culture and history of symbolic capital. Secondly, there is the second language they aim to incorporate in this symbolic capital, and thirdly, there is the linguistic marketplace in which the speakers operate. Erling and Bartlett define Europe, or New Europe, as the linguistic marketplace with new types of symbolic capital. In this community, the speakers can “cash in their new symbolic capital of being new Europeans” (Erling and Bartlett 2006: 33). The new linguistic marketplace is said to be an area of new symbolic capital without demands for L1 Englishes. Apart from studies that have investigated attitudes towards the use of English, there have also been a few studies that have focused on attitudes towards usage of English. One comprehensive study of this kind is Jenkins’s monograph on attitude and identity in ELF settings (Jenkins 2007). Jenkins takes into consideration the attitudes of teachers to both speech and writing in this work, with a special focus on attitudes to ELF accents based on large-scale questionnaire results on correctness, acceptability, pleasantness and familiarity of ELF accents. From the responses, BrE and AmE emerge as “unshakable”, the correctness or status of which were not questioned (Jenkins 2007: 179). They were preferred by the respondents in all respects. This appears as the most interesting finding of the questionnaire: “...despite the massive shift in the use and users of English over recent decades, many and perhaps the majority of teachers of English in the expanding circle countries continue to believe that ‘proper’ English resides in certain of its ‘ancestral homes’, principally the UK and the US” (Jenkins 2007: 188). Although not in the same comprehensive scale, there have been other studies on teachers’ (Decke-Cornill 2003; Grau 2005; Sifakis and Sugari 2005; Timmis 2002; Zacharias 2005) and learners’ attitudes to teaching and learning other varieties rather than standard native varieties respectively (Adolphs 2005; Friedrich 2000; Hakala 2007; Matsuda 2003; Shim 2002; Timmis 2002). However, there is very little on speakers’ attitudes in what are authentic ELF situations and not language-learning situations. Since it is mostly other non-native speakers who interact with nonnative speakers in ELF settings, it is appropriate to focus on non-native

58 2 Previous research on ELF speakers’ reactions towards non-standard production at different levels. A part of the present work addresses this issue among others (see section 4.3).

Chapter 3 Exploring an academic ELF setting in Sweden: The site

This monograph started with a description of the position of English today and provided a brief historical account of how it has become the language of science and technology. This brought the discussion to the notion of globalization with respect to higher education and multilingualism, and then to the reactions when one language gains such a powerful status. As we saw, there are strong reactions evoked by the fact that English has this position. After devoting a chapter to these important issues, in Chapter 2, I provided an overview of ELF studies that are of highest relevance to the present study with regard to normative elements, starting with work without normative elements on pragmatics, and then moved on to work including normative elements on form. After a brief section on pragmatics meeting form, I gave a review of some significant attitude and irritation studies. As the mentioned studies show clearly, ELF research is thriving with projects carried out at different linguistic levels from a variety of settings. At the same time, there is still need for more work on ELF used in academic settings. Although academic ELF has been described in a number of studies (e.g. Mauranen 2012; Smit 2010), academia is a domain that will certainly continue to become increasingly international, and there is more work to be done. In this chapter, I will undertake the task of describing the academic ELF setting in Sweden that I set out to explore. With a detailed description, I hope to provide for those who may be less informed about ELF a description of a typical academic ELF setting, and for those who are wellinformed, offer a possible point of comparison with their ELF settings. After this, our focus will move onto the research methodology and design that form the basis of the study. For purposes of transparency and clarity, I will explain all the steps of the research methodology as I took them during the field study.

60 3 Exploring an academic ELF setting in Sweden 3.1. An international university: Student and teacher body The present study focuses primarily on form and discourse and investigates the morphosyntax of spoken English as the academic language in Sweden and the pragmatic strategies speakers use. Sweden is an interesting setting for such an investigation, because it is one of the leading countries when it comes to English-medium tuition in continental Europe. There were 123 reported English-taught programs in Sweden already in 2007, which made the country number four on the list of the leading countries in continental Europe as providers of English-medium tuition (Wächter and Maiworm 2008: 29). This increased to 401 programs in 2011, putting the country in the third place after the Netherlands and Germany with 812 and 632 programs respectively (Institute of International Education). Up until relatively recently, Sweden was one of the few countries that did not require tuition fees from incoming students. This changed, however, when the Swedish government passed a law in spring 2010 outlining tuition and application fees for students outside EU/EEA countries. This change already took effect in the 2011/2012 academic year, causing a major decrease in the number of incoming students. The Swedish government says the fees were introduced for quality assurance purposes, in other words, to be able to attract top students instead of being a favorite destination for students who might simply be seeking free tuition. Another reason was to reallocate Swedish tax payers’ money to universities that already are or on their way to becoming centers of excellence. This development, needless to say, has caused serious debate in Sweden. A major argument in the debate has been that the student body in the country will be much less diverse, attracting students only from rich countries. In addition, some universities fear they will lose up to 80 or 90 per cent of their foreign students. Especially at universities and institutions where foreign students have been the majority, the new situation is likely to cause major changes in the student population. Some groups of students are exempt from these fees however. Exchange students can still receive free tuition, since their studies are financed by agreements between their home countries and Sweden. Doctoral programs will also continue to be tuitionfree. In addition, scholarships have been introduced to be able to continue attracting students from non-EU/EEA countries (Swedish National Agency for Higher Education). Despite the tuition fees and the possible decrease in the number of foreign students, studying in English is still seen as a natural step by the ma-

3.1 An international university 61

jority of students and scholars in Sweden to plan and prepare for an international career. Natural sciences, technology and engineering are the fields that have been most extensively anglicized. Gunnarsson and Öhman pointed out already more than a decade ago the position English has in Swedish universities, showing that English is very widely used in engineering, natural sciences and medicine (Gunnarsson and Öhman 1997). At the time of writing, the most recent sources report that 65 per cent of all Master’s programs in the country are offered in English (Salö 2010: 11). The situation is similar at the post-graduate level; about 87 per cent of all doctoral theses in Sweden are written in English (Salö 2010: 21). And among the European countries that offer the highest number of English-medium programs, Sweden is the only country that has engineering and technology as the most represented subject areas (Institute of International Education) unlike the other countries that have business and economics as the most represented areas, e.g. France. The present investigation was carried out at tertiary level. The university in question, which I will call the Technical Institute throughout, is responsible for a substantial proportion of Sweden’s technical research and engineering education. Students are offered a wide range of subjects and fields to choose from, ranging from natural sciences to architecture, industrial planning, work science and environmental technology. The annual report from 2012 shows approximately 4,600 employees, approximately 14,000 students enrolled in undergraduate programs and over 1,700 in postgraduate level programs. In 2011, there were 1,535 exchange students and a large number of international master’s students at the Technical Institute. This is a change from 1,700 exchange students in 2009, one year before the tuition fees took effect. In the university’s policy of internationalization report, communicating effectively in English is mentioned as an absolute requirement. English is used extensively in this setting, to allow for academic mobility of students and scholars and to prepare students for the global job market, among other reasons. As a consequence, there is a large number of exchange students and foreign scholars at the Technical Institute who speak English to communicate with each other. Swedish is the medium of instruction in the first three years, which is then replaced with English in the subsequent two years for purposes of internationalization. English is the only language of instruction in international Master’s programs and higher levels. Doctoral theses are almost exclusively written in English.

62 3 Exploring an academic ELF setting in Sweden The Technical Institute also introduced a Language Policy document in 2011. One of the important points in the document is that it states clearly the position English has as a global lingua franca today. Otherwise, the document aims to raise student and staff awareness of language-related issues. It is also stated that the document is meant as an aid to help the Technical Institute in their effort to become a multilingual academic setting. We will look more closely into language policy issues in Chapter 6 (see section 6.3)

3.2. Research questions and some important terms The previous section has shown that the Technical Institute is a highly international setting where English is used as a language franca on a daily basis. I set out to investigate the nature of ELF usage in this setting: • What, if any, are the morphosyntactic commonalities of non-standard usage in monologic and dialogic speech event types studied in the ELF setting examined? • Which of the commonalities found are the same as those described in the literature? • What kind of morphosyntactic non-standard usage results in overt disturbance in spoken ELF communication? • Which of the pragmatic strategies described in the literature are found in this setting? • What kind of morphosyntactic non-standard usage is perceived as irritating by speakers in ELF situations? In the initial stages of the present investigation, the hypotheses for each research question were as follows: • There will be commonalities of morphosyntactic non-standard usage across the two different speech event types. • The commonalities will be similar to the ones described in the literature. • Non-standard usage per se will not lead to overt disturbance. It is more likely that the combination of different types of non-standardness will create opacity in communication. • There will be similarities between the pragmatic strategies in the present study and those described in the literature; however, it is likely that

3.2 Research questions and important terms 63

there will be even more and different types of strategies used by the speakers in this setting, as the nature of interaction is very high-stakes in academic settings. • There will be some irritation towards non-standardness. The following terms will be used throughout this work and are therefore necessary for the understanding of the entire investigation. As they are also used in the description of methods in this section, it is best to include them here before going into research methodology and design in the next section. They are listed below in the order of frequency, from the most frequent to the least frequent: Standard/non-standard vs. incorrect: The term non-standard refers to what is traditionally considered incorrect in English usage according to prescriptive grammar books. It should be borne in mind at all times during the reading of this work that the term standard is not synonymous with native-speaker usage (the term non-standard will be elaborated on later in this section). The term incorrect is generally representative of the Second Language Acquisition (SLA) approach and is not relevant to ELF settings. Setting, speech event, situation and context: The term setting is the broadest of these four terms. It is used to refer to the place of interest, i.e. in the case of the present study, the Technical Institute. The term speech event is used in the sense used in MICASE (Simpson et al. 2002), simply as types of speech. Situation is narrower; there are many communicative situations in a speech event. Context is a segment in a situation, e.g. a context where two students are discussing a term. Disturbance vs. breakdown: The term disturbance refers to any turbulence during the communication process. It is unlike breakdown, which traditionally suggests that communication does not take place. In this sense, breakdown refers to a more severe problem during the communication process and is a bigger claim to make for a researcher. Disturbance is indicated by confirmation checks, repair requests and general requests for clarification through direct questions and repetition of troublesome items. The present investigation has studied overt disturbance only. Checking covert disturbance would require a different set of methods through which the comprehension of the subject matter is tested.

64 3 Exploring an academic ELF setting Irritation: The term irritation has been used here in the same sense it was used in the studies from the late seventies and early eighties (see Chapter 2). It refers to the listener-friendliness of a particular feature and attitudes towards it. Variety: The term variety has been used here to refer to autonomous varieties in the traditional way it is used in World Englishes research (Melchers and Shaw 2003: 28). A variety, therefore, can be codified and is consistent in terms of usage within itself. Intelligibility vs. comprehensibility: Some studies have used the term intelligibility synonymously with comprehensibility (Alptekin 2007: 267; House 1999; Kuo 2006: 216). In the present study, the term intelligibility is used to refer to the understanding of a word on the phonetic and phonological level only (Jenkins 2000), and for the understanding of content matters, the term comprehensibility is used. I believe that the reader will benefit from a distinction made between these two terms since the present study deals in part with both recognizing the sounds, i.e. intelligibility, and recognizing meaning, i.e. comprehensibility. Some earlier studies have made a similar distinction between the terms intelligible, i.e. recognizable, and comprehensible, i.e. recognizable in terms of meaning (Smith and Nelson 1985a11). High-stakes: The term has been used in the present study to refer to situations where there are consequences for speakers in the case of communicative turbulence and where it is critical for the speaker to convey the message to the audience, e.g. a high-stakes situation. A consequence of communicative turbulence in the present setting may be not being able to complete a group-work project, solve a problem, getting low grades or even failing a course.

11. Smith and Nelson (1985a) go into interpretability in addition to intelligibility and comprehension. They explain interpretability as being able to understand the speaker’s intentions, which has not been included in the present study.

3.3. Research methodology and design 65

3.3. Research methodology and design My aim in the present study was to investigate form and communicative effectiveness of ELF usage in spoken academic discourse. Clearly, this is a multi-layered research project which included several methodological details and decisions. This section will focus on the methods used to collect and analyze authentic spoken data from the academic setting in the present study in the order they were employed in the investigation. An essential preliminary here is the description of the methodology. The present study has used primary data and employed a variety of methods to obtain a detailed picture of the usage of English as a lingua franca at the institutional site. We will first visit the general research traditions and a main binary distinction of research methodology. A distinction is often made between qualitative and quantitative research. Basic as it may be, a good balance of both qualitative and quantitative methods contributes to the richness of data in the present work instead of only quantitative or qualitative data. Support comes from scholars who believe that a mixed methods approach “offers additional benefits to the phenomenon at question” instead of using only qualitative or quantitative methods (Dörnyei 2007: 47). The indistinguishability of these two research traditions was discussed also in some early work (Reichardt and Cook 1979). To start with, quantitative methods are known to be relatively fast and economical in terms of labor (Guilloteaux 2007: 123). They generally yield data that are reliable, replicable and generalizable. Qualitative methods, on the other hand, are more investigatory in the sense that they result in descriptive information. It is true that qualitative methods generally have a limited number of participants, but they provide the researcher with data that give a more nuanced view of the context. Another aspect to consider when it comes to the differences between quantitative and qualitative analyses is the selection of data. In quantitative research, points of investigation are generally preset whereas qualitative research is largely data-driven. In qualitative research, researchers do not work with preset analytical categories. This is practiced in order to avoid missing features that might be of interest. So, in the present study, using methods only of a quantitative nature would not yield rich and variable data that would be detailed enough with possibly important features going undetected, and qualitative methods only

66 3 Exploring an academic ELF setting would limit the study to a small, insufficient number of participants, which would not allow for the generalizability of the results. It is often the case that methods chosen for a certain type of investigation in a field mirror those of earlier studies. In some cases, the research field may be making use of mainly one or two methods, where data at other levels may go undetected. Discussing or considering other methods or combinations of methods, or carrying out parallel studies using different methods, can be a useful practice (Gustafsson et al. 2006). To obtain as much information as possible from a setting, it is useful to combine a set of methods. In the interest of generating maximally useful data, I made a decision to combine the advantages of both methods in a two-phase design where there was first a partly-quantitative phase that involves a large sample, the aim of which was to identify certain points of interest. This phase was followed by a section where the points of interest were investigated qualitatively in a subset of the large sample. Some previous studies have been designed in this manner (Dörnyei 2001), and similar methods have been employed in other studies, e.g. analyzing a small portion of a large corpus in an effort to make generalizations over a large population yet obtain detailed information (Charles, Pecorari, and Hunston 2009; Guilloteaux 2007: 124), or analyzing a big corpus and making deductions about a specific part of it to measure corpus adequacy (Nixon 1972). The present study adopted a similar two-phase approach. This twophase approach can perhaps best be explained by another binary distinction, namely ‘extensive’ and ‘intensive’ analyses, as discussed by Sayer (Sayer 1992: 243). I carried out the investigation of the large corpus extensively, and analyzed a subset of it intensively. The present study reports from a new, unexplored ELF environment with the aim of contributing to the existing knowledge of ELF in general and usage of English as a lingua franca in academia in particular. This required a large sample of recordings including a relatively large number of speakers that would allow for making general deductions from this ELF environment in particular. However, due to the nature of this work, data collection and analyses were handled by me as the only researcher, which naturally made it impossible for all the speech event recordings to be transcribed. Keeping this in mind, I made a combination of first extensive, and then intensive analyses, forming the two-phase nature of this study. For the extensive analyses of the large corpus, i.e. the entire material, I spotted all the instances of non-standard use along with an observation

3.3. Research methodology and design 67

protocol (Appendix 4) and transcribed them in their contexts, including their uptake. This, needless to say, was a laborious task. When doing so, I avoided having preset analytical categories in order not to miss other important data. The transcribing of the large corpus was done selectively, i.e. all the instances of non-standardness were transcribed. This allowed for getting information on usage on the entire corpus, a larger scale than is usually possible in research carried out by a single researcher. Since the instances were later on quantified, the extensive analyses are partly quantitative. In addition, I fully transcribed four lectures and four group-work sessions, resulting in a corpus of 46,662 words. I analyzed this subset intensively. This approach allowed for investigations at the discourse level in the selected speech events as well as for counting the number of standard features as opposed to non-standard features, both being useful additions to the study. Not only did this approach yield more data and details but it also provided the study with more transparency and clarity along with the proportion of instances of a given feature which deviated from standard forms. The combination of extensive and intensive analyses allowed the study both to outline general practices in morphosyntactic usage by the subjects and to give information on frequency and variability. Let us now turn to the several stages the study went through as it progressed from extensive to intensive analyses, which I will explain in chronological order.

Data collection My initial aim here was to collect primary data to find representative ELF speech by lecturers and students in the setting. In order to make sure the speech patterns would be as representative as possible of this specific ELF setting, I contacted several departments. Inevitably, the selection of departments was limited in part by their availability. Later, I recorded the lecturers that showed an interest in the investigation and responded positively, which provided the study with the lecture recordings. I found more lecturers through networking with the ones who had taken part in the study at an earlier stage, and this contributed to the number of lecture recordings. These together constitute the pool of primary data. In addition, a part of the lecture recordings came from already recorded lectures from a group of departments. These recordings were accompanied by the slides of the lecturers that could be viewed while listening to the sound files.

68 3 Exploring an academic ELF setting The lecturers that took part in the study had students from a wide range of linguistic backgrounds in their classes (see Table 3.1). I visited and carried out field analyses through recording and observation, which I will elaborate on later here in this section. A majority of the students from these intact classes agreed to take part in the study and to be recorded further for research purposes. So that the data would mirror the ELF situation as much as possible, I included all students in these intact classes; no selection of a certain profile of students was made. The lecture recordings total 42 hours and 44 minutes and the groupwork sessions, 28 hours and 41 minutes. There were 21 lectures given by 13 different lecturers, 14 of which were recorded by me. The group-work sessions total 24 group-work sessions from seven different courses, including 48 speakers. Altogether, the corpus captures an estimate of 502,000 words. There were nineteen L1s in total and 60 speakers, presented in Table 3.1: Table 3.1. The first languages and number of speakers in the present study, including the language typology for each language. LANGUAGES

Exchange students

Arabic (SVO,VSO) Bengali (SOV) Catalan (SVO) Chinese (SVO) English (US) (SVO) Finnish (SVO) French (SVO) German (SVO, OVS) Greek Icelandic (SVO) Italian (SVO) Persian (SOV) Punjabi (SOV) Russian (SVO) Somali (SOV) Spanish (SVO) Swedish (SVO, OVS) Turkish (SOV) Uzbek (SOV)

1 4 1 1 5 2 3 1 2 1 2 1 1 25

Ethnically non-Swedish students 3 1 1 1 3 2 11

Ethnically Swedish students

11 11

Lecturers

1 1 2 1 1 7 13

3.3. Research methodology and design 69

The subjects come from different first language backgrounds that belong to seven different language families with different sentence structure, namely Afro-Asiatic (Semitic: Arabic; Cushitic: Somali), Altaic (Turkish and Uzbek), Indo-European (Germanic: Swedish, German, Icelandic; Italic: French, Italian, Spanish, Catalan; Hellenic: Greek; Slavic: Russian; IndoIranian: Punjabi, Bengali, Persian), Uralic (Finnish) and Sino-Tibetan (Chinese). There was also one native speaker of English from the United States. Half of the speakers that took part in the study were exchange students from different countries. For exchange students, the language of interaction tends to be solely English, both in their social encounters in Sweden and in similar academic settings (Caudery, Petersen, and Shaw 2008). These subjects were all near-zero beginners of Swedish. Some chose to study the local language by taking language courses; however, most of them continued interactions in English throughout their stay in Sweden. A quarter of the students were Swedish with Swedish as their mother tongue. These Swedish students speak English in academic settings and Swedish in social encounters. The remaining quarter is made up of ethnically non-Swedish students, i.e. Swedish citizens who have another home language than Swedish. These students speak another language than Swedish at home, Swedish in daily interaction, and English in their lessons. Their first language backgrounds covered a wide range, none being predominant over the others. When it comes to lecturers, approximately half were Swedish (54%), and the other half were speakers of other languages (46%). The foreign lecturers had varying degrees of competence in Swedish, and English was the only language they used in academic settings. The students who volunteered for the recording did not receive compensation for their services. I offered feedback to the lecturers who volunteered for the study, provided that the feedback stage would be subsequent to the recording of their and their students’ speech. Most of them chose to get feedback and discussed English as the medium of instruction in depth. These data were not included in the results of the investigation. Another important matter was to ensure that the picture of the lingua franca usage at this university would be comprehensive and representative. For this reason, I made an effort to include several departments in the study and made arrangements to carry out recordings in five different departments. As we can see from the departments (Table 3.2), the present investigation dealt with spoken discourse in Applied Science only. The departments included all work with the application of knowledge from one or

70 3 Exploring an academic ELF setting more fields to solve practical problems, and the applications of the work they carry out lie within different areas in society. The subject areas ranged from chemical reaction engineering to rocket science. Table 3.2.

The departments included in the investigation.

Departments included in the Investigation Electrical engineering Industrial Engineering and Management

Energy technology Machine design Communication Systems

Information and Communication Technology

Chemical Science and Engineering

Computer and Systems Science Electronic, Computer and Software Systems Chemical Engineering and Technology Chemical Reaction Engineering

Engineering Sciences

Energy Processes Aeronautical and Vehicle Engineering

Recording Let us now turn to the recording of the data, undoubtedly a very important stage when analyzing spoken discourse. There is a considerable number of significant events in a classroom situation where many potential speakers are present. It is useful that the researcher does the recordings himself/herself rather than merely acquiring them from elsewhere in studies of this nature. Other studies that base their findings only on already-recorded speech events risk losing data that could be important in the analysis stage. My primary aim, naturally, was to perform the recording in an unobtrusive way. However, for ethical reasons, I informed all the subjects prior to the recordings.

3.3. Research methodology and design 71

Digital audio-recording has been the main method of data collection here. I recorded the data straight into the computer through a free crossplatform audio editor called Audacity (Audacity), using WAV and MP3 formats. WAV files require very large memory space up to about 85 000 000 bytes/ about 81 MB for a one-hour recording. However, they give the researcher good quality for formant analyses for phonetic analyses if need be. MP3 formats are compressed WAV files and take up much less space. Space precludes elaboration of further technical details of recording here. ‘Observation with protocol’ was another data collection method employed in this investigation along with the digital recording of both lectures and group-work sessions. Being present during the recording allows the researcher to truly observe the speech event. It is valuable to have an observation scheme to facilitate analyses to distinguish between different types of classroom activity and identify linguistic features of interest. For the observation in this study, I developed a simple observation scheme by adopting the COLT model (Communicative Orientation of Language Teaching) (Spada and Frölich 1995) scheme to the academic setting here (Appendix 4). As any researcher carrying out research including human beings, I also needed to ensure research ethics, as any research involving human beings can have short- or long-term implications for society. The present study adhered to the rules and regulations of the Swedish Research Council (Swedish Department of Education 1999). I followed the guidelines below: The name of the specific institutional site will not be mentioned in the thesis or in the publications that will result from the research, to protect the interests of the subjects and others that may be affected by the research. The names of the subjects were kept strictly anonymous in data archiving. Special codes were used for the speech events, which cannot be traced back to the speakers. Any other sensitive information regarding the subjects was kept strictly confidential. Only information that is useful for the investigation was kept, which is whether they had had extended stays in an English-speaking country, their first language, and whether they were bilingual or multilingual.

72 3 Exploring an academic ELF setting All the subjects confirmed in writing that they are volunteers. The subjects were given the right to withdraw even after the recording had taken place. I, as the researcher, agreed that the sound files obtained from the recordings were not going to be made available to the general public online.

Categorizing the speech events The next decision in the research design was to select and categorize the speech events. This can be done in different ways in different corpora. One way of dividing speech events up is taking location into account. The MICASE corpus, for example, includes sixteen different types of speech events, divided into two main groups: Classroom events and non-classroom events. The classroom events comprise small and large lectures, discussion sections, lab sections, seminars and student presentations whereas the nonclassroom events include advising sessions, colloquia, i.e. departmental or university-wide lectures, panel discussions or workshops, dissertation defenses, interviews for research purposes, meetings, office hours, service encounters, study groups, tours and tutorials (Simpson et al. 2002). However, what are subsumed under ‘non-classroom events’ in the MICASE corpus have proven to take place in classrooms in the corpus compiled for the present study. For example, student group-work, termed ‘study groups’ in MICASE, often takes place in classrooms in this Swedish setting with or without the presence of a teacher or a teaching assistant. Although students work on their own on a task they have been assigned, they do the work in classrooms. The present study uses the terms ‘monologic’, e.g. lectures, and ‘dialogic’, e.g. student group-work, to refer to the speech events recorded. Another way of organizing speech events is by taking interactivity into account. The ELFA corpus (www.eng.helsinki.fi/elfa), at the time of writing, comprises a number of types of speech events, e.g. lectures, seminars, conference presentations, etc. (Mauranen 2006a). The only ‘language’ related classification in the present corpus has been interactivity, namely the distinction between monologic and dialogic speech events. The speech events in this study have been classified by taking interactivity into consid-

3.3. Research methodology and design 73

eration, as in monologic and dialogic speech events, referring to lectures and student group-work respectively. The data in the present study come from authentic high-stakes technical speech from content courses, i.e. non-language-teaching settings. A typical example is the student group-work where the students working on a project together need to solve a problem, come up with a plan or system to better the existing situation etc. This can be compared to social speech where disturbance in communication has no or few consequences (Meierkord 2000), or to simulated ELF interaction where the participants feel as though they can change the topic at their convenience (House 2002a, 2002b). In the case of a student group project, however, any lack of communication or steering away from the topic would lead to the incompletion of the task, which would adversely affect the course performance of the students involved. So the aim in real high-stakes interaction is to communicate in a practical and functional fashion and get to the desired outcome. This allows for observations of real communication. All the recordings in this study are from naturally-occurring academic speech, contrary to practices in some previous work where set-up tasks for research purposes are used to generate data (e.g. Mollin 2006). Another point that distinguishes the present study from some of the earlier work is the fact that group-work as a genre has been included in the investigation. Such highly interactive speech events are quite rich in terms of examples of usage and therefore reveal considerable data that are not necessarily observable in other speech events. Such authentic group-work sessions have not been the target of small-scale individual research projects, most probably due to the difficulty of data collection. They are more typical of large corpora where a team of researchers is assigned to data collection. The inclusion of two extremely different speech event types should help cover many aspects of the ELF setting at hand.

Identifying non-standard usage The next step was to identify non-standard usage. I identified and transcribed the cases that did not follow prescriptive norms in terms of syntax, phonology and lexis as ‘non-standard’. I used the ELFA transcription guide to transcribe the selections. However, the size of the material yielded a large pool of data, which would not lend itself to intensive analyses. This presented a need to restrict the points to be investigated. I gathered all the

74 3 Exploring an academic ELF setting cases of usage that dealt with morphology and syntax in a list for further categorization, leaving phonology out of the main investigation (but see section 4.2.1 on question intonation). It was necessary to identify what kind of usage was to count as morphosyntactic. Any usage that is not exclusively morphosyntactic usage was discarded from the investigation. An example of this is: i found this /!!s/ results, where this could either be a case of phonologic non-standardness, the speaker’s failure to produce /!iz/ or lack of agreement between the pronoun and the plural noun. So any usage that could be the result of phonologic non-standardness rather than morphosyntactic non-standardness was left out. Broad variation among native varieties of English led to the exclusion of some features. It is very hard, even impossible, to point to standard usage where native speakers show variation. An example of this kind of usage is the area of prepositions, where there is variation between the British and American varieties. Prepositions were excluded from the study not only for variety differences but also because we simply cannot be sure what meaning is intended in some usage of prepositions. Another example of this kind of usage is the usage of the quantifiers ‘amount’ and ‘number’. Some native speakers, especially of US English, do not hesitate to use the noun ‘amount’ to quantify countable nouns whereas some other native speakers limit their usage only to uncountable nouns. See an example below from the corpus: amount of persons Another example of such a case is the existential ‘There is’ construction, which is commonly used in native varieties regardless of the subject. There is really some interesting success stories. There is a lot of clashes already. This usage would be described as ‘unacceptable’ in prescriptive grammar books (e.g. Fowler 1859) and as ‘informal’ in descriptive ones (e.g. Quirk et al. 1985), but it is frequently observable in native usage. Unlike in other studies where this feature has been included in the investigation, the present study excluded this feature because it is observable in native speech

3.3. Research methodology and design 75

and the ‘There is’/‘There are’ structures are used variably. The study carefully excluded cases where such differences were observed. The researcher’s possible lack of content knowledge or disciplinary traditions is another important factor to mention in such studies. Some disciplinary traditions and conventions are reflected in speakers’ usage of certain structures. For example, some terms are always used without a definite article in the field of Information Technology: ...if the prediction table is not as large as main memory then there will be... ...even if we get it from the instruction cache and not from main memory for the prediction table we need not have this In case of lack of familiarity with such practices, it is easy to misidentify some usage as non-standard. A typical indicator of such usage here was the teachers’ consistent usage of a feature. I made an effort to be wary of such specific usage as much as possible, e.g. by consulting subject specialists.

Categorizing all non-standard usage After all the instances of non-standard usage were listed, it was time to categorize all non-standard usage and draw distinctions that are systematic. The categorization applied here is shown in Figure 3.1:

Categorization of findings

Morphological!! Noun phrase Phrase level Syntactic

Verb phrase Clause level

Figure 3.1. The categorization of the findings.

76 3 Exploring an academic ELF setting It was important to have a clearly delimited scope, but equally important was to have a level of delicacy to allow for the majority of non-standard features to be included. The categories could not be so delicate that there would be too few instances in each category. The categories were devised so that a majority (75%) of the instances would be included. This practice resulted in the categorization shown in Figure 3.1. Related to categorizing non-standard usage was selecting features for further study and dealing with incidentalism, which can be a common problem in corpus studies (Swales 2002). To overcome this obstacle and to make sure only features that are frequent enough were included in the findings, a set of criteria was applied to each feature: The feature had to be used by five speakers of different L1s in both types of speech events and for a minimum of ten times. Only the features that met these criteria were included in the findings. Equally important was to check standard usage vs. non-standard usage. In order to have an idea of the frequency of the non-standard features in comparison with the standard structures that the speakers produced, I analyzed the four lectures and four student group-work sessions that I had transcribed for the intensive analyses. For each category (such as ‘subjectverb disagreement’) I noted all possible instances of non-standardness, e.g. all finite verbs, and compared the number of actual instances with that figure. I chose to do so, because I had observed variability of realization of these features during the recording and preliminary transcribing stages. Below are some examples of this: … in this floating unit four bits are quite enough, one, two, three and so on up to and including eleven four bits is enough to encode that… (taken from lecture, variability in agreement) ...i still don’t understand what what are the physics..... what is the physics behind this idea of putting a mass in front of the... (taken from group-work, variability in agreement) Another reason for documenting standard forms along with non-standard forms was reported variability in early ELF literature (Jenkins 2000) and in other ELF studies that came about a decade later (Firth 2009b). This variability, also observed here, alone calls for a thorough check of standardness versus non-standardness. Something that is certainly striking and makes it even more important to have a review of non-standardness is the surprisingly low percentage of non-standardness reported in usage in some stud-

3.3. Research methodology and design 77

ies: Breiteneder (2005), 20 per cent; Meierkord (2004), 9 per cent; Ranta (2006), 13 per cent. If non-standardness is so low in comparison with usage that conforms to prescriptive norms, can we talk about a set of features that cut across a wide range of settings? Why do some of the features seem to be identical with the features of learner language and New Englishes? Are these features then unique to the usage of English in ELF settings? For better understanding of ELF settings, these questions must be addressed. The present study aims to answer these questions in addition to the main research questions. The inclusion of this section will enable the study to show to what extent the discourse was characterized by variability, and if so, how stable the non-standard features were in the material, which would not be possible to obtain from the extensive analyses.

Checking for overt disturbance At this stage of the study, it was time to check for overt disturbance in the speech events and categorize the features as either ‘disturbing’ (D) or ‘nondisturbing’ (ND), indicating respectively whether the features cause overt comprehension problems or not. I examined the sound files carefully through listening by paying attention to the surrounding discourse and confirmation checks, repair requests and general requests for clarification through direct questions and repetition of troublesome items since such items are considered important indicators of misunderstanding in ELF research (Lappalainen 2001; Mauranen 2006a: 132). It is relatively easy to spot overt disturbance in student group-work. Even if they choose not to reveal problems in comprehension, miscommunication or lack of comprehension becomes apparent at a later stage of the group-work since, in such an event, they fail to get to the intended outcome. Lectures are, however, much harder, if not impossible, to analyze when it comes to comprehension. The present study relies predominantly on questions with regard to checking overt disturbance, both those raised during lectures and group-work. Questions are both direct and indirect signals of misunderstanding (Mauranen 2006a: 132–133). Other than questions, confirmation checks, requests for repair and clarification are considered signals of misunderstandings (Mauranen 2006a: 124). The general view of verbal communication is that understanding is generally not signaled (Mauranen 2006a: 128), except perhaps with backchan-

78 3 Exploring an academic ELF setting neling and nodding. However, if only overt disturbance is taken into account and investigated, the only behavior that is investigated is of a retrospective kind (Mauranen 2006a: 131). To get a better picture of the situation, speakers’ prospective behavior should also be taken into account, and to do so, speakers’ proactive talk should be considered (Mauranen 2006b: 131), which has been reported to be a prominent feature of ELF interaction (Mauranen 2006b: 135). In other words, what speakers do to avoid disturbance in communication to ensure information exchange is of great relevance. The present study investigates the discourse level by looking closely at the pragmatic strategies used by the speakers in both lectures and student group-work, and by doing so, goes beyond form. In the event of detecting overt disturbance, the study investigated the feature/s causing overt disturbance in depth in terms of morphosyntax and phonology if necessary. If analyses at other levels become necessary, e.g. analyses at the phonological level, I opted for the more advanced sound program Praat (Boersma and Weenink 2005) instead of Audacity, which was used during the recording stage. Praat allows for advanced speech analysis and speech synthesis, labeling and segmentation, speech manipulation, algorithms and has good graphics that are easy to read. After the investigation of non-standard morphosyntactic forms, the emphasis was put on the discourse level for further examination. This part of the investigation was designed subsequent to the completion of the first stage of this investigation. Because, during the recording stage, communication seemed to take place virtually without much disturbance despite frequent and considerable divergence from standard forms, it was quite likely that there were discourse strategies where meaning was negotiated or clarified among the speakers. Moreover, in order not to limit the work only to those cases where disturbance in communication is overtly recognized by the speaker(s), it was necessary to investigate the usage of discourse strategies and thereby look at ‘prospective behavior’ as well as ‘retrospective behavior’ (Mauranen 2006b: 131), also discussed by Kaur as preventative and remedial strategies respectively (Kaur 2009: 109). To reach a better understanding of the material at hand and to see whether a similar argument could be made, I analyzed the four lectures and four group-work sessions that I had transcribed earlier to ensure objectivity and to investigate variability in depth, to look into the number of occurrences of clarification strategies used by the speakers. Penz’s study (2008) on multicultural English-medium seminars at the European Center for modern languages was the starting point here, because her data show paral-

3.3. Research methodology and design 79

lelisms with the data in the present project: The multicultural nature of the setting, it being a specialized corpus, the incorporation of group-work discussions and data collection methods (participant observation and recording). The same criteria have been applied here. The pragmatic strategies investigated are comments on terms and concepts, on details and content of task, on discourse structure, signaling importance, comment on intent and common ground, with the additions of backchanneling, repair and repetition.

Perceived communicativeness and attitudes survey In a later stage of the present study, I investigated perceived communicativeness and attitudes towards non-standard morphosyntactic forms by means of a survey. I designed this survey after personal communication and interviews that revealed irritation towards non-standard English, from both the lecturers’ and students’ point of view. In other words, the survey was designed to see how common irritation was and what perceived communicativeness was like in this setting. This part of the study is different from the other parts with regard to data collection. The data came from an experimental situation and not from authentic communication. Although this might seem like compromising the quality of the general data here, it was important to get data on irritation since constant irritation can adversely affect communication. It was suggested as early as in the 1980s that comprehensibility and irritation are “intricately linked” (Ludwig 1982). When investigating attitudes and irritation in general, the problem the researcher faces is that the object is not observable (Albrechtsen et al. 1980). This requires a subjective data collection method, namely surveys and questionnaires where the subject himself/herself expresses an opinion on the point of investigation. This study investigated how ‘irritating’ or ‘non-irritating’ a feature was to examine attitudes and listener-friendliness using the data that were obtained through a survey, which is in a way a ‘tolerance test’. The method adopted here was to describe and report the interlocutors’ description of communicativeness and attitudes (objective), which has been regarded as a way to make attitude and irritation studies more objective (Chapter 2). When it comes to preparing the questionnaire sheet, I used a tripolar semantic differential scale (Osgood et al. 1957) for perceived communicativeness and attitudes. The information obtained from

80 3 Exploring an academic ELF setting semantic differential scales is regarded to be highly generalizable, which is the reason for applying this scale in the present study. The design of survey was subsequent to the investigation of the nonstandard morphosyntactic structures that were discovered in the investigation. I chose two or three sentences containing each non-standard feature (Appendix 3). Because most authentic sentences included more than one non-standard example of different types of morphosyntactic usage, I had to avoid exclusively authentic speech in some utterances, leaving only one non-standard example of usage in order to direct the respondents to a certain structure. Apart from this, I made sure not to manipulate the utterances in terms of form or lexis. The next issue had to do with the recording stage. I had two non-native speakers of English with indeterminate nonnative accents read the sentences into Audacity for recording which later made up the two different voices of the survey. One of the speakers was a native speaker of Swedish with a slight Swedish accent, and the other, a native speaker of German with a slight German accent. These two speakers were chosen to be representative of the type of English that is most frequently heard in the setting. In order to avoid getting responses on accents or the speaker’s voice instead of the intended morphosyntax, some groups listened to Speaker A and some, Speaker B, reading the same sentences from the material. The number of groups listening to each speaker was kept equal. I first piloted the survey with a group of respondents, all engineering students in the same setting. The piloting was undertaken to ensure the validity of the questionnaire in general and to make sure it would function effectively during the administration stage. According to the information obtained, some minor changes were required to the questionnaire as regards questionnaire design, especially to get the respondents to focus on the points of interest instead of other points that might stand out, e.g. unfamiliar words or unintelligible pronunciation of a word. The piloting stage also produced useful data on respondents’ views on speaking like native speakers (Chapter 5). In the administering stage, I selected courses with English as the language of instruction. After receiving information about the present research project, they all volunteered to spend some of the classroom time on this survey (convenience sampling). The respondents simply listened to the sentences and ticked the corresponding response on the questionnaire sheet under perceived communicativeness and irritation. Extra space was provid-

3.3. Research methodology and design 81

ed for those who wanted to write down additional comments. This completed the data gathering stages.

Chapter 4 Operating in a Swedish ELF site

4.1. Dimension 1: Form In the previous chapter, we looked at the different methods used to generate data for the study in the present monograph. In this chapter, we will examine the results, starting with this section on analyses in the form dimension, moving later on to the results on the communicativeness dimension. Finally, we will consider the results of the survey. With regard to form, the results regarding the phrase level will be followed by the results of the clause level. For a clear picture, the results for both levels in both lectures and student group-work sessions will be summarized in a tabular fashion. Finally, a comparison will be provided of the non-standard forms in the data with respect to standard production per hour. 4.1.1. Commonalities of usage We will start by considering the results in the phrase level, starting with the noun phrase (NP). The features found matching the frequency criteria (section 3.1) were ‘Not marking the plural on the noun’, ‘Article usage’, and ‘Double comparatives/superlatives’. One of the most interesting features perhaps is ‘Not marking the plural on the noun’, considering the importance and frequency of quantity bundles in engineering (Biber 2006: 170–171). In other words, engineers deal with numbers on a daily basis, and one would expect exact quantification to be critical, which would be reflected in their use of quantification. In the sample, many speakers seem to indicate the plural meaning merely by numbers or by adverbs or determiners before the noun but leave the noun itself without an inflexion. There were 159 instances of this type of usage in the lectures (3.3/hr) and 44 instances in student group-work (1.6/hr). This type of non-standard usage seems to occur in two different patterns in the material. In the first type, the speakers use quantifiers to indicate plural meaning. Some examples of this are given in (1–18):

Dimension 1: Form 83

(1) They have a range from 50 to 500 meter. Above 500 meter you have to consider Pelton turbines (2) Typically you want to have five kilogram of oxygen. (3) They are compatible even below 5 meter. (4) Here there are two type of equations. (5) I know one turbine with 1328 meter. (6) In our department we have three gasifier. (7) …two different reactor: reactor X and reactor Y. (8) We have three three three kind of reaction. (9) So we have two more condition for… (10) We have four parameter. (11) You can have ten glass vessel like we did. (12) … 4 cubic meter. (13) He said you can reduce from 8000 hour per year to 6000 hour per year. (14) You need to make sure it is 200 degree. (15) It is always ten digit. (16) And these are the destinations. Seven different destination. (17) Waterpower is very old. Hydroelectricity is 125 year. (18) It produces 700 megawatt for 11 months a year. As the examples above show, this usage is found both with regular nouns and units of measurement. In the second type, general quantifiers are used to indicate plural meaning, again without plural marking on the noun (19)– (27): (19) (20) (21) (22) (23) (24) (25) (26)

In many many case you can gasify it. … all the dynamic part. There are some difference between conventional… We have maybe several conclusion. You have several unknown. You also get a lot of infrastructure problem from wind turbines. ..., because we would need all the detail from the new equipments... You can actually compare several process and improve the use of the utilization and make it more… (27) In a couple of years I will have it in all three hydropower station. Apart from the two types exemplified above, there were some ambiguous cases where the type of non-standardness was not clear-cut. These cases are

84

4 Operating in a Swedish ELF site

unlike the cases where non-standardness can be observed at once in the NP. In such cases, I had to make a decision on what type of non-standard usage the instance had. Items (28) and (29) below are such cases. In (28), we cannot be sure whether the non-standardness is unmarked plural or zero article. Another example of such a case is (29) where the usage could be unmarked plural or lack of subject-verb agreement: (28) …just to get result. (29) Maybe there are some thing to improve. In such ambiguous cases, I studied the context closely, and I categorized the item as seemed most appropriate. The second non-standard usage in the noun phrase level deals with articles. Article usage can be deviant in terms of form or usage. There are two forms of the indefinite and definite articles in English, i.e. a/an and /!!/ /!i/, preceding consonants and vowels respectively. Not choosing the right form between the two alternatives for the definite or the indefinite articles might be regarded as leading to the same type of non-standardness. So saying ‘/!!/ efficiency’ instead of ‘/!i/ efficiency’ and ‘a efficiency curve’ instead of ‘an efficiency curve’ could be treated as falling into the same category of non-standardness. Nonetheless, an arbitrary decision was made to treat use of the incorrect allomorph a/an as morphosyntactic and incorrect forms of the as a pronunciation difficulty analogous to mispronunciation of other words, misplaced stress, etc. This is in line with treatment of misplaced stress in words like content or present. Because the aim in this section of the present study is to investigate morphosyntax only, the phonetic distinction for the definite articles has been excluded here. The following are examples where the article is simply in the wrong form (30–35): (30) (31) (32) (33) (34) (35)

You will have a efficiency curve… I have a exam then. So if you have a engine made for ethanol… This is a typical comparison between a old runner and a new runner. Duration curve is a important concept in hydrology. If you have a extremely efficient compressor…

The second type of non-standardness deals with usage, which has two subcategories. First there are articles that are superfluous, as in (36)–(42):

Dimension 1: Form 85

(36) You can use it in the different ways. (37) Pure hydrogen comes from the natural gas. (38) So question, what is renewables? Anybody can define the renewability? (39) If you don’t have the rainy climate it doesn’t work. (40) Bodies still exist but in the different form. (41) If you go to the Belgium, all the highways are lit. (42) The Einstein… There are also cases where the article is missing, exemplified in (43)–(55): (43) From those figures, you can have! idea what reasonable speed runner size is. (44) You can add timing interphase for ! memory system. (45) Who has paid for the infrastructure? That’s always! interesting question. (46) But they have! very good subway system. (47) It’s not !effective solution. (48) We come to !important conclusion. (49) It’s not that good !picture. (50) That solution is exactly !same solution. (51) So you got to be careful when you place the cells on !roof. (52) This is !argument they make. (53) You can find !discussion about the activation. (54) Even though there are thousands of people working in wind power, it is still very much !immature industry. (55) I made a remark about !internal grid. Altogether, there were 182 cases of non-standard article usage in the lectures (3.8/hr) and 41 in student group-work (1.5/hr). The last group of features at the noun phrase level, ‘double comparatives/superlatives’, includes adjective phrases such as “much more safer”, “much more cheaper”, “much more higher”, “more bigger”, “much more clearer”, “more stronger”, “more shorter”, “the most cheapest, “the most highest”, etc. There were relatively few cases of this type of non-standard usage in the corpus: Eight instances only in the lectures (0.2/hr) and 17 (0.6/hr) in student group-work.

86

4 Operating in a Swedish ELF site

Turning from the noun phrase level to the verb phrase level (VP), we see three different types of non-standard usage, namely ‘Subject-verb agreement’, ‘Tense and aspect issues’ and ‘Passive voice’. Non-standardness in subject-verb agreement (SVA) is a feature often found in L2 use. This type of non-standard usage occurs in two different patterns: The singular subject not agreeing with the verb and the plural subject in disagreement with the verb. Cases (56)–(67) are some instances of the first type: (56) I will talk about how a turbine operate in the system. (57) I call this A, which is a function of past history, what a catalyst have seen in its life. (58) One of the drawbacks is the amount of fluid this require. (59) So you end up with a nice mechanism which do not really contain the surface… (60) There is a further method which are sensitive to porosity in rocks. (61) What is dangerous is when the flow continue to another part of the surface where the pressure is slightly above the vapor pressure. (62) So the cost of electricity then come down to x kilowatt hour. (63) As you see, drilling cost increase (not) linearly. (64) We could also get destruction of the catalyst, mechanical destruction, which mean that we would actually spend the catalyst. (65) Chlorophyll generate the electrons. (66) The generation depend on... (67) If you’re in a country that do not have these kind of statistics,… Some examples of the second type, the verb not agreeing with the plural subject, are as follows, in (68)–(80): (68) (69) (70) (71) (72) (73) (74) (75) (76) (77)

However, the runner blades was not that good developed. Angle of the runner blades are reduced. All the different turbine types forms a curve. And many many parameters is affecting this one. Then comes Kaplan turbines. People gets mad because they want water there. That’s what essentially the electrons does. When the instruction window and the reorder buffer gets full… I will show you overhead pictures to explain how they works. Inorganic materials is very very poor in efficiency.

Dimension 1: Form 87

(78) (79) (80)

Two simulators I haven’t talked about here is A-sim and I-sim. I think all the creatures has two kind of activity. If incentives changes, the market can change…

There were 140 cases of subject-verb disagreement in the lectures (3/hr) and 59 in student group-work (2.1/hr). The most frequent feature when it comes to tense and aspect issues is the progressive, again a common feature in ELF usage (Ranta 2006). This is unlike native speaker academic discourse, for which the “simple aspect is overwhelmingly the preferred option” (Biber 2006: 63). At the same time, it has been shown that the progressive has been increasing for several hundred years and continues to do so (Smitterberg 2005). The speakers in the present setting often make sentences to refer to scientific or technical phenomena that are always true or valid, and despite this, they use the progressive instead of the simple form, as in (81)–(94): (81) A Francis turbine is using the whole turbine equation. (82) Typically the energy of the sun is emitting… (83) In water turbines the water is flowing along buckets before it leaves the runner. (84) My idea is to explain how this board is working. (85) But is that increase reflecting that we are dealing with a limited resource? (86) How much rain are you getting per year? (87) A power system is called a power system, because it is using different generator systems. (88) Do you understand how this is working not what it is but how it is working? (89) If the result is showing that one machine is better than the other… (90) Remember that part of the heating in Sweden is coming from… (91) But I don’t think so, if you go out you see a lot of trees which is living there. (92) So in average it is producing… (93) … the Francis turbine, which is the most famous turbine worldwide. It is producing more power than any other turbine type. (94) An asynchronous generator is depending on the net.

88

4 Operating in a Swedish ELF site

Examples like (95)–(97), where the simple form is used instead of the standard progressive form, were infrequent in the data even though they are typical of Swedish learner language (Smitterberg 2007): (95) Now I talk about optimized turbine with large… (96) You guys follow me? (97) We write about different things. Altogether, there were 145 cases of non-standard tense and aspect usage in the lectures (3/hr) and 36 in student group-work (1.3/hr). The third group at the VP level contains usage where passive voice would be expected. Although used much more frequently in engineering discourse than in other university registers, the passive voice is rare in spoken university registers (Biber 2006: 65). Correspondingly, there are few occurrences of deviant passive voice in the present material. However, there are enough cases to meet the criteria (98)–(106): (98) (99) (100) (101) (102) (103) (104) (105) (106)

And the plates get heat up very quickly. They are not directly affect by these concentrations. It can be happened that sometimes… What does it mean by molar fraction vapor? I think it’s a rather huge project that built underground. But we affect by the flow. Are we asking in the computer lab to find the minimum mass? When is this decision made what is going to include in this system. Certain steels are not corrode very quickly in alcohol environment.

There were 14 cases of non-standard passive voice usage in the lectures (0.3/hr) and 11 in student group-work (0.4/hr). We will now turn to the second main section in the syntactic analysis and consider the results in the clause level. The features found matching the frequency criteria were ‘Non-standard question formulation’, ‘Word Order’ and ‘Negation’. Non-standard question formulation is the first usage we will consider in the clause level. In the present study, I will use the term ‘question’ instead of ‘interrogative’ in classification and throughout. The reason for this is that the term interrogative is a way of classifying utterances by using structural criteria, i.e. that there is some change in intonation, that there is syntactic change in the sentence and finally, that there are insertions of words,

Dimension 1: Form 89

such as interrogative adverbs/pronouns, tags, etc. (Ultan 1978). In the present material, there are cases that do not follow any of these criteria but that are functionally questions, which is why structural data do not suffice and the term ‘interrogative’ is not fully appropriate in this case. In this study, I have defined questions functionally, following Quirk et al.’s definition, i.e. “utterances that seek information” (Quirk et al. 1985). The term ‘question’, therefore, is more appropriate here. There are two main types of questions, namely ‘Wh-questions also known as ‘Open questions’ and ‘Yes/No questions, also known as ‘polarity questions’ (Quirk et al. 1985: 808). I will refer to them as Wh-questions and Yes/No questions throughout. In Wh-questions, there are many possible answers that the question can generate while in Yes/No questions, two answers are possible: ‘Yes’ or ‘No’. There are many cases of non-standard question formulation of various types in the corpus as exemplified in (107)–(136), observed both in Whand Yes/No questions. See (107)–(124) for non-standard Wh-questions: (107) Why is such a difference in the diagram here between the blue and the red then? (108) So what kind of plant you have to consider? (109) Why is not good to combust directly? (110) Why this quotation only on one side? (111) Why it is X but not X0? (112) In the outlet…What we have in the outlet? (113) Why the function looks like that? (114) How many pages they have? (115) Where we are? (116) What other equation I would use? (117) Why it is like this? (118) What means two pages? (119) Why we place it there? (120) Which kind of the vortex we should have? (121) How many graphs we have? (122) How many pages you will work on this? (123) Why has ASTRA 18 hours? (124) So you started classification from which point you started? In (107)–(124) above, different types of complexities are mixed: Missing ‘there’ (107), missing ‘it’ (109), missing copula (110) and issues related to

90

4 Operating in a Swedish ELF site

the usage of ‘do’ (113, 118, 121, 124). The rest of the non-standardness deals with word order (115, 116, 117, 120, 122). (125)–(136) below are instances of Yes/No type questions, many of which are Declarative questions, i.e. questions with declarative word order: (125) (126) (127) (128) (129) (130) (131) (132) (133) (134) (135) (136)

Either one or two we have? Then I remove it? Anybody knows what is black liquor? We should go through every topic? Also J’s part after my part? Then I should remove it? You know what the main difference? T11 is the combustion? Refer to a real accident do you mean? You just start measuring more early? Anyone knows why you wouldn’t see that? So anybody can read the first sentence?

In (125)–(136) above, there are declarative questions (126, 130, 132, 134, 136), questions that deal with reported speech (127) and questions missing a verb (131). From these examples, it seems that speakers in ELF settings sometimes disregard question word order and other complexities. They ask questions mainly by using interrogative adverbs/pronouns e.g. what, why in Wh-questions and follow the declarative sentence word order in Yes/No questions. This type of usage is present mainly in the dialogic material in the present study. It is highly likely that the speech event type is the main factor behind this: In group-work, speakers often raise questions to complete the task whereas in lectures, it is not so often that the lecturer directs questions to the floor. This type of deviance differs from all others discussed here since it is the only one that overtly disturbs communication and leads to repetition and rephrasing in some cases (section 4.2). There were 20 cases of non-standard question formulation in the lectures (0.4/hr) and 62 in student group-work (2.2/hr). Non-standardness in questions are often word-order related. We will now turn to word order problems that are not in direct questions. The material has examples of non-standard word order, most of which have to do with indirect speech such as the ones in (137)–(153):

Dimension 1: Form 91

(137) (138) (139) (140) (141) (142) (143) (144) (145) (146) (147) (148) (149) (150) (151) (152) (153)

One of them is energy; another is how fast can you recover. Still we have to find out what sources do we have on this bus. Here you see how does it look like. We have to look at what did we here. The system shows what kind of a system are you going to be using. So I had to find out where is the registers. Salinity can affect what kind of material can you use and what type of X can you use. It depends on what do you have and what your needs are and so on. So that’s more or less what was it. Then you know how much is the flow. Here you see in terms of plant size, how does it work and how does it look like? You get a feeling of how is the cost developing for wind power. We can’t judge how solid is the ground. Maybe my problem I don’t know what does this mean. I don’t know what is the aim. You should mention what X is in your text because everyone doesn’t know what is it. The degradation of the EPI is proportional to how long will you keep it at high temperature.

The above examples (137)–(153) are all non-standard reported questions and all deal with indirectness. Of course, there were other cases of nonstandard word order in the material such as in (154)–(156) below. However, they were infrequent. (154) You are able to carry directly gas. (155) You cannot produce directly alcohol from potatoes. (156) ... to supply oil enough. There were 38 cases of non-standard word order in the lectures (0.8/hr) and 23 in student group-work (0.8/hr). Another category that meets the frequency criteria is negation. The type of non-standardness in negation is what one might call absence of raised negation from the subordinate clause to the main clause. This usage seems to be common especially in the dialogic material with some examples also in the monologic material, as in (157)–(168):

92

4 Operating in a Swedish ELF site

(157) (158) (159) (160) (161) (162) (163) (164) (165) (166) (167) (168)

I think it is not right to plot these in the same diagram. I think it’s not a proper way to describe it. This point is supposed to not move. I think my X is not OK here. But that caught fire not because of the hydrogen. I think he won’t be here. I think they cannot be… Do you have any nons that lead not to 0 here? No this can’t be. That would be not ten minutes. It has not always a low complexity. It is a not very good generator. It looks not good.

There were 13 cases of non-standard negation in the lectures (0.3/hr) and 20 in student group-work (0.7/hr). The third area of non-standardness in the form dimension deals with morphology. Non-standard morphology forms only a small part of the present investigation. The commonalities here have been categorized as ‘Nonstandard word formations’, ‘Analytic comparative’ and ‘Non-standard plural forms/Countability’. The speakers produce forms as in examples below (169)–(175): (169) (170) (171) (172) (173) (174) (175)

discriminization forsify levelize unlogical boringdom inofficial argument (verb)

Within morphology, there are also instances of the analytic comparative, such as in (176)–(181): (176) (177) (178) (179) (180) (181)

more narrow more cheap…more easy more clear more stiff more tiny more close

Dimension 1: Form 93

I have grouped non-standard plural forms, such as in (182)–(191) also under morphology. (182) (183) (184) (185) (186) (187) (188) (189) (190) (191)

Instead of showing how much liters will pass through,… … how many energy I need… We did very much things. I don’t know why you guys had less samples. One unit has very much disadvantages. … to give you an idea of how much efforts... … how many hydrogen you have around these carbons. That’s getting less students. He has a very good statistics about it. And many people say the belt gear it’s much simpler and less problems that can occur.

Some of the cases were overlaps between ‘article usage’ and ‘not marking the plural on the noun’. Example (190) above is an example of such a case. This example could be subsumed under ‘incorrect article’ or ‘non-standard plural’. I treated such uses of the indefinite article as cases of ‘non-standard plurals/countability’. I categorized cases with the definite article however, such as examples (41) and (42), as cases of non-standard article usage since there is no countability issue in these cases. Altogether there were 58 cases of morphological non-standard usage in the monologic material (an average of 0.4/hr) and 48 in the dialogic part (an average of 0.6/hr). Considering the size of the material, there are few cases of divergence from standard usage of morphology per hour. Providing per hour figures may not necessarily allow for a comparison with findings from other studies where the figures are given per number of words. This is naturally a consequence of carrying out extensive analyses (Chapter 3), which I would not have been able to do had I employed other methods. However, to mitigate this, I have provided figures per number of words for the intensive analyses in the present study (section 4.1.2). Table 4.1 shows the number of occurrences for each feature as exemplified above, in the noun phrase (NP), verb phrase (VP) and clause level (CL). The figures show clearly that the phrase level is richer in terms of non-standard usage compared to the clause level. The figures also show how little non-standard usage there is of morphology in comparison with syntax:

94

4 Operating in a Swedish ELF site

Table 4.1. The number of non-standard (NonS) morphosyntactic features found in the monologic (M) and dialogic (D) corpus.

M

D

!

Not marking the plural Article usage Double comparatives/superlatives

159 182 8

44 41 17

203 223 25

Subject-verb agreement Tense and aspect issues Passive voice

140 145 14

59 36 11

199 181 25

Non-standard questions Word order Negation

20 38 13

62 23 20

82 61 33

NonS word forms NonS analytic comparative NonS plural forms/countability

6 18 34

10 16 22

16 34 56

Non-standard morphosyntactic features Syntax

PL

NP

VP

CL

Morphology

In this section, we have seen that the present lingua franca corpus has a number of non-standard morphosyntactic features. This is, however, not unique to ELF usage (Chapter 5). A corpus of native speech would turn up such non-standard usage as well (section 3.3). The implications of the results in the form dimension will be elaborated on in Chapter 5.

4.1.2. Non-standard usage vs. standard usage Subsequent to the previous steps, I decided to calculate the average number of occurrences per 60 minutes of the non-standard morphosyntactic features in both the monologic and dialogic parts of the corpus. The following table, Table 4.2, shows these. The figures are relatively low, which indicated that the majority of the morphosyntactic constructions in the material actually conformed to standard forms. In addition to the figures per hour, I carried out intensive analyses and manually counted all the cases of standard and non-standard usage in the four lectures and four group-work sessions that I had fully transcribed ear-

Dimension 1: Form 95

lier (Chapter 3). This was considerable manual work, but it allowed for comparisons of standard vs. non-standard usage. Tables 4.3 and 4.4 show the numbers and percentages of the standard and non-standard features in the monologic and dialogic corpus respectively.

Table 4.2. The number of occurrences of the non-standard morphosyntactic features found both in the monologic (M) and dialogic (D) corpus. Non-standard morphosyntactic features Syntax

PL

M (/hr)

D (/hr)

NP Not marking the plural

3.3

1.6

Article usage

3.8

1.5

Double comparatives/superlatives

0.2

0.6

Subject-verb agreement

3

2.1

Tense and aspect issues

3

1.3

Passive voice

0.3

0.4

Non-standard questions

0.4

2.2

Word order

0.8

0.8

Negation

0.3

0.7

NonS word forms

0.1

0.4

NonS analytic comparative

0.4

0.6

NonS plural forms/countability

0.7

0.8

VP

CL

Morphology

Marking the plural Articles Comparatives/ Superlatives SVA Tense and aspect Passive voice Question formulation Negation Analytic comparative Plural forms/ countability Totals

3

7 0

6 2

2

1

2 0

3

26

270 8

183 187

36

7

47 10

49

826

NA

94

96 100

88

95

97 99

97 100

91

Lecture 1 NS S%

29

S

NA

6

4 0

13

5

3 1

3 0

NS % 9

1515

69

38 15

8

16

405 469

455 15

25

S

25

2

0 0

0

0

4 4

15 0

0

NA

97

100 100

100

100

99 99

97 100

10

Lecture 2 NS S%

NA

3

0 0

0

0

1 1

3 0

NS % 0

0

1 0

1

3

2 5

5 0

1

NA

100

99 100

90

89

99 99

99 100

95

Lecture 3 NS S%

1348 18

173

66 7

9

24

324 333

361 30

21

S

NA

0

1 0

10

11

1 1

1 0

NS % 5

1958

171

46 10

11

35

541 563

514 24

43

S

48

6

1 0

0

0

10 19

6 1

5

NA

97

98 100

100

100

98 97

99 96

90

Lecture 4 NS S%

NA

3

2 0

0

0

2 3

1 4

NS % 10

Table 4.3. The number of occurrences of the non-standard (NS) and standard (S) morphosyntactic features found in the monologic corpus. NA: Not applicable.

Marking the plural Articles Comparatives/ Superlatives SVA Tense and aspect Passive voice Question formulation Negation Analytic comparative Plural forms/ countability Totals

2

2 0

8 1

0

3

2 1

0

19

9

391 36

441 478

10

15

62 30

26

1498

NA

100

97 97

83

100

98 100

99 100

82

Group-work 1 S NS S%

NA

0

3 3

17

0

2 0

1 0

NS % 18

1535

61

130 14

63

10

511 519

161 17

49

29

1

0 0

0

1

6 1

12 1

7

NA

98

100 100

100

91

99 10

93 94

88

Group-work 2 S NS S%

NA

2

0 0

0

9

1 0

7 6

NS % 13

1876

65

103 8

42

23

554 555

504 12

10

15

0

1 0

7

0

0 0

6 0

1

NA

100

99 100

86

100

100 100

99 100

91

Group-work 3 S NS S%

NA

0

1 0

14

0

0 0

1 0

NS % 9

833

21

54 0

15

1

266 278

187 0

11

55

2

3 0

12

1

11 0

21 0

5

NA

91

95 0

56

50

96 100

90 0

69

Group-work 4 S NS S%

NA

9

5 0

44

50

4 0

10 0

NS % 31

Table 4.4. The number of occurrences of the non-standard (NS) and standard (S) morphosyntactic features found in the dialogic corpus. NA: Not applicable.

98

4 Operating in a Swedish ELF site

Tables 4.3 and 4.4 show all the morphosyntactic features in the data with the exception of word order and word forms. This was done due to reasons of practicality. As the figures show, non-standard features were not frequent in the data. Although there is more non-standardness in the student group-work sessions, the standard forms are much more common overall. I will discuss the implications of these results in further detail in Chapter 5.

4.2. Dimension 2: Communicativeness This section focuses communicativeness in the present study: First where communicativeness is not achieved, and next, by which means it is achieved when it is achieved. As touched upon in the methods section, in spite of the divergence from standard forms, there were very few cases of overt disturbance in the material. More significantly, among all the forms analyzed closely, only non-standard formulation of questions resulted in overt disturbance in communication and affected communicative effectiveness. This naturally calls for a detailed investigation, which I will provide in the immediate section (4.2.1). From overt disturbance, we will turn to how speakers in the present setting seem to achieve communicative effectiveness by using pragmatic strategies (4.2.2).

4.2.1. Investigating overt disturbance Since questions proved to be critical for communication and they were the only features that caused disturbance when produced in a non-standard way, I made a decision to expand the study and include phonological analyses to explore questions in more depth. When investigating questions, undoubtedly, question intonation shoulders an important responsibility. Therefore, a phonological investigation proved necessary. For the phonological analyses of question intonation, I investigated examples of questions that caused overt disturbance, questions that did not cause disturbance as well as questions that seemed to cause disturbance for extra-linguistic reasons, e.g. content-related reasons, phonologically. As I had planned at the outset of the study for possible phonological analyses (Chapter 3), I used the program Praat (Boersma and Weenink 2005). I made a separate sound file for each question, opened them in Praat win-

4.2 Dimension 2: Communicativeness 99

dows and got the program to draw the pitch contours in a different window. From these windows, I was able to study the pitch contours separately for the cases of Yes/No questions and Wh-questions. My aim here was to investigate what role question intonation played in the registering of a question by the listener if the speaker makes use of question intonation. Questions were present predominantly in the dialogic part of the material. This is quite natural since lectures, which are almost exclusively monologic speech events, do not usually have frequent instances of interaction but mostly questions that are posed by the lecturer to the audience. The following instances are all examples of non-standard questions from the dialogic material. In some of the cases including questions, non-standard questions were paraphrased (192), or repaired by the speaker (193)–(194), changed slightly and repeated (195), or simply repeated in the same form (196): (192)

<S1> how is the length of this . i mean how many pages you will work on this . i mean if you count for how much part in your work <S2> ahh , just two or three pages i think

(193)

<S1> it is given <S2> huh <S1> is it given <S2> yes

(194)

<S1> we should go through every topic <S2> mhm <S1> should we go through every topic <S2> not this one this one we already have

(195)

<S1> on the outlet , what we have on the outlet <S2> mhm , huh <S1> (in) the outlet what we have <S2> (xx)

(196)

<S1> so what other equation i would use <S2> yeah <S1> what what other equation i would use

100 4 Operating in a Swedish ELF site <S2> this one In (192) we have an example of a paraphrased question. S1 first asks the question by saying “how is the length of this” and then paraphrases to “how many pages you will work on this”, followed by “I mean if you count for how much part in your work”. S1 tries to provide clarification marked by “I mean”, which he does twice. In (193) and (194) we have declarative questions, i.e. “it is given” and “we should go through every topic”. S1s, i.e. speakers numbered 1, repair their questions by applying standard syntax when they ask the questions for the second time, which provides them with the answers. In (195), S1 repeats the question he asks in the first turn in a slightly different form to get an answer and changes from “in the outlet what we have in the outlet” and “in the outlet what we have”. When he rephrases the question, he topicalizes “in the outlet”, which seems to be an important piece of information in this interaction. He gets an answer after this rephrasing. In (196), we see the same form repeated by S1 with basically no change in syntax, i.e. “so what other equation I would use”. However, there is something S1 does differently when she asks the question for the second time: She makes changes to her prosody, which is generally a key difference between a declarative statement and an interrogative, apart from syntactic differences. At this point, questions appeared as significant features in ELF communication in the analyses. The initial research questions did not cover this area, and it would not have been logical to include previous research on questions in Chapter 2. I will undertake this task here. Asking questions, which is a speech act, may involve a variety of lexical, syntactic and prosodic means (van Heuven and van Zanten 2005). Among these, prosodic means play an important role in identifying questions. It is true that languages differ considerably with reference to interrogativity in terms of sentence melody. However, it has been suggested that most languages include some type of high pitch somewhere in the question, indicating interrogativity (Bolinger 1989; Lindsey 1985; van Heuven and Zanten 2005). In comparative studies on word and sentence prosody, findings have shown that interrogativity is always accompanied by the use of high pitch either locally, i.e. distributed somewhere in the utterance or globally, i.e. present throughout the utterance (van Heuven and Haan 2000; van Heuven and van Leyden 2003). So, it is common to use final rise in questions; however, it also has been suggested that it is by no means compulsory (Hirst 1998: 65).

4.2 Dimension 2: Communicativeness 101

To have a closer look into questions, we will consider Yes/No questions and Wh-questions separately: In Yes/No questions, there are three characteristics that tell declarative statements from interrogatives: Global characteristics, recurrent patterns and local characteristics (Hirst and Di Cristo 1998: 25). A pitch contour that is defined as global is placed on the entire utterance or on a part of it. US English, Swedish, Brazilian Portuguese, Finnish, Hungarian, Western Arabic, Vietnamese and Thai are examples of languages where utterances have global characteristics that differentiate interrogatives from declarative statements. There are also recurrent patterns that one can tell questions by, i.e. recurring pitch patterns on stressed parts in the utterance, such as in French. Finally, there are local characteristics, and this is best recognized by the high final rise in an utterance. It was reported long ago that about 70 per cent of a large sample of languages, i.e. almost 250, have utterancefinal rising pitch (Bolinger 1978). In the same investigation, the remaining 30 per cent had globally higher pitch, i.e. higher general pitch in the entire question compared to declaratives, for interrogatives than for declaratives. Wh-questions, on the other hand, are generally similar to declaratives rather than to Yes/No questions in terms of pitch patterns. In English, both Yes/No and Wh-questions have been reported to have falling intonation almost as often as they rise (Bolinger 1998: 50). There seem to be two possible cues for Yes/No questions and three for Wh-questions: Interrogative adverbs/pronouns (in Wh-questions only), syntax (especially word order) and intonation. To start with, in the present data in Wh-questions, we see that interrogative adverbs/pronouns are generally in place. The second cue is syntax. In this material, there is considerable non-standardness in question word order in both Wh- and Yes/No questions (e.g. 115, 116, 117, 120, 122), missing ‘there’ (e.g. 107), missing ‘it’ (e.g. 109), missing copula (e.g. 110) or issues related to the usage of ‘do’ (e.g. 113, 118, 121, 124) (section 4.1.1). The third cue is intonation. Although it has been suggested that rising intonation in questions in English is not compulsory in native speaker usage (Hirst 1998: 65), the plausible assumption is that it is most likely to be helpful to the listener, especially to those who use English as a lingua franca, where there are different types of non-standardness and complexities in speech. Despite varying results on the necessity of rising intonation to indicate interrogativity, the lack of question intonation when combined with other non-standardness in morphosyntax may be one of the factors that leads to disturbance in communication. Very often in the material, the non-standard

102 4 Operating in a Swedish ELF site syntax in the questions was accompanied by flat intonation. It may be that it is a combination of non-standard syntax and lack of clear question intonation that creates opacity for the interlocutors. The following is a Praat window, showing the pitch contour for “it is given” (193) in Figure 4.1. It should be noted that the pitch range is traditionally set at 75–300 Hz for male voices and 100–500 Hz for female voices, which is what I have done in the present study. it_is_given0 300

0

Pitch (Hz)

255 210

it

is

given

165 120 75

0

0.1361

0.2721

0.4082

0.5443

0.6804

Time (s)

Figure 4.1. The pitch contour for the question “it is given”.

The pitch contour for “it is given” is quite flat as Figure 4.1 shows. During the conversation, S2 cannot register the question. S1 repeats the question, this time with standard syntax, i.e. “is it given” and with clear rising intonation on the word “given” (Figure 4.2): is_it_given 0.00779516564 300

Pitch (Hz)

255

given

210

is it

165 120 75

0

0.1662

0.3325

0.4987 Time (s)

0.6649

0.8312

4.2 Dimension 2: Communicativeness 103 Figure 4.2. The pitch contour for the question “is it given”.

Notice the difference between 0.4082–0.5443 in Figure 4.1 and 0.4987– 0.9949 in Figure 4.2. It is not clear whether it is the corrected syntax or the clear rising intonation that helps S2 register the question. However, two cues are present in the second version of the question, which are both likely to have helped the listener register the question. In (194), the case is somewhat similar. The pitch contour here indicates that S1 is asking a question with non-standard syntax, i.e. “we should go through every topic” and then repairs it, i.e. “should we go through every topic” after S2 asks for repetition. The first time the speaker does this, the word order is non-standard, but there is rising intonation on different parts of the sentence. By doing so, the speaker has one of the two cues present, rising intonation, with a mean pitch of 139.6 Hz. See Figure 4.3 for the pitch contour of “we should go through every topic”: we_should_every_topic0 0.0126262547 300

2.3753288

Pitch (Hz)

255 we should

210

go through

every

topic

165 120 75

0

0.4751

0.9501

1.425

1.9

2.375

Time (s)

Figure 4.3. The pitch contour for the question “we should go through every topic”.

The second time S1 asks the question, he corrects the syntax of the question to “Should we go through every topic?” and keeps rising intonation (mean pitch 148.4 Hz), and in doing so, he has two of the three cues present: Rising intonation and standard syntax. It is highly likely that these two cues help the listener register the question. The pitch contour is shown in Figure 4.4:

104 4 Operating in a Swedish ELF site should_we_2_ 0.0121882556 300

Pitch (Hz)

255

topic

210

should we

go through

165

every

120 75

0

0.3318

0.6636

0.9953

1.327

1.659

Time (s)

Figure 4.4. The pitch contour for the question “should we go through every topic”.

In (195), S1 asks the question “what we have on the outlet”, and when doing that, he has two of the three cues present: The interrogative pronoun “what” and rising intonation. The do has been omitted. There is some rise in intonation (1.001–1.252), as shown in Figure 4.5 below: ELF_Question_1 0.0041107636 300

Pitch (Hz)

255 210

what

we have on the out

let

165 120 75

0

0.2503

0.5007

0.751

1.001

1.252

Time (s)

Figure 4.5. The pitch contour for the question “what we have on the outlet”.

4.2 Dimension 2: Communicativeness 105

This question with non-standard syntax does not get a response. S1 repeats the question, but he does not make any changes to the syntax or the intonation. However, he topicalizes “(in) the outlet”, which is an important piece of information. The importance of this piece of information is signaled by his repetition of it in the first utterance. Topicalization is an explicitness strategy (section 4.2.2), so S1 most probably wants to be more explicit and uses topicalization as a tool to achieve explicitness, a tool available to him at the time of speech in the conversation. He might not have been able to correct the syntax or have clearer question intonation. The pitch contour of the question in its second form, “in the outlet what we have”, is shown in Figure 4.6: what_we_have 0.00955361448 300

Pitch (Hz)

255 210

in the out

let

what we have

165 120 75

0

0.2601

0.5201

0.7802

1.04

1.3

Time (s)

Figure 4.6. The pitch contour for the question “in the outlet what we have”.

In (196), S1 repeats the question “so what other equation I would use” after getting a response which clearly is not the answer to the question. The nonstandard syntax is kept in the repeated version. There is a change in pitch, however. See the pitch contour of the question the first time S1 asks it in Figure 4.7, and the second time she asks the same question in Figure 4.8:

106 4 Operating in a Swedish ELF site equation_1 500

0

Pitch (Hz)

415 330

so what other

245

equation

I would

use

160 75

0

0.3095

0.619

0.9284

1.238

1.547

Time (s)

Figure 4.7. The pitch contour for the question “so what other equation I would use”.

The pitch contour of the repetition, “what other equation I would use” is shown in Figure 4.8: equation_2 0.0106954087 500

Pitch (Hz)

415 330

what other

equation

I would use

245 160 75

0

0.4024

0.8048

1.207

1.61

2.012

Time (s)

Figure 4.8. The pitch contour for the question “what other equation I would use”.

The second time S1 asks the question there is a clear rise (a rise from a maximum pitch of 193.8 Hz and a mean pitch of 168.4 Hz to a maximum pitch of 238 Hz and a mean pitch of 200.5 Hz). It seems reasonable to assume that this is the speaker’s attempt to make the question clear. Subse-

4.2 Dimension 2: Communicativeness 107

quent to the repetition with rising intonation, S2 appears to register the question. Question intonation, as suggested, is important for the question to be registered by the listener. So in some cases, despite non-standard syntax, if the intonation is sufficiently clear, the message seems to be conveyed to the listener. Excerpt (197) below is another example of such a case. S4 asks a question to S3 on line 1, with non-standard syntax but clear question intonation. S3 registers the question immediately and answers on line 3. See Figure 4.9 for the pitch contour. (197) <S4> why you always miss the lecture , sorry just curious <S3> (curious) er sometimes i have some other lectures <S4> but other lectures there’s some conflict in the timetable <S3> yeah <S4> mhm Elf-_why-QI 0.00988847232 500

Pitch (Hz)

420 340

why you

260

always miss

the lecture

180 100

0

0.4911

0.9822 1.473 Time (s)

1.964

2.456

Figure 4.9. The pitch contour for the question “why you always miss the lecture”.

In (197), S4 provides S3 with the interrogative adverb and question intonation (mean pitch 203 Hz), thereby providing the listener(s) with two cues. S4 asks another question on line 5, following up on the answer she got from S4 to her previous question. This time the question is a Yes/No question in the form of a declarative: “but other lectures there’s some conflict in the timetable” (line 3) (Fig. 4.10):

108 4 Operating in a Swedish ELF site declarative_QI 500

0

Pitch (Hz)

420 340 but

260

lectures in the timetable other there is some conflict

180 100

0

1.001

2.002 3.003 Time (s)

4.004

5.004

Figure 4.10. The pitch contour for the question “but other lectures there’s some conflict in the timetable”.

Although the syntax is not in the form of an interrogative, the rising intonation in the utterance is clear (mean pitch 211.5 Hz), between 4.004 and 5.004 Both questions asked by S4 in (197) are immediately registered by S3. Another example of a Wh-question where the syntax is non-standard, but the question intonation is present (mean pitch 207.8 Hz) is “what means two pages” on line 3:

1 2 3 4 5 6

(198) <S1> two pages conclusion it’s too much <S2> yeah i mean if i can conclude something in one paragraph i am gonna write my conclusions there what means two pages (…….) you can imagine i was trying to add words and phrases everywhere <S1> @yeah@ <S1>

4.2 Dimension 2: Communicativeness 109 what_means_two_pages 300

0

Pitch (Hz)

255

what means

210

two pages

165 120 75

0

0.2322

0.4644 0.6966 Time (s)

0.9288

1.161

Figure 4.11. The pitch contour for the question “what means two pages”.

The following is another example of a Yes/No question with non-standard syntax but with some rising intonation: (199)

<S1> and also J’s part after my part here <S2> yeah yeah

The question has non-standard syntax, but there is local question intonation at the end of the utterance (3.214–4.017) and a mean pitch of 207.8 Hz. part 500

0

Pitch (Hz)

420

and also J’s

340

part

here after my part

260 180 100

0

0.8034

1.607 2.41 Time (s)

3.214

4.017

Figure 4.12. The pitch contour for the question “and also J’s part after my part here”.

110 4 Operating in a Swedish ELF site In all the examples given in this section, it seems that the presence of as many of the three cues as possible is very helpful, but in some cases when the intonation is sufficiently clear, the question is registered and communication takes place without overt disturbance although other cues are missing. It is interesting that there should be overt disturbance in Wh-questions. This deserves special attention. The presence of an interrogative adverb/pronoun is a clear signal of a question in an interaction. It has also been suggested that the first words in an utterance are easily registered in comparison to the words that come later on (Giora 1988). In this respect, non-standard syntax in Yes/No questions or declaratives that are meant as questions with no question intonation are riskier than Wh-questions. However, this does not seem to be the case in the present study. As the previous section explained, registering the question seems to be much easier for the listener if all the three cues are present. The interrogative adverbs/pronouns alone do not ensure communication without overt disturbance when dealing with questions. The duration of an interrogative adverb/pronoun in a question is another matter that needs to be taken into account. The interrogative adverb or pronoun is the first word in the utterance and has considerably shorter duration in comparison to the other words in the question and the total duration of the question. To show this, I measured the duration of each interrogative adverb/pronoun in the Wh-questions given. See the following figures for the interrogative adverb/pronoun (IA/P) durations in comparison to the total duration (TD) in the Wh-questions given above: Table 4.5. The interrogative adverb/pronoun IA/P duration and the complete utterance duration in the Wh- questions. Question What we have on the outlet What other equation I would use Why you always miss the lecture What means two pages Why it is black 1 and 2 12

IA/P duration /secs 0.143 0.067

TD of utterance 1.252 1.547

IA/P duration as a percentage of TD 11% 4%

0.223

2.451

9%

0.239 0.220/0.183

1.161 1.153/0.825

20% 19%/ 22%

12. This question will be mentioned later in this section.

4.2 Dimension 2: Communicativeness 111

As the IA/P duration figures in Table 4.5 show, the interrogative adverbs/pronouns take up between 4 per cent and 22 per cent of the entire question in the examples given above, an average of 14.1 per cent. This surely depends on the length of the question; nevertheless, the figures give one the idea of how short the duration of an interrogative adverb/pronoun can be in a question, even in very short questions. In addition, the first words in an utterance are more likely to be missed than the subsequent words due to any number of reasons, e.g. lack of attention, background noise and unpreparedness for the utterance. Along with the nonstandardness in syntax and intonation, it is perhaps not so surprising that the interrogative adverbs/pronouns do not or cannot prevent disturbance. Another point may have to do with the intonation of Wh-questions. Whquestions have been reported to have falling intonation as often as rising intonation (Hirst 1998: 64). This is so even when used by native speakers; Wh-questions lack rising intonation. The combination of non-standard syntax, the short duration of the interrogative adverb/pronoun and falling intonation, which are three different types of complexities for the listener, can very well lead to overt disturbance in communication. The overall conclusion here then could be that speakers in ELF situations, because of other types of non-standardness in their interactions, need extra-explicit intonation to register and convey interrogativity. It was important during the analysis stage to attempt to make a distinction between overt disturbance caused by linguistic elements and disturbance caused by extra-linguistic elements, i.e. content-related issues in this case. Undoubtedly, especially in group-work where the students are working on challenging tasks, many content-related issues arise which can lead to disturbance in communication. The three excerpts below (200–202) are examples of cases where overt disturbance is caused by extra-linguistic elements. In (200) and (201) below, the question has standard syntax and the question intonation is there; nevertheless, there is overt disturbance. The reason for such kind of disturbance surely could be any number of reasons such as lack of attention from the listener etc., but it is likely that it is the difficulty of the content that leads to such cases of disturbance. In excerpt (200) below, S1 is asking a question on pressure change on lines 1 and 2. S2 responds on line 3 by saying “this one”, which shows that he registered the question. Her question is not registered with the intended meaning, however, and she repeats it along with some explanation on lines 4 and 5. Getting the response “ok” to her Yes/No type of question, she actually labels her speech act by saying

112 4 Operating in a Swedish ELF site “no it’s a question” on line 7. This is followed by laughs from both speakers, leading finally to S2’s attempt to answer the initial question on lines 8 and 9: 1 (200) <S1> but will will the pressure change be be the same with the 2 equation 3 <S2> this one 4 <S1> no the pressure change , the outcoming pressure will it 5 be the same in for different (pros) 6 <S2> ok 7 <S1> no it’s a question @@ 8 <S2> @@ , /öö/ it’s supposed to be change 9 but…

ELF-_ss-QI 0.0165951372 500

Pitch (Hz)

420

340 but will will the pressure change be be the

same with the equation

260

180

100

0

0.8204

1.641 2.461 Time (s)

3.282

4.102

Figure 4.13. The pitch contour for the question “but will the pressure change be be the same with the equation”.

Figure 4.14 shows the second question in excerpt 200, “outcoming pressure will it be the same in for different (pros)”, lines 4 and 5:

4.2 Dimension 2: Communicativeness 113 question_formulation_11 500

0

Pitch (Hz)

420 340 260

the outcoming pressure will it be the same in

for different pros

180 100

0

0.8181

1.636 2.454 Time (s)

3.272

4.09

Figure 4.14. The pitch contour for the question “outcoming pressure will it be the same in for different (pros)”.

If we look at what happens a few minutes later in the same recording, excerpt (201) below, the same two speakers encounter another case of overt disturbance in communication: 1 (201) 2 3 4 in 5 6 7 8

<S1> yeah but this is the total inlet flow isn’t it <S2> only . no <S2> <S1> ah is this in comparison one one flow one total flow comparison with one other total flow <S2> yeah <S2> <S1> is it <S2> it’s a question also <S2> <S1> @@ i think let me see let me see

In (201), the question “yeah but this is the total inlet flow isn’t it” (line 1) has standard (as opposed to non-standard) syntax. The question is only signaled by a tag question, i.e. “isn’t it” (line 1). In addition to standard syntax accompanied by the tag question, the speaker has rising question intonation (mean pitch 183.6Hz). See Figure 4.15.

114 4 Operating in a Swedish ELF site ELF-_with_question_intonation0 500

0

Pitch (Hz)

420 340 260

yeah but this is

the total inlet

flow isn’t it

180 100

0

0.5576

1.115 1.673 Time (s)

2.23

2.788

Figure 4.15. The pitch contour for the question “yeah but this is the total inlet flow isn’t it”.

So despite all the three cues being present, i.e. the standardness in syntax, the presence of question intonation and the question tag, which further signals that a question is being asked, the message is not successfully conveyed to S2. This suggests that there are likely to be extra-linguistic factors at play here, e.g. lack of knowledge of subject matter. A third example of a question causing overt disturbance due to extralinguistic elements is the question “but why the function looks like that” (line 1) below (202). 1 (202) 2 3 4 5 6 7

<S1> but why the function looks like that <S2> huh <S1> try to put the F1 on the other side , Fn or something <S2> on the other side <S1> mhm <S2> but we are not defining X3 <S1> ok

4.2 Dimension 2: Communicativeness 115 ELF_question_2 0.00646584587 300

Pitch (Hz)

255 210

but why the

165

function looks

like that

120 75

0

0.4203

0.8406 1.261 Time (s)

1.681

2.101

Figure 4.16. The pitch for the question “but why the function looks like that”.

There is some final rise in intonation at the end (1.681–2.101), with a mean pitch of 161.6 Hz. Although some rise is expected locally in the question, in native usage, Wh-questions have been reported to have falling intonation virtually as often as rising intonation (Hirst 1998: 64). So, in this respect, the intonation in these questions cannot be described as non-native-like. The speaker provides the listener with two of the three cues, some rising intonation and the interrogative adverb. Despite this, the listener needs some help with the question. S1 provides S2 with some assistance on the subject matter instead of repetition, correction or rephrasing. The issue is resolved through clarification of the subject matter. S2’s response “but we are not defining X3” (line 7) indicates that S1’s question was unexpected in terms of the task the speakers are working on. So, the overt disturbance in the question was probably not caused by non-standardness in linguistic elements but by an unexpected question in the course of the group-work exercise. Finally, we will discuss cases where there is no overt disturbance despite non-standardness. When the speaker prepares the listener(s) for the question to come by giving sufficient contextual clues, even if the syntax is non-standard and the question intonation is not identifiable, communication does take place, the question is registered by the listener(s) and the speaker gets an answer to his/her question, as in (203):

116 1 (203) 2 3 4 5 6 7 8 9 10 11 12 13 14

4 Operating in a Swedish ELF site

<S1> i am surprised look to the colors here . all of them in black only do you have an idea why <S2> mhm <S1> why it is black <S2> /öö/ <S3> (xx) @@ <S1> why it is black , you see <S2> because they are not running the (metafile) <S1> yeah <S2> this is ASF <S1> exactly <S2> this is back-up or something <S2> <S1> so we should /öö/ run it from (xx) <S3> mhm

In extract (203) above, S1 first draws the listeners’ attention to the colors and prepares the listener (line 1). He then makes a declarative statement, saying they are all black (line 1). Then he asks a standard question with some embedding, but this does not seem to be registered by S2 or S3 (line 4). He repeats his question, this time with non-standard syntax by saying “why it is black” (line 7). This question would not be non-standard in terms of syntax if it were part of the first turn (lines 1 and 2; Do you have an idea why it is black?); however, in isolation, the syntactic structure is nonstandard. Another feature of this question is the absence of rising intonation (mean pitch 146.6 Hz, maximum pitch 184.6 Hz). See Figure 4.17 below: why-black_short 300

0

Pitch (Hz)

255 210 why

165

it is

black

120 75

0

0.2254

0.4508 0.6763 0.9017 1.127 Time (s) Figure 4.17. The pitch contour for the question “why it is black” on line 4.

4.2 Dimension 2: Communicativeness 117

The second time S1 raises the same question (line 7), he keeps the nonstandard syntax, but seems to pay more attention to intonation. This is visible in the pitch contour when he asks the same question for the second time, with rising pitch on “black” (0.4949–0.6598), with a mean pitch of 156.8Hz and a maximum pitch of 197.9Hz, as shown in Figure 4.18 below: black_2 300

0

Pitch (Hz)

255

it is black

210

why

165 120 75

0

0.165

0.3299 0.4949 Time (s)

0.6598

0.8248

Figure 4.18. The pitch contour for the question “why it is black” on line 7.

When he asks the question for the second time, there is clearer rising intonation on the word “black”, a rise from a mean pitch of 154.8 Hz in the first question (line 4) to a mean pitch of 174.8 Hz in the second question (line 7). See Figures 4.19 and 4.20 for the pitch contour of “black” in the first (line 4) and second (line 7) utterances: black-1

black-2

0.000473295355 300

300 255 Pitch (Hz)

Pitch (Hz)

255 210 165 120 75

0

210 165 120

0

0.08045

0.1609 0.2413 Time (s)

0.3218

0.4022

Figure 4.19. The pitch contour of “black” on line 4. Mean pitch: 154.8 Hz.

75

0

0.05963

0.1193 0.1789 Time (s)

0.2385

0.2982

Figure 4.20. The pitch contour of “black” on line 7. Mean pitch: 174.8 Hz.

118

4 Operating in a Swedish ELF site

The same is true for the interrogative adverb “why”. See Figures 4.21 and 4.22 for the pitch contour of “why” in the first and second utterances, showing a rise from a mean pitch of 141.2 Hz in the first question (line 4) to a mean pitch of 172.8 Hz in the second question (line 7): why_1

why_2

0.000266223634 300

300 255 Pitch (Hz)

Pitch (Hz)

255 210 165 120 75

0

210 165 120

0

0.04525

0.0905 0.1358 Time (s)

0.181

0.2263

Figure 4.21. The pitch contour of “why” on line 4. Mean pitch: 141 Hz.

75

0

0.03925

0.07849 0.1177 Time (s)

0.157

0.1962

Figure 4.22. The pitch contour of “why” on line 7. Mean pitch: 172.8 Hz.

Subsequent to these changes, he gets an answer from S2, which he agrees to (lines 9–13). Following this, he can suggest a course of action toward solving the problem they are working on as a group “so we should /öö/ run it from (xx)” (line 13). If we look at the number of standard and non-standard questions in student group-work, we get to the following figures: Table 4.6. The number of standard and non-standard questions in the four groupwork (GW) sessions and whether they have cause disturbance or not. S/NS GW1 GW2 GW3 GW4

S NS S NS S NS S NS

Number of questions 11 1 56 5 52 5 13 17

Overt Disturbance 1 2

4.2 Dimension 2: Communicativeness 119

The conclusion we arrive at from the figures in Table 4.6 is that it is not questions in general that cause overt disturbance; the non-standard questions seem to cause overt disturbance. In the four group-sessions, none of the standard questions caused overt disturbance. The non-standard ones, however, led to overt disturbance in three cases. The finding that non-standard questions were the only features that caused overt disturbance indicates that questions per se are important realtime signals that show comprehension or disturbance. In this respect, the differences between lectures and student group-work are also likely to be of importance. To investigate these differences, I decided to investigate further the eight speech events that I had transcribed fully for Stage 7 of the present study. The following table, Table 4.7, shows the frequency of questions in the four lectures and four speech events: Table 4.7. Questions per speech event in lectures (L) and group-work (GW) and questions per hour.

Questions per speech event Questions per hour

Lectures L1 L2 16 7

L3 11

21.3

14.6

9.3

L4 9

Group-work GW1 GW2 12 61

GW3 57

GW4 30

12

16

76

40

81.3

The figures in Table 4.7 show that questions in lectures are much less frequent than in student group-work. A subsequent issue regarding lectures is how often students ask questions compared to lecturers. In student groupwork, naturally, it is the students who ask questions to each other. In lectures, however, it could be either. To investigate whether it is lecturers or students who ask more questions, I examined the eight speech events. It was equally necessary to distinguish rhetorical questions lecturers usually ask in teaching from authentic questions. Table 4.8 shows the percentage of questions asked by the students (QSs) and questions asked by the lecturers (QLs) in lectures and distinguishes rhetorical questions (QRLs) from genuine questions (QGLs) asked by the lecturers.

120 4 Operating in a Swedish ELF site Table 4.8. Questions asked by the lecturers (QLs) and the students (QSs) in the four lectures along with the rhetorical questions asked by the lecturers.

QSs QLs

QGLs QRLs

L1 N=16 4 (25%) 6 (37.5%) 6 (37.5%)

L2 N=7 0 3 (42.8%) 4 (57.1%)

L3 N=11 2 (18.2%) 0

L4 N=9 7 (77.7%) 0

9 (81.8%)

2 (22.2%)

The second row, QLs, shows the percentage of questions asked by the lecturers and what percentage of the lecturers’ questions were rhetorical (QRLs). The figures show, first of all, the percentage of the questions in the four lectures that were asked by the students, the rest being the questions asked by the lecturers. Among the questions asked by the lecturers, only 30 per cent of the questions were regular questions, as in (204) below:

1 2 3 4 5 6

(204) <S1> why do you think we use catalysts , in the first place , do you have any idea about that why would we use a catalyst . varför ska man använda katalysator , is there any point . well it’s not really totally obvious

In (204), the teacher pauses only for 2–3 seconds to get an answer to the question why catalysts need to be used (line 1). Following this, he reformulates the question (line 2) and immediately after repeats the same question in Swedish (line 3). These efforts together indicate that he does want to get an answer to his question from the students. He pauses for another 3–4 seconds (line 4) after he asks the question in different ways and in two different languages, which again indicates that he is waiting for an answer. When no student answers the question, he acknowledges this by saying “well it’s not really totally obvious” (lines 4 and 5). Example (205), is another case of a regular question: (205)

<S1> and and that of course involves other parts and which other parts could that be could be interesting in the design of catalytic reactor , apart from the catalyst isn’t life just about chemi-

4.2 Dimension 2: Communicativeness

121

cal reactions are there anything more in life , is there <S3> mixing <S1> mixing very good point (206) and (207) below are examples of rhetorical questions: 1 (206) 2 3 4

<S1> that’s considered a bad idea and why is it a bad idea well , branches occur frequently so it will , very quickly happen that , the processor will have to branch another time . here

In (206), the lecturer asks the question “why is it a bad idea” (lines 1–2) and answers the question himself (line lines 2–4) without waiting long enough for the students to give an answer. In (207), the lecturer asks the question “what can happen” (line 5) and answers it immediately without any pause (line 6). 1 (207) 2 3 4 5 6

<S1> for the prediction table we need not have this , how come , well the worst thing that can happen is this case we are to be we are to execute this branch instruction but we’re looking at the prediction bit corresponding to this instruction because they use the same prediction bit in the table what can happen well the prediction can be the same no problem

In (208), the lecturer first asks the question ‘why’ (line 4) and answers the question himself immediately without a pause (line 4). In the next sentence, he uses the word ‘endothermic’ and asks his next rhetorical question “what means endothermic” (lines 5 and 6). He explains the meaning of the term by answering his own question (the rest of the excerpt): 1 (208) 2 3 4 5 6 7 8 9 10

<S1> the most interesting part of this process is that final the final result carbon-monoxide plus pure hydrogen as a high energy value much higher than the mixture of methane and water much higher than the methane itself why because the (process) is very very endothermic what means endothermic the (xx) it needs heat so process swallows heat you must supply energy in terms of heat from (xx) this energy is usually supplied in the industrial processes by burning some part of the methane if you supply that energy with a very very low quality fuel then you would have an (energy) upgrade here

122 11 12

4 Operating in a Swedish ELF site

because this mixture has actually upgraded the energy coming from the solid fuel

From the figures given in tables 4.7 and 4.8, it is clear that questions are not very frequent in lectures. Previous research has shown that students ask and answer fewer questions in lectures given in English, and the results of this section of the present study confirm this (Airey 2009). Lecturers also ask few questions compared to the students in interactive group-work sessions. An important addition to the present study is the section on the frequency of rhetorical questions asked by the lecturers. This concludes the section on question intonation in the present investigation. The results of the intensive phonological analyses on questions can be summarized as follows: •

Non-standard syntax in question formulation can cause overt disturbance in ELF settings.

•

When syntax is non-standard in questions and causes overt disturbance, question intonation helps achieve meaning and register the question.

•

There are three cues that help the listener register a question in Whquestions (i.e. the interrogative adverb/pronoun, syntax and question intonation, and two for Yes/No questions, i.e. syntax and question intonation.

•

The failure to register a question can be caused by extra-linguistic issues such as issues related to content.

•

If there is enough contextual information in a situation, the question can get registered despite non-standard syntax and the absence of question intonation. Earlier research has shown the importance of shared contextual knowledge, i.e. shared disciplinary knowledge, in ELF communication (Mauranen 2007).

•

The findings appear to suggest that, despite what native speakers traditionally do, rising question intonation helps speakers ask and answer questions in ELF settings.

4.2 Dimension 2: Communicativeness

123

4.2.2. Analyses at discourse level: Pragmatic strategies In the present study, in an effort to investigate the material as thoroughly as possible, I wanted to go beyond form and investigate pragmatic strategies. At the outset of the study, I used Penz’s study (2008) on multicultural, English-medium seminars at the European Center for modern languages as a starting point due to parallelisms with the data and methods in the present project (Chapter 3): The multicultural nature of the setting, using a micro and specialized corpus, the incorporation of group-work discussions, and the data collection methods (participant observation and recording). I applied Penz’s criteria here to the best of my ability. However, since there always are borderline cases, the numbers given in the present section must be realized as approximations. I investigated pragmatic comments on terms and concepts, details, intent, discourse structure, discourse context and common ground alongside backchanneling, repair (self- and other-repair), repetition (for emphasis, repetition due to disfluency and repetition of others’ utterances). The study includes analyses of both prospective and retrospective usage in order to provide the reader with a better picture of the pragmatic strategies in the corpus. The pragmatic comments on discourse structure comprise both prospective usage, i.e. pointing forward in the speech event, and retrospective usage, i.e. referring back to an earlier part of the speech event and summarizing. In addition to these strategies, I have investigated predislocation/fronting and post-dislocation/tails, which I will discuss separately later on. Previous research has shown that these are elements of explicitness that help speakers achieve communicative effectiveness (Mauranen 2007), which is why it was important to incorporate them into the present study. In addition, topic abandonment was important in investigating the high-stakes nature of the speech events and in observing whether linguistic elements were causing topic abandonment. The number of cases for each feature in the four lectures and four group-work sessions are presented in Table 4.9:

124 4 Operating in a Swedish ELF site Table 4.9. The pragmatic strategies used in the four lectures (L) and four groupwork sessions (G). PRAGMATIC STRATEGIES Comment on terms and concepts Comment on details and content of task Comment on discourse structure Comment on discourse content Comment on signaling importance

L1

L2

L3

L4

G1

G2

G3

G4

7

2

9

1

0

5

0

0

0

45

18

24

11

6

21

7

2

16

8

7

23

4

4

0

22

17

23

16

21

43

11

0

5

7

6

9

0

4

0

5

Comment on intent Comment on common ground

7

14

2

1

20

5

6

0

149

110

33

0

46

83

107

149

Backchanneling

4

4

3

5

102

217

246

208

Repair

Self

7

9

0

4

20

5

4

12

Other

0

0

0

0

0

0

0

0

Emphasis

0

17

27

3

9

26

4

12

Disfluency

10

60

23

75

125

103

76

80

Other

2

3

0

1

14

7

10

5

215

304

152

146

391

508

489

478

Repetition

!

Because there may be differences among the speakers of speech rate, all the figures have been normalized (per 10,000 words). Topic abandonment has been excluded from Table 4.9 since it is not a discourse strategy but a discourse phenomenon. The ‘L’s from this point onwards will refer to figures and examples of usage by the lecturers, and ‘S’s will refer to figures of student usage and

4.2 Dimension 2: Communicativeness

125

examples from student usage. Although there are some parallels between the two speech event types with regard to the type of pragmatic strategies used, there are some differences worthy of attention. The results given in Table 4.9 above show, to start with, that there are very few cases of pragmatic signaling with regard to explaining terms and concepts. Examples of such usage are as follows, in (209)–(213), where speakers explain terms and concepts to the listener(s): (209) what is steam reforming , this is as i already mentioned a very very commercial way to produce hydrogen (210) why because the process is very very endothermic what means endothermic the (xx) it needs heat so process swallows heat you must supply energy in terms of heat from (211) solvolysis is another example of a possible method it is not commercial that’s another approach to try to produce liquid directly from solids er what is it it’s some kind of dissolving of the wood with the acid organic solvents at a certain condition (212) <S1> i can ask them if they have have a lease a lease program <S2> lease <S3> lease like you <S1> rent <S3> rent <S2> rent <S2> yeah but what is the <S1> leasa i don't know <S3> lease is in english she should know (213) <S4> what's null space <S1> null space is <S2> he said that er but you have a matrix like this er <S4> mhm <S2> you want to (xx) this matrix well this is er this has to be the the the base what's it called is it a base for the (xx) <S4> yeah ok <S1> mhm <S2> ok and er you want the image of you have how to explain that you have a space (you) n by n matrix ok it just n and space like n by n

126 4 Operating in a Swedish ELF site The second type of pragmatic comment deals with the details and content of the task. In this example, the speaker comments on the task, specifically on the instructions of the task. This seems to be a frequently employed technique in student group-work sessions. In (214), a student is commenting on how unsure s/he is about the details and content of task. In (215) and (216), students are discussing the content of the task they are working on. In (217), a student raises a question regarding the content of the task. (214) <S1> i don’t know if we’re supposed to know the code during the lab i think we’re just supposed to know it so we can model it later (215) <S2> and then just invert it and then solve the problem but i don’t see the point here of introducing i mean that’s that’s mathematical stuff that’s <S1> yeah , no u don’t think it’s got anything to do with physics it’s just a mathematical , i think so at least does anyone agree or diagree <S3> i agree <S1> because well that’s the way i interpret it it’s just a way of making physics come into mathematics in a good way <S2> of course but er i mean why do we do that why do we do another way it’s another way to [solve] the problem but <S1> [yeah] <S2> what’s the point of it (xx) and it doesn’t (xx) anything to just make it more <S1> yeah <S2> difficult <S1> yeah <S3> doesn’t it make it more general (216) <S1> ok we’re supposed to know the code well enough what code didn’t get it @@ <S3> @@ (xx) <S2> the code er at the er <S1> the lab codes <S2> yes yes yes (217) <S4> what are we supposed to do The next pragmatic strategy included both prospective and retrospective signaling of discourse structure. Examples (218)–(224) show how speakers

4.2 Dimension 2: Communicativeness

127

make discourse structure explicit by pointing forward in the speech event, thereby showing discourse structure prospectively: (218) we will start the third iteration of the loop (219) i would like to move on to branch prediction . ok so moving onto branch prediction (220) but one thing is er which i will explain soon is that using a catalyst er you will normally decrease the reaction temperature (221) i will talk a little about mechanisms and (xx) i will talk about how we can relate this to reactor design because that’s like the ultimate goal for for using this type of calculations (222) just before break i would like to say couple of words (223) part we will concentrate now is how to make fuel out of bio mass direct combustion goal the first step is gasification of biomass now we start from a solid fuel and gasify it (224) i will explain in a few seconds what green certificates are The following examples, (225)–(236), show how speakers make discourse structure explicit to the listener by referring back to previous points in the speech events and summarizing, thereby commenting on the discourse structure retrospectively: (225) where were we we have executed two , no that’s wrong we haven’t executed anything yet but we have issued two complete i iterations of this this small loop , nothing has been executed really because the memory unit hasn’t delivered load data (226) a few minutes ago i said that (227) you have now seen one way of er handling renaming an outof-order execution (228) following the one i pointed out (229) we now produced the b (230) to summarize the first part of the lecture here i would say the catalyst … (231) pyrolysis you have touched upon a little bit last time with anaerobic digestion (232) this is as i already mentioned a very very commercial way to produce hydrogen

128 4 Operating in a Swedish ELF site (233) (234) (235) (236)

if we go back to the options in biomass conversion so we saw that er that’s what i just tried to explain i tried to emphasize this one before ok coming back to the fixed quotes combined with green certificate trading

The next strategy investigated is pragmatic comment on discourse content, which can be seen as labeling the speech act, as follows (237)–(241): (237) (238) (239) (240) (241)

<S1> yeah that’s my question <S1> <S1> that’s what i asked myself <S4> i had the same question my in my mind when i read that because it’s <S3> so i don’t know i i agree with you <S2> yeah yeah i see the (point) but] i am just wondering what could be the the the complicated

A pragmatic strategy that is relevant to discourse context/content is about signaling importance, which is a strategy used predominantly by the lecturers. Below, (242)–(250) are some examples from the corpus: (242)

(243) (244)

(245) (246) (247) (248) (249) (250)

decoder it is important that incorrect (xx) codes are never delivered to the processor because then it will go away and execute instructions it is noteworthy that instruction are taken from the … one thing which is very important er from an industrial point of view is that you normally will end up with a more selective system meaning that you will produce less by-products i would say the heat exchange is very important clear physical process er but there’s a very important point here it’s to er we really doesn’t change the equilibrium in a reaction methanol is a very important first step product in many many oil industries of course the most important thing is economy some other very important thing is actually you have to connect wind turbines whenever possible this is the main discussion at the moment for all the offshore windfarms

4.2 Dimension 2: Communicativeness

129

The next pragmatic comment is on intent, which in a way is labeling the speech act. The speakers sometimes make the intent explicit as done in the examples below, (251)–(253): (251) (252) (253)

<S1> yeah i just wanted him to repeat that <S1> this was what i wanted <S1> because i just wanted to know how how you motivate continuous equations of motion

The speakers in group-work sessions often have pragmatic comments on common ground, where they comment on what they need to do together as a group (254)–(258): (254) (255) (256) (257) (258)

<S1> we have to check the distillation process <S1> then we need batches for four cubic meter then then it’s not (downscaled) any more <S1> so basically what we need to think about is cause everyone got one about point two kilos right <S1> yes but we want to compare them the experiment <S1> yeah we can have this as the lecture topic and the extra request <S3> true

The study, in addition to the pragmatic comments and strategies above, includes backchanneling, repair, repetition, topic abandonment and preand post-dislocation. The rest of this section will give examples of such usage from the present material: Backchanneling is the acknowledgement of what the other speaker has said, found often in the group-work sessions here, as shown in examples (259)–(263) below: (259)

(260) (261)

<S2> it means we can only have flutter results with computers <S3> yeah <S2> it should derive at once you'll have some kind of <S1> yeah <S1> i said really two thousand dollars he said yeah i said for the whole for the whole unit <S2> yeah yeah yeah

130 4 Operating in a Swedish ELF site (262)

(263)

<S1> i don't know maybe he said so under four cubic meters and he said it's eight not four <S2> yeah yeah i know i know <S1> mhm <S2> but i think it’s right because when you when you are in a plane you can see that the wing is is vibrating it’s not going so it’s right that <S3> mhm

When it comes to repair, the second present investigation set out to study both self- and other-repair. There were, however, quite strikingly, no cases of other-repair in the eight speech events that were fully transcribed. The following, (264)–(274) are all examples of self-repair: (264) (265)

(266)

(267)

(268) (269) (270) (271) (272)

(273) (274)

<S1> but it should be less , lower <S3> yeah you see flutter is , on the real part everything is zero it’s flutter , that is not really unstable not really stable to be (xx) so , let me see now label <S1> well perhaps you can insert more constrange constraints in this matrix and we use the matrix to collect all the constraints [in er visualized visual visibly] <S1> you add points on so here down here you end up at the wing tail i guess so it's the first one it's the wing root the wing tip sorry no wing root wing root and er there were others er innovating , innovators such as er for example studying the heterogeneous process why is that why can it can it stick to the why can’t it stick to the surface but there is are also examples it’s still very very interesting but we have a long time long way to go to make it commercial so this reservation station enables its right signal and reads data of the bus and stores it in the reservation station the other locations just sinently silenty we are predictly correct predicting correctly every time it will requires it will require two mispredictions again before we change the prediction

4.2 Dimension 2: Communicativeness

131

The examples given above show different types of repair by the speakers: most of the self-repairs are repairs of linguistic elements (264–266, 268– 272, 274). There are, however, also repairs of occasional slips of the tongue (273) and repair that seems to be related to content (267). Repetition is the next pragmatic phenomenon I will consider here. It differs from the rest of the strategies as regards its nature: It is a formal category. The cases of repetition found in the transcripts were categorized as ‘repetition for emphasis purposes’, ‘repetition of disfluencies’ and ‘repetition of parts of others’ utterances’. Cases of repetition for backchanneling purposes, e.g. “yeah yeah”, have been excluded from the repetition figures for reliability purposes. (275)–(282) below are examples of repetition for emphasis, often indicated by ‘very’: (275) (276) (277) (278) (279) (280) (281) (282)

if you have a homogeneous process it’s very very tricky to separate er er the catalyst from the the product solution upgrade it liquid to gaseous fuels and the very very commercial standard process of today why because the process is very very endothermic it should be handled very very very very carefully that’s why we cannot use really hydrogen now because it’s a very very very very poor energy per volume ratio <S1> it's the wing root the wing tip sorry no wing root wing root <S1> it's nothing it's nothing <S3> it’s very strange it’s very strange

The following, (283)–(288) are examples of repetition that seem to have been caused by disfluency: (283) (284) (285) (286)

(287) (288)

<S2> yeah yeah if i am if i am [right] <S1> and i said i said yeah i am sitting in front of a picture for the whole thing and i said only two thousand he said yeah <S1> if it's it's reasonable if they can they can provide us with three hundred units <S1> and then he said you can speculate for the sensitivity analysis the degradation is safety if you if you if you take a hundred hundred degrees over <S3> we can so we can connect now can you connect the <S4> it's it's for water water treatment

132 4 Operating in a Swedish ELF site

The last type of repetition was repetition of parts of others’ utterances, as shown in the examples below. (289)–(290) are examples from student group-work, and (291) and (292) are from lectures: (289) (290)

(291)

(292)

<S4> everything which is in that (half plane) is stable <S3> stable <S1> but we didn't we didn't understand that the buildings was included and the equipment was included <S2> everything <S1> everything <S> erm , the err what is it called alltså efter instruktionen , jag kommer inte this picture <S> yeah , reorder buffer reorder [buffer] <S> [where is that] , is that in this picture where is the reorder buffer mhm it’s not there are there anything more in life , is there <S> mixing mixing very good point

The final set of features investigated as discursive strategies are ‘pre- and post-dislocation’ and ‘fronting’. The reason for doing so was the fact that, despite the negatively charged term ‘dislocation’, these usages are common in native usage and are typical features of spoken language (MICASE; Mauranen 2007: 253). They have been described as explicitness strategies (Mauranen 2007). They occur frequently in academic speech and are used to topicalize or highlight information both by native and non-native speakers. The present study treated pre- and post-dislocation as explicitness strategies. In the ELFA data, the basic construction is: (Demonstrative+) NP1+ coreferential subject pronoun1 (Mauranen 2007: 254). Although this is not the only construction in the present study, there are many examples of it, especially in the monologic material. (293)–(299) below are examples of regular fronting or heads (also termed ‘topicalization’, in e.g. Meierkord 2006: 24). The topicalized pieces of information are given in bold:

4.2 Dimension 2: Communicativeness

(293) (294) (295) (296) (297) (298) (299)

133

And what n1 is I don’t know. The batch you have to use always 75% of. Or function pipe is it also called. So these three types of crust I mentioned. What is endothermic you already know. The char residue you can recycle and use as energy. Why you can’t get more than 59% I will explain on Wednesday.

(300)–(311) below are examples of pre-dislocation. The reason the following examples under ‘Pre-dislocation’ but not ‘Fronting’ or ‘Heads’ is that the pronoun is present, unlike the case in the examples (294)–(299) above. It should be noted that others have not made this distinction and have termed both constructions as ‘headers’ (Carter and McCarthy 2006: 192– 194). However, I believe that distinguishing the two types might be necessary. My aim here is only to show the difference between the two constructions and not to make any prescriptive claims. (300) (301) (302) (303) (304) (305) (306) (307) (308) (309) (310) (311)

This report we’ll do it later. The composition of the liquid it’s the same. All these chemical reactions they are reversible. The size of the runner and the size of the turbine it is decided by the flow. Because the heat peak you need electricity for that. So the weak combination of things that’s what we don’t want them to happen. And these guys they claim that for example ether which could replace diesel fuel. But these type of uses you can have it anywhere else in Sweden. The reaction turbine it basically works with pressure difference. So biomass it cannot supply alone all fuel and food needs of the ever growing population in the world. Natural gas we can replace it by biogas. So having a fixed structure it doesn’t work in most places.

Pre-dislocations and headers were more frequent than post-dislocations and tails in the material, which is in line with what was reported by Maybaum and Swales (2006). Their investigation on MICASE data showed that only 13 per cent of the dislocations in total in the MICASE material were postdislocations. In the present material, post-dislocations and tails make up

134 4 Operating in a Swedish ELF site about 15 per cent of the monologic examples and 13 per cent of the material in total. This seems to indicate that pre- and post-dislocations are genrerelated, which makes it hard to claim that they constitute an ELF feature. I applied the same principle to the cases of pre-dislocation and headers here: When the pronoun is present, it is termed ‘post-dislocation’, and when it is not present, it is called ‘tails’. The following are some examples of post-dislocations (312)–(315) and tails (316)–(319): (312) (313) (314) (315) (316) (317) (318) (319)

This could be 80 per cent the margin efficiency. Well it is not so emission-free hydropower. So this is very important the design the specific design of the free board. What is the problem when it gets too big the vessel? You have very big parts of it flatland. But you manage much better the float control. Because you are putting at risk the safety. Here you see very clear the natural gas supply.

Although it is not a strategy, topic abandonment as a discourse phenomenon has been included in the present investigation. There were very few cases of topic abandonment in this corpus. The following, (320) and (321) are two examples of topic abandonment. In the first extract taken from a group-work session, S4 and S1 are wondering why S3 does not attend the lectures (lines 1 and 2). S3 answers their questions for a while but then abandons the topic. S/he does not answer S1’s comment “but you’re busy” on line 13, and S4 moves on to talking about another topic on line 14, i.e. what time it is in China:

4.2 Dimension 2: Communicativeness

1 (320) 2 3 4 5 6 7 8 9 10 11 12 13 14

135

<S4> i have (further) part to write at the same time . why you always miss the lecture sorry just curious <S3> yes sometimes i have some other lectures <S4> other lectures there’s some conflict in the timetable <S3> yeah and sometimes <S4> but it’s impossible for all time there’s a conflict for all lecture maybe you don’t want to have lecture you don’t want to attend this <S3> yes sometimes <S1> it’s not interesting to you <S4> @@ <S3> it’s but <S1> but you’re busy <S4> what’s the time now oh it’s from china time

The following, (321), is another extract from a group-work session where two speakers, S4 and S3, start talking about another subject than the content at the end of the group-work. Perhaps the most interesting point in this conversation is that S3 appears to agree that he is from Greece on line 3 although he is from Cuba. He does so most probably because he does not understand what S4 says on lines 1 and 2, and uses “yes” as a filler, in an attempt to let the matter pass. S3 later on in the conversation has to abandon the point because he cannot make a contribution to this conversation on Greece after line 13. The topic dies out after line 18, and S1 moves on to another topic: 1 (321) <S4> (xx) we just talk about there’s some island in Greece 2 you’re from greece 3 <S3> yes 4 <S1> we should prepare a chair for the teacher 5 <S3> @@ 6 <S4> this one this one this one is for the teacher i will stand 7 beside you 8 <S1> i think (xx) other chair 9 <S4> mhm i think (greece) is a country (full of) charming 10 11 <S3> mhm

136

4 Operating in a Swedish ELF site

12 13 14 15 16 17 18 19 20 21

<S4> (full of) charming and that’s why the greece refer to that kind of how to say <S1> greece <S4> greece yes beautiful yes greece <S1> but er i have never been to greece <S4> have been to there he says he recommend to be there (later) <S1> (yesterday) we write the teacher <S4> yesterday <S1> invite the teacher to our group

In her discussion of the Let-it-pass principle, House says ELF interactants first “pretend to understand”, provided that “they are not forced to reveal their non-understanding at some later stage of the talk” (House 2003b: 141). In cases like (321) above where non-understanding does not have serious consequences, ELF interactants can let it pass, but this is very different in discussions of content matters where the stakes are high. Unlike in the two examples above taken from small talk, the subjects generally did not abandon the topic in high-stakes spoken exchange on content. Extract (322) below is an example where three students are working on a group task, calculating the cost of a project. This is a convergent task, i.e. there is only one correct answer. When S2 and S3 think S1 has made an incorrect statement on line 5, they react strongly (lines 6 and 7). S2 and S3 try to explain what they mean to S1 (lines 10–13), but S1 does not seem to be convinced (line 14). S2 and S3 give further explanation (lines 15–17). Nevertheless, S1 pursues what he said earlier (line 18). S3 finally code-switches to Swedish in an effort to convey to S1 that s/he is wrong (lines 22 and 23): 1 (322) 2 3 4 5 6 7 8 9 10 11

<S1> but does this include material <S2> no , material , you’re doing that <S1> yeah but i mean for reference <S2> no <S1> i think it include(d) in here <S2> nej [no, no no no] <S3> [no, no no no] no no nooo <S1> but why should they include everything when , without <S3> because material <S2> it’s it’s to each process

4.2 Dimension 2: Communicativeness

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

137

<S3> yeah <S2> you’ve been doing <S1> yeah but no, this is for <S3> yeah yeah but but it depends <S2> no it’s not in this case <S3> no <S1> no it’s for <S3> yeah but this is not material that’s what you’ve been calculating . come on <S1> this is for <S3> i know this is why we det var därför vi har gjort det här materialkostnaden [That’s why we have calculated the material cost] <S1> yeah but but because

So in the example above, speakers do not abandon the topic, most likely because there would be consequences of this kind of abandonment for the project work, which would show in the calculations. Clearly, the members of the project team need to be in agreement as to what and how to calculate. Overall, in the present study, not all speakers employed a wide range of pragmatic strategies. While speakers in group-work sessions very frequently made use of these strategies, the lecturers had fewer instances of such usage in their speech. Since these differences were observable in the analyses, a decision was made to investigate the differences between the two speech event types in terms of pragmatic strategies. An investigation of the two speech event types in the eight speech events transcribed earlier revealed some important differences. Pragmatic strategies were abundant in group-work while the situation was different for lectures (this section, Table 4.9). The figures provided in Table 4.9 show that the group-work sessions investigated are richer than lectures with regard to pragmatic strategies. The absence of some linguistic phenomena in lectures is due to the nature of lectures as speech events, such as backchanneling. Nevertheless, the overall numbers are higher in group-work. I will discuss these interesting differences between these two spoken genres in Chapter 5.

138 4 Operating in a Swedish ELF site 4.3. Perceived communicativeness and attitudes: A survey of student attitudes In this section, we will look at the results of the attitude survey. Table 4.10 shows all features that met the criteria and were therefore commonalities in the present material and one hundred respondents’ reactions toward these non-standard features (Chapter 3). Since my aim here was to see what perceived communicativeness was like among the respondents and what degree of irritation there was toward each non-standard morphosyntactic commonality, I divided the table into two main sections: Communicativeness and Attitudes. The columns ‘Perfectly OK’ and ‘Comprehensible but wrong’ have been merged in the ‘Communicativeness’ results. Similarly, the columns ‘Not irritating at all’ and ‘Somewhat irritating’ have been merged under ‘Attitudes’. This was done in case the respondents had been steered by the wording of the questionnaire towards more irritation in the experiment. Table 4.10. The results of the attitude survey in percentages.

No irritation

Very irritating

Attitudes

Incomprehensible

Word order Question formulation Countability/Plural forms Passive voice Negation Subject-verb agreement Analytic comparative Articles NonS word formulation Tense usage Double comparatives/ superlatives Not marking the plural

Communicativeness Comprehensible

Feature

71 83 80 80 89 73 80 76 85 81 97 88

29 17 20 20 11 27 20 24 15 19 3 12

72 77 80 81 81 85 86 87 88 90 91 91

28 23 20 19 18 15 14 13 12 10 9 9

4.3 The survey 139

The figures show that category ‘very irritating’ is not a frequent answer. However, it seems to indicate a general pattern of irritation. Table 4.10 has been sorted according to what the respondents thought was very irritating, from the highest to the lowest figure. It is interesting that non-standard question formulation, which is the only feature that seems to disturb communication, is also high up on the irritation scale. The only feature that is more irritating than non-standard question formulation is word order. Both these categories are actually concerned with the order of constituents in a sentence. Respondents who express their irritation towards the divergent forms reported here had interesting quotes in the space provided on the questionnaire sheet “I feel my comprehension of all these incorrect sentences is problem-free but it is quite annoying with such basic grammar mistakes and I hope I won’t have lecturers with this poor English grammar”. Some were quite critical of their teachers, criticizing their general proficiency of English although there were no remarks on the survey on whether the non-standard forms had been produced by teachers or students. Some complained about their teachers’ English: “The teachers’ English isn’t that good, leading to difficulties listening to what they are saying, rather than how they are saying it. (the student’s own underlining)”, ”I wish those giving lectures in English would go through some sort of training before taking on something like this.” and ”Understanding bad English is challenging.”. For others, the subject matter was the only important point, and language was not an issue. Some of the respondents thought what they had heard was perfectly natural and it mirrored the type of speech they were exposed to frequently: “I do not understand what this (this task) is about this sounds like any professor talking that is deep in thought. So as long as you understand it is ok.” And some pointed to the importance of subject matter instead: “I just care about the course itself, benefits that I get depends on the topics of the course.” Undoubtedly, this survey differs from the rest of the study in terms of data where only naturally-occurring speech has been analyzed. It is important to note that in an experimental situation, there is certainly more focus on form than there would be in a real-life situation. In this sense, this part of the present research project did not investigate attitudes towards authentic naturally-occurring speech. However, it investigated the reactions to the forms produced in naturally-occurring speech in the same setting, which brings it much closer to authenticity than perhaps would be possible

140 4 Operating in a Swedish ELF site in a survey. The results show that there is some irritation towards nonstandard forms. Figure 4.23 has been arranged according to the band ‘incomprehensible’, from the feature that was reported as the least incomprehensible to the one that is the most incomprehensible. The graph, along with all the graphs in this section, was prepared using raw figures.

##!" #!!" '&!" '%!" '$!" '#!" '!!" &!" %!" $!" #!" !"

()*+),-./"01"

2*345"67-"01" 84,39:*);)4<=6.)"

Figure 4.23. The distribution of perceived communicativeness of the nonstandard morphosyntactic commonalities.

The results show what students say is comprehensible or incomprehensible, discussed in Chapter 3. As can be seen from Figure 4.23, the highest features on the incomprehensibility scale were subject-verb agreement, article usage and word order. The rest, from highly ‘incomprehensible’ to perfectly OK were pre- and post-dislocation, passive voice, tense usage, nonstandard word formulation, issues related to countability and plural forms, analytic comparative, not marking the plural on the noun, non-standard question formulation, negation and double comparative and superlatives. The following figure, Figure 4.24, shows the reported results on irritation, arranged from the least irritating to the most irritating. As the figure demonstrates, the highest reported irritation was caused by word order, passive voice and question formulation while the lowest irritation scores

4.3 The survey 141

belonged to double comparative and superlatives, the analytic comparative and not marking the plural. The significance of these responses will be discussed in Chapter 5. >3-"=**=-?-=45"

@"6=-"=**=-?-=45"

A)*/"=**=-?-=45"

'&!" '%!" '$!" '#!" '!!" &!" %!" $!" #!" !"

Figure 4.24. The distribution of irritation towards non-standard morphosyntactic commonalities.

4.4. Summary of results Let us now briefly summarize the results of the study by revisiting the research questions. The following are the research questions which the present investigation had as its starting point (section 3.2), followed by the answers based on the results of the analyses: (i) What, if any, are the morphosyntactic commonalities of non-standard usage examined? The results of the present investigation showed commonalities of morphosyntactic non-standard usage across the two different speech event types. So the first hypothesis that there would be overall commonalities in the material was confirmed.

142 4 Operating in a Swedish ELF site (ii) Which of the commonalities found are the same as those described in the literature? The results are in general in line with the features described in the literature. However, there are some differences between what has been described in the literature earlier and what the results here maintain as commonalities. Not all the commonalities reported in the literature were found in the present study. This indicates that the hypothesis that the commonalities found here would be similar to the ones in the literature was only partially confirmed. First, there are features that were discarded here that the earlier literature reported on, e.g. prepositions. The present study excluded prepositions from the analyses for reasons discussed earlier (Chapter 3). Secondly, there are features that were not sufficiently frequent to be included in the list of commonalities, e.g. invariant question tags, redundant prepositions and the relative pronouns which and who used interchangeably, i.e. which used for people and who for non-living things. Finally, there are features found in the current study that met the criteria and proved to be commonalities but had not been reported on in the literature earlier, i.e. non-standard question formulation and unraised negation. See Table 4.11 for a comparison of the features presented by Seidlhofer as hypotheses of usage (2004: 220) and the ones reported here:

Table 4.11. A comparison of morphosyntactic features described in the earlier literature and features found in the current study. NA stands for ‘Not Applicable’ since these features were not included in the investigation. Y stands for ‘Yes’ and N for ‘No’. Features from earlier literature (Seidlhofer 2004: 220)

Features found in this setting

Dropping the third person -s

Y

The interchangeable use of the relative pronouns which and who Invariable tag questions, e.g. isn’t it? No? Higher frequency of some verbs of “high semantic generality” e.g. do, have, make

N

Some cases, NA NA

4.3 The survey 143 Features from earlier literature (Seidlhofer 2004: 220) Increased explicitness, e.g. black color

—

Features found in this setting Some cases. Increased explicitness, expressed through: 1) Heads and Tails, Preand Post- dislocation13 2) Unraised negative. 3) Repetition Non-standard question formulation

The results in the present study were included on the basis of a strict set of criteria. It is uncertain what kind of criteria were applied to the commonalities reported in previous literature. I took Seidlhofer’s list of “regularities that point to at least some hypotheses” (Seidlhofer 2004: 220) here as a point of comparison. Seidlhofer maintained herself that the theses and seminar projects conducted on VOICE data “have brought to light certain regularities that at least point to some hypotheses, which in turn are proving useful for formulating more focused research questions” (Seidlhofer 2004: 220). Since then, several studies on ELF usage have been carried out, showing commonalities of different types (Chapter 2). It is argued here that the differences between the findings in earlier literature and the present study are likely to have been caused by methodological differences. The present study applied a strict set of criteria to all the features found in the material, resulting in a list of features as candidates of commonalities. Overall, when the results on form are considered, the features found in the present study could be subsumed under the following categories in terms of their communicative effectiveness: (1) Non-standard usage that leads to overt disturbance in communication, i.e. non-standard question formulation (2) Successful reductions of redundancy, e.g. not marking the plural on the noun and (3) Devices that increase comprehensibility and create extra explicitness, e.g. unraised negative, pre- and post-dislocation, heads and tails, and finally (4) Other non-standard features that do not interfere with communication. Categories (2) and (3) seem to be strategies

13. Mauranen’s 2007 paper included Heads and Tails (Mauranen 2007: 254), and several ELF studies have included repetition as an explicitness strategy, e.g. Mauranen 2006b; Lichtkoppler 2007; Cogo 2009, as discussed in Chapter 2 here.

144 4 Operating in a Swedish ELF site speakers in lingua franca settings employ to get the message across. In this respect, transparency seems to be a characteristic of usage of English in settings where it is used as a lingua franca. The form features seem to be based on functionality. In other words, those features that do serve a function seem to be preferred by speakers in lingua franca settings whereas those that are redundant are dropped. A feature is kept if it does not interfere with communication, if it aids communication and is functional. (iii) What kind of morphosyntactic non-standard usage results in overt disturbance in spoken ELF communication? The only overt disturbance caused by non-standard morphosyntax was in fact caused by non-standard syntax, namely non-standard question formulation. This type of usage created overt disturbance that could be observed by the parties involved. However, the results also demonstrate that it might be a combination of factors, i.e. question intonation, the presence or absence of the interrogative adverb/pronoun and syntax, that causes overt disturbance and not solely syntax. The results here indicate that the listener needs more than one cue to register a question, confirming the hypothesis that it would be more likely that the combination of different types of nonstandardness would create some opacity in communication. (iv) Which of the pragmatic strategies described in the literature are found in this setting? The discourse and pragmatic strategies investigated here were clarification of terms and concepts, clarification of details and content of task, comment on discourse structure, discourse context, intent and common ground, and backchanneling, repetition (other) and topic abandonment. As I touched upon earlier, topic abandonment was not regarded as a strategy but a phenomenon of interest in the communication process. The lectures and student group-work revealed differences in the use of these strategies. As hypothesized, the group-work sessions were much richer than the monologic lectures. The backchanneling and ‘other repetition’ in the lectures were limited to the very few questions raised, as is generally the case in monologic speech. An interesting finding was that there was no topic abandonment in either lectures or group-work with the exception of topic abandonment in social talk (section 4.2.2). The fact that topic abandonment was found in social talk only and not in any discussion of the content signals the

4.3 The survey 145

high-stakes nature of the speech events in the present material. In addition, there were heads (fronting) and tails and pre- and post-dislocation, and I have argued here that they serve as discourse strategies speakers use in order to achieve extra explicitness. There were similarities between the discourse and pragmatic strategies mentioned in the literature and the ones found here (section 4.2.2). At the same time, there are some differences due to the nature of the situation. One of the studies that investigated characteristics of ELF discourse prior to the present study looked at data from recordings done at the European Centre for Modern Languages, from group discussions concerned with European language education (Penz 2008). These group-discussions were carried out by people who were probably not very familiar with each other or the practices from each other’s settings, which indicates the absence of a strong common ground with the exception of the common aim they had for the task. The current investigation is different by its very nature. Both speech event types in the present investigation have a relatively strong common ground shared by the participants. In lectures, students have access to the course book, lecture notes if any and information regarding the course, and lecturers are familiar with the subject matter. In group-work sessions, students together have sufficient contextual information. All the students involved had the same frame of reference, had received instructions on the task or the expected outcome of the group-work, knew about the task they were working on to a lesser or greater extent and had course books. There are more pragmatic strategies in the current material based on common ground compared to earlier studies and fewer strategies on clarification of terms and concepts. Apart from these differences, the pragmatic strategies reported in the literature and the ones here were similar, confirming the hypothesis. (v) What kind of morphosyntactic non-standard usage is perceived as irritating by speakers in ELF situations? The results of the survey on perceived communicativeness and irritation revealed the types of non-standard morphosyntactic commonalities that were perceived as ‘perfectly ok’, ‘comprehensible but wrong’ and ‘incomprehensible’ along with ‘not irritating at all’, ‘a bit irritating’ and ‘very irritating’. The features that the respondents rated as highest on the incomprehensibility scale were subject-verb agreement issues, article usage and word order. The features that scored the highest on the irritation scale were

146 4 Operating in a Swedish ELF site word order, passive voice and question formulation. These results show that the hypothesis was confirmed due to reported irritation towards some non-standard forms. The results here are overall in line with the findings of earlier ELF research in the sense that there is a considerable number of cases of nonnative-like usage but very few cases of overt disturbance in communication based on morphosyntax. The next chapter, Chapter 5, will include a detailed discussion of these findings.

Chapter 5 Theoretical and practical implications

5.1. Discussion of the findings In the previous chapter, I reported the results of the study in each dimension, namely, form, communicativeness and perceived communicativeness and attitudes. The findings presented in the previous chapter have a number of theoretical and practical implications for different parties, which I will discuss in the present chapter. However, before doing so, a prerequisite is a discussion of the results pertaining to the three dimensions of the study, reported in Chapter 4 previously. I will undertake this task here, starting with the form dimension. Form is one of the main areas that should be investigated when attempting to describe and analyze language usage. Correct understanding of language requires a certain level of knowledge of form. In a discussion on variation in form, the background of the subjects must be taken into account. The subjects in the present study come from a large variety of first language backgrounds from a range of different language types, as is the case in lingua franca settings by definition (Table 3.1). Another factor has to do with the admission of students to the present setting. Since admission requirements traditionally consider central exam scores only (see the discussion on this later on in this chapter, section 5.2), and since these exams do not all include the testing of certain productive spoken skills, there is little or no information on students’ speaking skills. Due partly to this factor and partly to the wide range of first language backgrounds, the subjects have varying degrees of oral proficiency, as in most lingua franca settings. In such a diverse setting, one would expect certain deviance from standard forms. As I mentioned earlier here (Chapter 2), variability has been regarded as a characteristic of the usage of English as a lingua franca (Jenkins 2000). The deviant usages would be described as learner language by second language acquisition researchers. It is a fact that the lists of learner language features, features of World Englishes and the features found in the present study have some overlap. A quick run-through of five different sources on the features of World Englishes (Crystal 1995; Mesthrie and Bhatt 2008; Mesthrie et al. 2001), creoles (Crystal 1995; Mesthrie and

148 5 Theoretical and practical implications Bhatt, 2008; Mesthrie et al. 2001) and learner language (Hultfors 1986; Swan 1980) reveals forms that cut across these groups. See Table 5.1. Table 5.1. The types of morphosyntactic variation found in the present setting, in World Englishes features, creoles and learner language. Y stands for ‘Yes’ and indicates the presence of that particular feature. This ELF setting NonS question formulation Pre- , Postdislocations/ Heads and Tails

World Englishes –

Creoles

Y (Mesthrie and Bhatt 2008: 81–82; Mesthrie et al. 2001: 307) Y (Mesthrie and Bhatt 2008: 85) –

Y

Y (Swan 1980:498)

Y

–

–

–

Not marking the plural

Y

–

Subject-verb agreement

Y

NonS analytic comparative

(Mesthrie and Bhatt 2008: 85)

Y (Crystal 1995: 347; Mesthrie and Bhatt 2008: 52) Y (Crystal 1995: 347; Mesthrie and Bhatt 2008: 65–66) –

NonS word forms

Y

–

Tense and aspect issues

Y

Y (Mesthrie and Bhatt 2008: 60)

Y (Hultfors 1986: 148) Y (Swan 1980: 414–430, 457– 463; Hultfors 1986: 96)

Double comparative/ superlatives Unraised negation

–

Learner language –

Y

–

5.1 Discussion 149 This ELF setting Passive voice

World Englishes Y

Creoles

Word order

Y

NonS Plural forms/ Countability

Y

Y /N Question word order not included. (Mesthrie and Bhatt 2008: 78–81) Y (Mesthrie and Bhatt 2008: 52)

Article usage

Y

Y

Y

Learner language Y (Swan 1980: 407–414) Y/N Question word order not included. (Hultfors 1986: 139) Y (Swan 1980: 137,138; Hultfors 1986: 59) Y (Hultfors 1986: 42–56)

Table 5.1 compares the findings in the present study that met the criteria and were therefore commonalities with the features of World Englishes, creoles and learner language as reported in four sources. What is of interest are the features that run across the features found in the present study, in World Englishes, creoles and learner language. With regard to devices that increase comprehensibility and create explicitness, the present study has pre- and post-dislocations/heads and tails. This is a feature that has been included in World Englishes, creoles and learner language. Another example of usage that creates extra explicitness is the double comparative and superlatives, which has been covered in World Englishes and creoles. The last feature in this category is unraised negation. The sources that were referred to here did not cover unraised negation. As regards reductions of redundancy, we see ‘not marking the plural on the noun’, ‘subject-verb agreement issues’ and the ‘non-standard analytic comparative’ in the present study. ‘Not marking the plural on the noun’ has been covered in World Englishes and creoles, but not in the learner language source consulted here. ‘Subject-verb agreement issues’ are present in all the categories. ‘Non-standard analytic comparative’ is included in

150 5 Theoretical and practical implications World Englishes in Mesthrie and Bhatt’s work (2008: 35), but not in the sources for creoles and learner language. The final category is about other non-standard features that are found in the present investigation, namely, ‘non-standard word forms’, ‘tense and aspect issues’, ‘passive voice’, ‘word order issues’, ‘non-standard plural forms/countability’ and ‘article usage’. All these non-standard forms are present in the sources consulted for World Englishes, creoles and learner language, with the exception of non-standard word forms in creoles in the source consulted. In addition, question word order is not included in the sources for creoles or learner language. The information given in Table 5.1 is by no means exhaustive, and it is not a review of a large number of sources. My aim here was not to outline all the features that run across all the reported features of World Englishes, creoles and learner language, which would be beyond the scope of the present study. Such an analysis would need to be much more detailed, covering several sources and would need to touch upon issues such as language contact, variation and change, linguistic norms and their acceptability, ownership of the language and expression of social identities (Seidlhofer 2009a: 236). I took the features in this study as the basis and checked them against the features found by other sources in World Englishes, creoles and learner language. Although this is a relatively quick run-through of few sources, the overlaps that surface are interesting. The reason for this overlap is not that these language forms are all the same, but that the overlap is due to the demands of functional communication. A close analysis of the types of non-standard usage in the present setting reveals more than merely non-standardness; there are tendencies such as reducing redundancy and increasing explicitness. The results of the present study in the form dimension are in compliance with the ‘functional hypothesis’ and the Prague school of the ‘functional sentence perspective. If a feature shoulders an important role in communication, i.e. if it is communicatively dynamic, it seems to be retained whereas in the opposite case, it is omitted. It must be true that the diachronic source of these features is learner language, but a closer look at them is crucial since lingua franca settings are by nature not language-learning or language-teaching settings. This review also answers the question whether ELF usage is sui generis, a question raised in a number of studies (e.g. Firth 2009b). The features that run across World Englishes, learner language and ELF settings show clearly that ELF usage is formally not sui generis since none of these features are unique to usage of English as a lingua franca. If ELF were sui

5.1 Discussion 151

generis (a detailed discussion of this will follow in section 5.1), there would be very large numbers of unique and consistent non-standard usage. This is not the case in this setting. Apart from a few studies, there has been little reported on this in ELF research. Let us now discuss the findings of the communicativeness dimension of the study. One of the most interesting findings here is overt disturbance caused by non-standard question formulation. It is crucial to first touch upon the main differences between questions and assertions in an attempt to reach an understanding of the communicative difficulties non-standard question formulation seems to create. Questions and assertions fulfill different functions. In the case of an assertion, at the simplest level, the speaker states his knowledge of something and wishes to convey this piece of information to the listener. In questions, the speaker is in need of a piece of information, and his/her main goal is to fill the information gap by obtaining that particular piece of information. So functionally, questions and assertions have different relations to the speaker (Firbas 1976: 12). In the present study, I treated non-standard questions together with all the other morphosyntactic features in its earlier stages but dealt with them separately in the results (Chapter 4). Because they fulfill the important function of asking for essential information or providing the listener with the information s/he needs, a main function in spoken academic discourse, the comprehension of questions is much more critical than that of assertions. Although it has not been possible to report statistical results here, this suggestion can be safely made, considering the results of the present study. Questions seem critical for the further development of communication in this ELF setting. The first part of the present study focused on non-standard morphosyntax, and to continue in the same fashion, I will discuss the syntax of questions before the role intonation plays on the registering of a question. The non-standard questions in the corpus compiled for the present study, as explained before, were categorized traditionally as Wh-questions and Yes/No questions. Let us first consider Wh-questions: As discussed in Chapter 4 briefly, it might be unexpected that the Whquestions did not elicit the preferred response despite the presence of the interrogative adverb/pronoun. There could be two different explanations for this: First, it may be that the interrogative adverb/pronoun goes unnoticed. It was argued more than three decades ago in the literature that it is predominantly the interrogative adverb/pronoun that fulfills the function of express-

152 5 Theoretical and practical implications ing the need for information (Firbas 1976: 12). In this respect, if the interrogative adverb/pronoun goes unnoticed and the most critical element in the question is absent, it is highly likely that the question is not registered. The short duration of the interrogative adverb/pronoun compared to the total duration of a question (Chapter 4, Table 4.5), the fact that the interrogative adverbs/pronouns took up about an average of 14 per cent of the questions studied here, shows that it may well have been that the listeners missed them. Especially when combined with background noise, which is quite usual in student group-work, initial words in an utterance may go unheard. In addition, if there is no preparedness for a question in discourse, there is an increased risk for missing the beginning of an utterance. However, Firbas’s view is not undebated. An alternative explanation could be that interrogative adverbs/pronouns have little semantic load and it is the rest of the utterance that is crucial for the smooth flow of communication. It has been generally accepted that it is the final position in an utterance that carries heavier semantic load (Giora 1985, 1988). So it may be that speakers focus less on the first parts of an utterance and more on the parts that come later, which plays down interrogative adverbs/pronouns. It is necessary to refer to the Prague-school concept of functional sentence perspective here. In the functional sentence perspective, all elements contribute to communication, but they do so in different degrees. Some elements are more dynamic than others in communication, and this notion is termed ‘communicative dynamism’ (Firbas 1992: 7). Communicative dynamism is regarded “an inherent quality of communication”, and it “manifests itself in constant development towards the attainment of a communicative goal; in other words, towards the fulfillment of a communicative purpose” (Firbas 1992: 7). So, in this school of thought, every element plays a role in this communicative goal and shows different degrees of dynamism. In the light of this information, I will argue that there might be a recency effect in these questions that makes the later part of the questions more critical for understanding than the interrogative adverb/pronoun, which seems to further suggest that the presence of the interrogative adverb/pronoun does not necessarily aid communication. So, it would be as likely for a non-standard Wh-question to go unregistered as it is for a nonstandard Yes/No question. Since the present study is based primarily on naturally-occurring communication and not speech events set up for research purposes, there were no controlled experimental situations. There-

5.1 Discussion 153

fore, it was not possible to measure precisely which elements in a question caused overt disturbance. As mentioned in Chapter 4, there appear to be three cues that help the listener register a Wh-question: The interrogative adverb/pronoun, word order and intonation. The ideal case for the listener would be to be provided with all these three cues so the risk for disturbance is decreased. So the importance of the interrogative adverbs/pronouns is not disregarded here. They are one of the cues present for Wh-questions, but the probability that they may go unnoticed or unheard is taken into consideration. The second cue for the registering of Wh-questions is syntax. This second cue was not available for the listeners since only questions that had been formulated with non-standard syntax were analyzed. There were different types of syntactic non-standardness in the material. The types of non-standard syntax in Wh-questions found in the material are missing ‘there’ (107), missing ‘it’ (109), missing copula (110), issues related to the usage of ‘do’ (113, 118, 121, 124) and issues related to word order (115, 116, 117, 120, 122) (Chapter 4, section 4.2.1). Finally, there is the third cue, which involves intonation. Although in English, both Yes/No and Wh-questions have been reported to have falling intonation almost as often as they rise (Bolinger 1998: 50), the addition of a third cue is most likely to help listeners register the question. The analyses here maintain that rising question intonation is indeed an important cue, and it is best to try and provide the listener with this cue in lingua franca settings, regardless of what the practice is in native varieties of English. In the case of Yes/No questions, the types of non-standard syntactic features are declarative word order (126, 130, 132, 134, 136), issues related to reported speech (127) and missing verbs (131). Syntax is only one of the cues that may be available to the listener. While the listener has three cues in Wh-questions, there are only two cues in Yes/No questions: Syntax and intonation. Some studies with their focus on teaching situations have suggested that standard syntax alone suffices for a Yes/No question to be registered regardless of intonation: The intonation of yes/no questions in English language is not an essential issue for most learners. Although yes/no questions in English language teaching are taught primarily in terms of intonation, it is clear that, whatever the intonation, a fully-formed yes/no question remains a question and will be interpreted as such. Thus, it appears that the grammatical form and communicative use of yes/no questions, which are de-

154 5 Theoretical and practical implications emphasized or ignored in most textbooks, are likely far more important for learners of English. It is not unusual for even advanced students often to have trouble with both form and use. Inaccurate formation can mask the fact that a question is being asked, often so much that listeners are not sure how to respond. Lack of inversion and missing words or auxiliaries can short-circuit a question far more effectively than lack of rising intonation. (Levis 1999: 377) [!] The results indicate that intonation of yes/no questions should be an unimportant issue in English language teaching because intonation of these questions appear to play a minimal role in the success of interactions between speakers from different varieties of English (Levis 1999: 373).

It is undoubtedly true that syntax is important in the registering of a Yes/No question. In fact, in Yes/No questions, there are both auxiliary addition and inversion (Firbas 1976: 18), acting as two syntactic cues for the listener. However, Levis’s argument is based on differences of intonation patterns in Yes/No questions in British and American English. Lingua franca settings are highly diverse with a wide range of morphosyntactic non-standardness and other complexities, and any other aid to help speakers become effective communicators is of significance. What seems to be the point made by Levis is that the differences between the British and American varieties of English make it pointless to teach intonation to learners of English. He suggests that one should rely on the stable syntax instead of question intonation, which varies in different native varieties of English. This may be true for some learners of English, but Levis’s findings do not necessarily apply to lingua franca settings. In ELF settings, syntax ceases to be a stable element with all its added complexities, and the universal question intonation remains as a more stable and therefore more reliable one. Even regardless of the setting, the speaker who asks the question must make his angle14 clear to the listener, and this is achieved by intonation. Otherwise, the angle will be unknown to the listener (Firbas 1976: 13). Recipients simply go by the signals given to them in a sentence, and mean-

14. Firbas uses the term ‘angle’ to refer to the differing values of importance in a question. This piece of knowledge is not necessarily shared by the speaker and the listener, and therefore, the speaker must make the important words in a question clear to the listener by making use of intonation (Firbas 1976: 13).

5.1. Discussion 155

ing is traditionally conveyed by means of a combination of prosodic and non-prosodic features (Firbas 1992: 219). It has been claimed that in Yes/No questions, if there is non-standard syntax, e.g. lack of inversion or the auxiliary ‘do’, this needs to be compensated for by question intonation (Firbas 1976: 18). Although it has also been argued that intonation cannot operate on its own independent of other elements (Firbas 1992: 219), we know that in new Englishes, e.g. in South African English, a question can be signaled by only intonation rather than a change in word order (Mesthrie et al. 2001: 313). The results of another study show that speakers of Singapore English prefer SVO word order when it comes to syntax, and they signal interrogativity by changes in intonation and using questions tags (Williams 1990: 159). The findings here reveal that intonation is equally important and therefore should be factored into any analysis of questions. The conclusion for the Yes/No questions in the present material is the same as that for Whquestions. It seems best to try to provide the listener with all cues available, but if the syntax is non-standard, intonation can shoulder the task of conveying the question to the listener on its own. Let us now turn to the results of the discourse level. The findings of the analyses at the discourse level clearly show that speakers in ELF settings tend to use a wide range of pragmatic strategies and comments to convey the message to the listener. There is empirical evidence from ELF research that speakers in ELF settings make frequent use of such strategies and thereby ensure communicative effectiveness despite the at least equally frequent non-standardness in the form dimension (Cogo 2009; Kaur 2009; Mauranen 2006b). In this respect, pragmatic strategies can be presumed to play a profound role in the communication process and in preventing misunderstanding since significant communicative efforts are made through these strategies. Insofar as it can be ascertained, such proactive work prevents a large part of the potential disturbance in group-work ELF interactions. It is through these strategies that speakers in ELF settings achieve negotiation of meaning and communicate a message successfully, which is supported by evidence in ELF literature. Kaur reports in her study that despite the fact that there are frequent ungrammaticalities in the data “the lack of communication breakdown is remarkable for it reflects on the participants’ success in compensating their lack of linguistic competence in the language with increased use of interactional practices to bring about shared understanding” (Kaur 2009: 119).

156 5 Theoretical and practical implications It is hard to determine to what extent it is the pragmatic strategies that helped along the communication process in the present setting and to what extent non-standard morphosyntax is simply insignificant. It is highly likely, however, that it is the frequent use of these strategies that aid communicative effectiveness in group-work sessions and prevent potential disturbance despite the higher number of non-standard features in group-work compared to lectures. As discussed in earlier sections (sections 3.3 and 4.2), pragmatic signaling plays an important role in the communication process and in preventing misunderstanding since speakers do significant work through these strategies (Mauranen 2006b: 131). It is very possible that it is the frequent use of these strategies that ensures communicative effectiveness in group-work and prevents disturbance despite the higher number of non-standard features in group-work compared to lectures. Lecturers seem to employ fewer pragmatic strategies in lectures than students do in group-work when overall results are considered. This difference is not unexpected. Lectures are traditionally made up of long stretches of speech, which does not allow for opportunities for pragmatic speech (Flowerdew 1994: 11). In this regard, lectures do not have the same affordance group-work sessions have. They are monologic speech events, and it is only up to the lecturer to make use of these strategies. Groupwork, on the other hand, is by its very nature interactive with many instances of turn-taking and plenty of opportunities for signaling understanding or disturbance and negotiating meaning. This is one of the main differences between monologic and dialogic speech. So, naturally, there is little or no backchanneling in a lecture since there is traditionally no backchanneling in monologues unless prompted by questions. There is also little or no ‘other repetition’ in lectures since there is little or no ‘other talk’. Even when these two features were excluded from the present investigation, the group-work sessions here would have more instances of pragmatic strategies than the lectures. The differences in the number of pragmatic strategies between these two types of speech events, however, cannot solely be attributed to the speakers, which indicates that the frequent occurrence of pragmatic strategies in dialogic speech seems to be a genre feature. In the speech events investigated for the present study, the two groups of speakers had almost equal numbers of comment on discourse structure and discourse content and repetition for emphasis purposes. Some of the differences between the lecturers and students concerned backchanneling and self-repair, of which there were more cases in the group-work sessions. Some of the features are naturally genre features, i.e. features that are tradi-

5.1. Discussion 157

tionally not found in monologic speech, e.g. ‘other repetition’. There were also few cases of backchanneling in lectures. While the lack of backchanneling could be assigned to the nature of the speech event here, the lecturer’s use of questions and his/her preference for interaction is equally decisive here. A lecturer can create opportunities for student interaction, e.g. by asking questions, which in turn would create an equal number of opportunities for backchanneling. Overall, however, even when the features that are typically not associated with monologic speech, e.g. other repetition, are excluded from the present investigation, the group-work sessions appear to have more instances of pragmatic strategies than the lectures. Students had more cases of ‘comment on intent’, ‘comment on common ground’ and ‘repetition for emphasis’. This suggests that the lecturers in the present study paid less attention to pragmatic strategies than the students. It is especially interesting that the students made equally frequent use of repetition for emphasis as the lecturers, signaling their efforts to convey the message. Perhaps more was at stake for them, as they needed to complete the task in the group-work projects. The results also seem to indicate that speakers with varying degrees of proficiency and fluency are perfectly capable of making use of pragmatic strategies. Although the speakers in the present study had morphosyntactic non-standardness in their speech and disfluencies, they were able to make frequent use of pragmatic strategies. This supports House’s notion of pragmatic fluency and how it is not necessarily dependent on the speaker’s proficiency (House 1996). Management of the form dimension of speech was not decisive for one’s ability to make use of pragmatic strategies in the present study, unlike what has been suggested in some studies (Pérez and Macia 2002). Despite non-standardness of form, speakers in ELF settings can, without much trouble, negotiate meaning and check their understanding through these pragmatic strategies. So, irrespective of whether the speech event is monologic or dialogic, it is crucial that speakers in ELF settings create opportunities in spoken communication to deploy pragmatic strategies. This naturally holds true for any high-stakes spoken discourse, but it is argued here that such strategies are of specific importance to speakers in ELF settings due to the added complexities in the form dimension (Björkman 2009). Moreover, it would not be unreasonable to expect the lecturer to pay more attention to conveying the message to his/her audience. This, however, was not the case in the speech events investigated here. Similar results have been reported in an-

158 5 Theoretical and practical implications other study where only 60 per cent of the lecturers paid attention to pragmatic signaling (Fahmy and Bilton 1990). I will now discuss the results of the last section of the study on perceived communicativeness and attitudes. The survey on perceived communicativeness and attitudes, although it was not naturalistic, was a valuable complement to the rest of the study. One of the strengths of the survey lies in the nature of its respondents. The survey aimed to investigate what speakers who speak English as a lingua franca had to say about different types of non-standardness that were observed in the same lingua franca setting. Unlike the irritation and attitude studies of the 70s and 80s, non-native speakers’ attitude to other non-native speakers’ production was investigated. In addition, the present questionnaire was based on the results of the present study and the respondents were from the setting in question, unlike in other studies where respondents from language classes expressed their opinion on non-standardness in lingua franca settings (Kuo 2006: 218). It is likely that a group of people studying English in a country where it is the native language will have as their primary aim acquiring native-like English. The goals will differ considerably for those who use the language mainly for vehicular purposes, i.e. communicative effectiveness when speaking to other non-native speakers in non-English-speaking countries. The focus in the first part of the experimental survey was on perceived communicativeness, i.e. what students reported as comprehensible and incomprehensible. Undoubtedly, in surveys, one is left with what the respondents choose to reveal, but in this case, what the respondents reported as incomprehensible revealed important information on their attitudes. In this sense, what they reported as incomprehensible gave additional information on their attitudes towards non-standardness. The three non-standard features that scored highest on the incomprehensibility scale are subject-verb agreement, article usage and word order. When it comes to reported incomprehensibility on subject-verb agreement issues, it is suggested here that this result is purely due to affective factors since this is a redundant feature and does not play a role in comprehensibility. As the student quotes also show (section 4.3), some of the respondents perceived the non-standard forms as produced by lecturers only and expressed irritation although the survey was not about teachers’ production. Students tend to be critical towards their teachers and expect ideal delivery of the subject matter. They may have thought this type of basic nonstandardness would be unacceptable in a lecturer’s speech.

5.1. Discussion 159

The second highest feature on the incomprehensibility scale is article usage. The absence of articles and extra articles caused reactions among the respondents. The situation in the case of articles is likely to be similar to subject-verb agreement issues, i.e. respondents may have reacted to the basic nature of this type of non-standardness. It is interesting, however, that word order comes third right after subject-verb agreement and article usage, which is quite high on the incomprehensibility scale. This study earlier reported non-standard question formulation as the only feature that caused overt disturbance, however, and one of the elements that caused overt disturbance in questions is non-standard word order. When it comes to low scores on the incomprehensibility scale, the lowest three scores are shared among double comparative and superlatives, negation and question formulation. Double comparative and superlatives and negation are both non-standard types of usage based on explicitness. This appears to suggest that explicitness-based non-standardness was not perceived as incomprehensible. The second part of the survey focused on reported irritation towards non-standard forms. Word order issues were the highest on the irritation scale, followed by passive voice and non-standard question formulation. It is striking that word order and non-standard question formulation issues were among the three highest features on the irritation scale since it is precisely the combination of these two features and intonation that leads to non-standard question formulation, being the only feature that caused overt disturbance. Word order scored the third highest incomprehensible feature in the survey, which gives these results more weight. This indicates that what the respondents reported as incomprehensible were in fact some of the features they were irritated to have been exposed to. Irritation towards a speaker’s production can have a negative effect on communication and should be factored in when investigating the communication process. The lowest reported irritation was caused by double comparative and superlatives, the analytic comparative and not marking the plural on the noun. This strengthens the suggestion made earlier here that explicitnessbased non-standardness was not perceived as irritating or incomprehensible by the respondents. In addition, the consistency between the results regarding overt disturbance and irritation seems to indicate in fact that the questionnaire was effective in measuring students’ overall attitudes. Having said that, the two scales measured the same variable.

160 5 Theoretical and practical implications In the next section, I will discuss the theoretical implications of these findings.

5.2. Theoretical implications From the results of the present study emerge some theoretical implications pertaining to the status of English as a lingua franca, the norms and standards for spoken English and what the notion ‘good English’ appears most strongly associated with in such settings. These notions are central to any discussion of ELF. 5.2.1. The status of ELF While discussing the issues regarding the status of ELF, I will be attempting to answer some of the most-debated questions regarding ELF that have been brought up in several publications and discussions. The first question deals with whether ELF speakers should be considered learners. Whether ELF speakers are learners of the language and whether their production is equivalent to learner language have been topics of much-heated debates. ELF researchers have strongly rejected this view (Jenkins 2006b; Ranta 2009; Seidlhofer 2004) and discussed some of the differences between ELF speakers and learners of English. Mauranen summarizes the view shared by most corpus linguists as follows:

As a corpus linguist, I think it is sad that although corpus study has challenged the supreme status of the native speaker, it has not dropped the ideal of native speaker intuition; it has in effect transferred the description of the native speaker from ideal competence to actual performance. It is nevertheless still firmly the native speaker who holds the reins. On the whole, despite the mixed feelings that ELF has aroused in English linguistics, one thing is shared: The native is the “speaker”, while the non-native is eternally the “learner” (Mauranen 2005: 272).

To be able to answer this question, one needs to consider the notions of situation, code, self-image and intentionality. The attempts to answer this question have included discussions of situation (Firth and Wagner 1997), self-image and intentionality (House 2002a; Jenkins 2006b; Seidlhofer, Breiteneder and Pitzl 2006) and partly, code (Breiteneder 2005; Meierkord

5.2. Theoretical implications 161

2004; Ranta 2006). I will outline the main differences between the two groups in the mentioned order, in the setting investigated here. When it comes to situation, there are large differences between speakers in ELF settings and formal learners of English. The ‘learner’ label all nonnative speakers seem to bear regardless of the situation is not justified. It is argued yet again here that ELF speakers are not formal learners in the traditional sense, because there are important differences between the two groups with regard to the setting, the level of the speakers, the norm, feedback and the way non-standardness is dealt with (see Table 5.2). Table 5.2. Some differences between formal learners of English and speakers in the present ELF setting.

Setting Level Norm Feedback

Dealing with nonstandardness

Other

Speakers in the present ELF setting Authentic communicative situations Most levels present (A2, B1, B2, C1, C2) No overt norm Little (other repair) or no negative feedback in case of nonstandard production Non-standardness that does not interfere with communication or aids communication kept. Little or no consequences in the form dimension (language generally not assessed). Explicitness strategies added for optimal communication

Formal learners of English Classroom situation Homogeneous level Norm presented overtly Negative feedback in case of non-standard production Non-standardness not kept= There are direct consequences in the form dimension (testing: grades etc.)

-

Table 5.2 shows a skeletal view of some of the main differences between learners of English and speakers in the present ELF setting. The differences are in the main areas of setting, level, norm and feedback. In addition, non-standardness is handled differently in speaker and formal learner groups. Finally, there are some behavioral differences by speakers of English as a lingua franca, included in ‘Other differences’ below. These together make up fundamental differences between speakers in ELF set-

162 5 Theoretical and practical implications tings and those who have chosen to take language courses and are, in the traditional sense, learners of English. To start with, ELF settings are truly authentic communicative situations. This is very different for formal learners of English whose only territory of practice is the language classroom, with provided help from the teacher and the textbook. The second issue deals with the level of proficiency of speakers. In ELF settings, speakers come from a range of levels (Jenkins 2000: 73, 74). Due to admission requirements set by the universities, students are almost never within the Common European Framework15 language level A1; however, there are students who are within A2 or B1 in terms of speaking skills. One hears different levels of proficiency in lingua franca settings, ranging from A2 all the way to the highest level, C2. This is also precisely why ELF is not and cannot be referred to as a level of proficiency. Some ELF speakers are highly proficient while others may be communicating with difficulty. So, ELF has “its own cline of proficiency” (Jenkins 2007: 30). The situation is completely different for formal learners of English. Ideally, learners are placed in groups in terms of their levels to make both the teaching and the learning process more effective. It is not uncommon that there is some range of levels in a learner group; however, the range is never as large as that of lingua franca settings. ELF settings are authentic usage settings; therefore, language is not of primary concern. It is the task that has the participants’ main focus. In this sense, no overt norm is imposed on ELF speakers. Naturally, this is very different in language classrooms where the main aims are to become more proficient users of the language and to achieve language complexity. Formal learners are always presented with a norm they are instructed, to which they are expected to conform. As a direct consequence of this, they aim to follow the norm they are prescribed to follow. If they fail to do so, there are consequences for them in their learning situations, such as the grades of achievement tests and negative feedback etc. Speakers in ELF settings, on the other hand, aim for functionality before any other factor such as conforming to what may appear as the norm, accuracy or language complexity. They pay attention to the functional bits of the language that helps them

15. The goal in the Common European Framework is native-like usage for any language including English, which is not in accordance with the use of English as a lingua franca. Therefore, it is not seen as relevant to ELF research where the aim is communicative effectiveness.

5.2. Theoretical implications 163

convey the message and do not pay particular attention to the redundancy of English, as is clear from the results of the present investigation. There are no consequences for them in the language dimension since they do not get language feedback from content teachers unless their production is incomprehensible. In addition, because communicative effectiveness is their main goal, they employ pragmatic strategies frequently to assure that communication takes place. In connection with the situation, one should also mention that learner language is used to refer to an individual’s language development and is inadequate for describing contact languages. In addition, the use of English as a lingua franca is not about language improvement, and ELF users do not necessarily have a continuum before them, unlike learners of English. The sense of the term ‘learner’ here is the traditional sense of the word, i.e. formal learner. Naturally, human beings are receptive to input. So, it is perfectly possible that incidental learning may be taking place in ELF settings, e.g. that speakers may be learning vocabulary from each other. It is also equally likely that they may be learning non-standard forms from each other. The following excerpt from the present corpus is an example of such a case: 1 <S1> the methods , the results of the interviews 2 <S2> the analyses [!n!!l!!z!z] you mean 3 <S1> is that what it’s called analys 4 <S2> yeah analyzes [!n!"l!#z#z] yeah it’s analyzes [!n!"l!#z#z] you’re 5 analyzing the result In the above extract, On line 2, S2 pronounces the word ‘analysis’ as [!n!"l!"z!z]. This is not the way the word is pronounced. S1 asks whether the word he is looking for is what is provided for him by S2 on line 3. When he asks his question, he wants to make sure and provides S2 with the Swedish equivalent of the word he is looking for: ‘analys’. S2 confirms several times on lines 4 and 5 that the word S1 needs is [!n!"l!"z!z]. As a result, S1 might walk away from this conversation, thinking the word he needs to use for the Swedish word ‘analys’ is [!n!"l!"z!z]. In another example, one of the speakers attempts to correct another speaker’s syntax but provides the speaker with another non-standard form: 1 <S1> the (phosphoric) acid 2 <S2> is not rised [!r!"z"d] 3 <S1> is not rise [!r!"z], just that the amount…

164

5 Theoretical and practical implications

On line 3, S2 corrects ‘rised’ and produces ‘rise’. Both forms are nonstandard, so there is no real correction or repair here. This excerpt is another example of speakers providing each other with non-standard forms. The second notion to consider is the code. The data here showed clear overlaps between learner language and ELF code features. In this respect, one could surely suggest that ELF commonalities are like learner language. Code, however, is only one of the notions to consider when attempting to answer the question whether ELF speakers are learners or not. The third notion, the notion of self-image, is an important one. Apart from learning that might result from correction, it is also possible that some speakers in ELF settings have the additional goal of learning more English and speaking more sophisticated English along with getting content education. Especially in Scandinavia, where English is used in a large number of domains, speakers might have this additional aim in mind. In other words, some speakers might see themselves as learners of English. There is evidence for this in the literature. A study that investigated exchange students’ behavior studying in Scandinavian universities showed that language improvement was among the aims of some of the exchange students. For some students, English was a vehicular language while others saw Scandinavia as the second best setting to improve their English (Shaw, Caudery and Petersen 2009). In this sense, it is up to the individual speaker whether s/he is also a learner of English or not. Connected with the notion of self-image is intentionality. Intentional learning and incidental learning are two very different concepts. Naturally, we are surrounded with different types of information at all times. Equally natural is our inability to block information, which results in incidental learning in most cases, i.e. learning indirectly in an unplanned manner as a result of continuous input. Incidental language learning has been documented in different settings, e.g. a business setting where English serves as a lingua franca shows that speakers reject the role of ‘learner’, but incidental learning takes place in the course of communication (Firth 2009a). This issue deserves a much more detailed discussion and a reference to neurolinguistics than space and topic allow here, but in an attempt to briefly explain the difference between incidental and intentional learning, we can say that we pay attention to other things than what is learned in incidental learning whereas in intentional learning, as the word suggests, our consciousness is activated, with our focus on what we are learning. Another important dichotomy is the one between intake and input. What speakers in the present ELF setting are exposed to is content input, i.e. explicit and

5.2. Theoretical implications 165

consciously perceived input of content, but possibly language intake, i.e. implicit intake of language. Is ELF usage learner language then? So far, I have discussed the nature of speakers in ELF settings and aimed to show clearly the differences between learners of English and speakers who use English as a vehicular language in real-life settings. Although it is clear from the discussion here so far that ELF usage is not simply learner language, I will now expand on why this term is not appropriate for ELF situations and speakers. The first reason this term is unsuitable has to do with the very meaning of the term. The term learner language is used to describe an individual’s language proficiency and is temporal since, along with an individual’s language proficiency, comes the continuum of development. ELF situations are typically contact situations. Moreover, as I have mentioned in this section earlier, ELF is not a level of proficiency. Undoubtedly, speakers can get better at the language they need to use on a daily basis. However, negative learning can also take place, as we have seen in the two excerpts in this section, which also should be taken into consideration. These factors alone suffice to explain that the term should be used exclusively for learners. There is evidence also from the form dimension of the present investigation that ELF usage is not learner language. Although there is some overlap between ELF usage and learner language features (Table 5.1), learner language research shows that learners act differently when it comes to language production. It was shown by a study on twelve Chinese learners’ plural marking that informants applied the pluralization rule almost categorically in a limited set of prototypical expressions consisting of numerals or quantifiers modifying plural nouns, with the NP functioning syntactically as an adverbial (Young 1991), and the subjects in general applied plural marking to the nouns regardless of the quantifiers preceding the nouns. Young’s results contradict the functional hypothesis, discussed earlier by Kiparsky (1972), Guy (1981) and Littlewood (1981) among others. The functional hypothesis, originating from creole studies and quantitative sociolinguistics (Young 1991: 117), suggests that a feature is more likely to be kept if it fulfills a function and more likely to be omitted if it carries unimportant information. In the present study, as explained before, many speakers indicate the plural meaning merely by numbers or by adverbs or determiners before the noun but leave the noun itself without an inflexion (section 4.1). This finding contradicts starkly with Young’s results. Young himself explains that the results of his study are contrary to the functional hypothesis he tested, and he suggests this hypothesis be rejected (Young

166 5 Theoretical and practical implications 1991: 155). In ELF settings, functionality plays a major role to the extent that it seems to be decisive of usage whereas in learner language, learners (in the classroom learning situation) have the primary aim of approximating their production to native speaker production. This is a key difference in the code. In the piloting stage of the perceived communicativeness and attitudes survey (sections 3.3 and 4.3), the respondents expressed their views on approximating their spoken usage of English to native speakers’ as follows, with original punctuation and use of capitals: Speaking like a native speaker is not my aim, because

(i)

the content (…) is more important, it is a technical topic. Clear delivery is the aim, not be like native speaker.

(ii)

as long as the speaker is easy to understand and has a good knowledge of English — in international contexts I don’t expect him to sound like a native speaker.

(iii)

the most important thing is to be understood.

(iv)

it is difficult and I don’t think it is necessary as long as people understand what you say.

(v)

I don’t want to sound like something I am not.

(vi)

it’s nice with some variation.

(vii) it’s completely understandable that a non-native speaker doesn’t sound like a native speaker.

(viii) be clear and precise is enough! Native speakers tend to speak too fast for me, and always with some slang, which are hard to follow.

Investigating the question of approximating spoken usage to native speaker production is a separate matter which I have excluded from the main study here; nevertheless, the above quotes suggest that there is definitely some awareness among the speakers in this setting of the importance of functionality in ELF settings. Their primary aim is to get the message across and maintain clarity as opposed to conforming to native speaker standards. They show that they are aware of the nature of international settings (quote ii), by saying that they do not see it necessary for speakers to sound like

5.2. Theoretical implications 167

native speakers in settings that are by nature international. In addition, they seem to reveal other issues, such as identity (quote v) and variability (quote vi) issues. Another important question pertaining to the status of ELF is whether ELF is sui generis or not. To be able to even attempt to answer this muchdebated question, one first needs to see how the term sui generis is defined in the answers provided so far in the literature. There appear to be two senses of the term sui generis in ELF literature: First, a set of features unique to ELF, in other words the dictionary meaning of the term, and ELF usage that deserves to be studied in its own right, which is the way it seems to be used in some studies in ELF literature. Those who use the term in its dictionary meaning seem to refer to the features or strategies observed in ELF settings while those who prefer the second meaning appear to refer to ELF situations. Jenkins, in a paper discussing the differences between ELF and SLA states that ELF is indeed sui generis, and she seems to refer to the second sense of the term as discussed here by saying “The essential point here is that – as empirical research is already demonstrating – ELF is sui generis, then ELF proficiency levels must also be sui generis” (Jenkins 2006a: 141). In the introduction to a special issue on the pragmatics of ELF, it is claimed that ELF is a “phenomenon sui generis” (House 2009: 141), and the meaning of the term is similar to the way Jenkins uses it: This special issue deals with the pragmatics of ELF. It contributes to current conceptualizations and descriptions not the nature of ELF as a phenomenon sui generis. The papers in this issue do not therefore look upon English as a second or a foreign language to be measured against an ideal native speaker norm, but as a special form of language operating under different conditions than both native/native and native/non-native interactions. (House 2009: 141)

What House seems to mean is that ELF does not need to be studied with reference to second language or foreign language studies. The studies in the issue mentioned regard ELF as a special form of language which should be dealt with accordingly. Firth, in the same special issue, answers those who say ELF cannot be sui generis because of its by mentioning variability as one of the characteristics of ELF (Firth 2009b), using the first sense of the term. Variability was mentioned by Jenkins earlier as a characteristic of ELF usage (Jenkins 2000). I will use the term in its dictionary meaning here. The answer I can provide to the question whether ELF is sui generis or not is limited to the form

168 5 Theoretical and practical implications dimension. With regard to form, a comparison is made between the commonalities found in the present study, features of World Englishes, creoles and learner language. In the light of this, it appears that ELF is by no means sui generis. If the features are shared with varieties of English, it cannot be claimed that the commonalities found in ELF usage form a unique set of features. Among the thirteen non-standard features subsumed under the four categories, i.e. non-standardness that causes overt disturbance, devices that increase comprehensibility and create extra explicitness, reductions of redundancy and other non-standard features, ten were shared with World Englishes, nine with creoles and seven with learner language. This large overlap shows without a doubt that ELF usage is not sui generis in the form dimension. ELF literature has suggested that World Englishes and ELF are not disparate paradigms (Cogo 2009: 59). When it comes to the pragmatics of ELF settings, the issue is more complicated. Pragmatics is not a relatively closed system like morphosyntax, which makes it very hard to assess whether the pragmatic features seen in ELF settings are sui generis or not. There is, however, little reason to think that the pragmatic strategies (sections 4.2.2 and 5.1) observed in the present corpus should be absent in other types of high-stakes interaction where speakers strive to get the message across to be able to solve a problem together. In this sense, ELF usage is not sui generis, or unique, if we go back to the dictionary meaning of the term. Uniqueness, I will argue here, comes with being different from the rest, in this case other types of usage. As ELF is the global use of English, it is inclusive, and it is a type of usage that includes, e.g. different L1 backgrounds, levels of proficiency, rather than excludes. Accordingly, ELF usage will naturally have everything in it that the speakers bring to the table, native and non-native, resulting in inclusion rather than exclusion. ELF does not need to be sui generis to receive the attention it deserves to be studied separately from World Englishes, creoles and learner language. In fact, the question seems to be receiving undeserved attention. It is true that ELF has great similarities to the varieties mentioned, but ELF situations are different from the situations in World Englishes, creoles and learner language. For validity and reliability of any research, it should be dealt with separately as a legitimate topic of research. Another important question to address here is whether ELF is a variety or not. The results of the present investigation show clearly that ELF usage is not sui generis, and that variability is a characteristic of ELF usage. In

5.2. Theoretical implications

169

fact, the commonalities in the present study were a minority at all times and in both types of the speech events investigated. These results indicate that ELF is not a variety, if it is the traditional meaning of the term variety to which one intends to refer. It is useful to refer to what is meant by the term variation. Quirk et al. discuss five major types of variation with reference to varieties of English: Region, social group, field of discourse, medium and attitude (1985: 16). The first two types of variation refer to people using a certain variety because of living in a certain area, and they use a certain variety due to belonging to a group of social variation, e.g. variation with reference to education, socioeconomic situation and ethnicity (Quirk et al. 1985: 17). The last three types of variation are of specific interest here since they are defined by language use. In the field of discourse, speakers have a certain set of features available to them that they deploy within their field, be it their profession, their interest or training (Quirk et al. 1985: 23). For the speakers that took part in the present investigation, the field of discourse was engineering. Variation with reference to medium, i.e. speaking and writing, can be equally decisive of variability in one’s usage. Speaking is the natural form of communication and is, needless to say, different from writing in many respects. Finally, attitudinal variation refers to the formal and informal binary distinction in usage. Regardless of the type of variation in usage, a set of solid features, referred to as ‘common core’ (Quirk et al. 1985: 16) is required that cut across the entire variety. A common core or nucleus is present in all the varieties so that, however esoteric a variety may be, it has running through it a set of grammatical and other characteristics that are present in all the others. It is this fact that justifies the application of the name ‘English’ to all the varieties (Quirk et al. 1985: 16). In the sense of the term variety used by Quirk et al. above, ELF usage documented in the present investigation does not fulfill the requirements, for there is great variability of usage and the ratio of non-standardness to standard features is relatively low. Based on the results of the present study, ELF cannot be referred to as a separate variety. Figure 5.1 adapted from Quirk et al. shows variability within a variety on a scale from “relatively uniform” to “relatively diverse”:

170 5 Theoretical and practical implications

relatively uniform Variation in individual’s usage (ELF usage)

Any given variety of English relatively diverse (ELF usage)

Variation in community’s usage (ELF usage) Figure 5.1. Variation within a variety (adapted from Quirk et al. 1985: 31).

If the variety is “relatively diverse”, then one could talk about diversity within an individual’s usage and the community’s usage. The results of the present investigation on ELF usage show relatively high diversity both within an individual’s and the community’s usage, as marked in Figure 5.1. The confusion and the debates around whether ELF is a variety are caused by perhaps unintentional but careless usage of the term ELF and suggestions of “teaching ELF” (e.g. Alptekin 2007; Kuo 2006; Sifakis 2007, 2009), which suggest that ELF is a variety, consisting of a set of features that can be taught. ‘Integrating ELF into one’s teaching’ is not synonymous with ‘teaching ELF’. It is important that the learners who will use English as a lingua franca need to be made aware of ELF in general: ELF should be integrated into the curriculum in the teaching of English (see section 5.3), and the growth of English as a lingua franca needs to be factored in English language teaching (Jenkins 2000; McKay 2002; Seidlhofer 2001; Widdowson 2003). In addition, undoubtedly, one can talk about the ELF core, which has been the topic of Master’s theses (Noorman 2008). As phonology/phonetics are closed systems with a set of features, it is possible to teach students the sounds one needs to produce correctly for intelligibility. However, I will argue here one should talk about ‘teaching ELF’ with caution, as it has ‘variety’ implications. Teaching an awareness of ELF, of course, is an entirely different matter. Naturally, commonalities of usage can be outlined provided that there is a clear set of criteria (section 3.3), and that the ratio of non-standard usage

5.2. Theoretical implications

171

to standard usage is shown clearly (section 4.1), as the present study strived to do. However, the variability would have needed to be very low for ELF usage to be regarded a variety. With the dynamic nature of its speakers and the numerous L1 backgrounds, stability is unlikely to be found in ELF settings. ELF research so far, including the present study, tells us that ELF usage is variable and very context-bound. The usage will depend on who is talking to whom, in what type of setting and for what purposes. If ELF is not a variety, what is it and which term can be used to best describe it? In the light of what has been said, it is reasonable to suggest that ELF is a functional term and not a linguistic one, i.e. it does not refer to a set of codified features that show stability of usage, as discussed before in literature (Kirkpatrick 2008: 28). ELF refers to a situation where speakers from different L1 backgrounds use English as a vehicular language. This definition does not presuppose the presence of a stable set of features that can be codified with ease. The notion of ‘speech community’ versus ‘discourse community’ is certainly of relevance to any discussion of variety. In the earliest definition, a ‘speech community’ was defined as a group that shares a set of linguistic rules and conventions (Bloomfield 1933). In a less strict definition, it was described as a group sharing a set of norms (Labov 1966: 7). There have been other definitions of a ‘speech community’ since the first one; however, they all point to a relatively set group in terms of rules, norms and functions. The concept ‘discourse community’ is much more suitable for describing communities that are in a relative state of flux (Bizzell 1982). Discourse communities, as Swales discusses them, are “centrifugal”, i.e. “separating people into occupational or speciality-interest groups” and speech communities, “centripetal”, i.e. “absorbing people” (Swales 1990: 24). Swales continues with the characteristics of a discourse community: 1. A discourse community has a broadly agreed set of common public goals. 2. A discourse community has mechanisms of intercommunication among its members. 3. A discourse community uses its participatory mechanisms primarily to provide information and feedback. 4. A discourse community utilizes and hence possesses one or more genres in the communicative furtherance of its aims. 5. In addition to owning genres, a discourse community has acquired some specific lexis. 6. A discourse community has a threshold level of members with a suitable degree of relevant content and discoursal expertise.

172 5 Theoretical and practical implications The community in the present investigation seems to reflect the six characteristics of a discourse community as described by Swales above. First, the speakers in the present study have common goals they wish to attain (1). Second, they have certain participatory mechanisms, i.e. lectures and group-work sessions (2). Next, the speakers in the present ELF setting take advantage of the “informational opportunities” (Swales 1990: 26), for they are in high-stakes situations where they must ensure information exchange (3). The next characteristic (4) refers to the adjustment of “topics, form, function and positioning of discoursal elements” (Swales 1990: 26). The speakers in the ELF settings described in the present study show certain adjustments, as described in Chapter 4 and discussed here in Chapter 5. A discourse community can also have its own lexis (5). Lexis, in this depth, has not been a part of the present investigation. It is, however, common knowledge that members of a speciality-group have their own jargon. Finally, the speakers in the present setting have a certain level of content, owing to their shared background, and they display certain discoursal strategies, pointing to what may be described as “discoursal expertise” (6) (Swales 1990: 27), as discussed in sections 4.2.2 and 5.1. In consideration of the evaluation made above, it is reasonable to suggest the ELF community in the present setting is a discourse community rather than a speech community. ELF usage is better relieved from pressures of forming a variety and speakers in ELF settings, from the obligation of belonging to a speech community. To repeat the conclusion of the discussion on the previous question, ELF usage is not a variety nor does it need to be a variety to be studied separately from World Englishes, creoles and learner languages. It is the largest use of the English language today by legitimate discourse communities, and that fact on its own suffices for ELF settings to be investigated thoroughly. Where should ELF speech be sought? The fundamental differences between learners and speakers in ELF settings show clearly that if one aims to undertake an investigation on ELF usage, the only place to do this is an ELF setting and not language classrooms where language is the primary concern for everyone involved. Learners in language classrooms will be paying attention to as many aspects of the language as possible at all times. Any research resulting from language classroom data will fail to show what speakers in ELF settings do and the results arising from such research cannot be extended to any other context than the language classroom (also discussed briefly in Cogo and Dewey 2006: 65, Cogo and Dewey 2011),

5.2. Theoretical implications

173

since it is not possible to simulate a real-time authentic speech situation. As Schlegoff very rightly says in Wong and Olsher (2000): The talk that learners are going to have to do when they’re not in the hothouse of the classroom is situated in the real world where they have real things to do, and that’s the talk that people ideally should be recording and studying if they want to understand what the real world problems are for those who are speaking a language that is not their native language. (Schegloff 2000:122)

The settings and where data came from must be taken into account before any research results are considered. If it is usage of English as a lingua franca that one needs to investigate, the setting in which to undertake such an investigation cannot and should not be a classroom setting. Compilation of such spoken data is by no means straightforward; it is time-consuming, complicated and costly. However, if we want to understand how lecturers and students interact by using English as a lingua franca, we must investigate lingua franca settings. In this sense, even set-up speech events where speakers use English as a lingua franca can come closer to documenting ELF usage than data collected in language classrooms.

5.2.2. Norms and standards for speech I mentioned earlier in this study (section 3.3) that there is variation within native varieties of English, and it is by no means easy to draw clear boundaries between what is standard and non-standard. Some features that are considered incorrect can be observed in native speakers’ speech frequently. When native speakers have non-standard usage in their speech, it is generally termed ‘variation’ whereas when non-native speakers have the same usage, it is considered an ‘error’. This study aimed to show that non-native speakers too can have variation in their speech, especially that the nonstandardness observed shows clear patterns of reducing redundancy and increasing explicitness, both aimed for communicative effectiveness. In the light of this, the important question for this study and all studies dealing with the form of speech is what the standards are for speech. If we approach the general concept of language as a set of rules, a clear and solid set of rules to prescribe for standard usage of the language becomes necessary. What is considered standard in traditional grammar books is traditionally based on written English. In other words, the standards for

174 5 Theoretical and practical implications spoken English have long been set by written English and what would be considered correct and therefore acceptable in written discourse. This is recognized by the MICASE team, as one of the reasons for undertaking such a project and thus as one of the ways MICASE “benefits (our) community”. The team expected that they would find “many divergences from those described in current grammar and vocabulary books, which have largely relied on introspection or on features of written texts” (Simpson et al. 2002). As Preisler says “By definition, a standard or norm seeks to minimize diversity, and diversity is usually discouraged in the teaching of a written standard” (Preisler 2011: xiv). Written norms are not appropriate for speaking, for speech and writing are two very different types of discourse. The fact that speech is generally impromptu and requires real-time production and processing creates the biggest difference between these two types of discourse with regard to one’s production. So evaluating a speaker’s communicative competence by his/her adherence to standard forms based on writing would be a practice that is against the nature of speaking, it is unjust to the speaker, and it is a practice that does not comply with the way natural speech is. This has been discussed in earlier work. Brazil’s work was important since it was a prominent work in discussing spoken grammar, i.e. a different model for speech rather than using the same model for speech and writing although he does not provide a detailed description of spoken grammar (Brazil 1995). His work was supported by Carter and McCarthy, who discussed the inappropriateness of standard grammars for speech since they were based on writing (Carter and McCarthy 1995). The CANCODE project (Cambridge-Nottingham Corpus of Discourse in English), mainly led by Carter and McCarthy, was compiled with the aim of enabling the study of “spoken grammar in particular relation to different genres of speech” (Carter and McCarthy 1997: 8). Most importantly, too, such a project design allows insights to be developed concerning differences and distinctions between spoken and written language use, with particular reference to grammar. The major contemporary grammars of the English language are largely based on written examples; the new data are exemplifying common, ‘standard’ patterns of use alongside patterns more familiar from written-languagebased grammar. For example, new data […] reveal key interpersonal functions in the use of tense, modality, ellipsis, deixis and clause structure; the crucial role of spoken discourse markers, acting as a kind of conversational punctuation, is also illustrated. Additionally, variations in

5.2. Theoretical implications 175 canonical (written) word order can be shown to be common across a wide range of genres of speech.

The opposite view suggests that it is the same grammar that forms the basis of both speaking and writing and therefore no distinction needs to be made (Biber et al. 1999; Leech 2000). The present study finds itself closer to the first approach, the Nottingham school, as Leech calls it (Leech 2000: 688). Evaluating a speaker’s spoken production based on writing norms would not be an issue of “emphasis and attitude” only as Leech claims (Leech 2000: 690); it would in fact be denying the very nature of speaking. The first preliminary in describing speech properly is naturally the compilation of spoken corpora, e.g. BNC, LLC, MICASE (Simpson et al., 2002), BASE, VOICE, ELFA and ACE. These corpora all serve the important purpose of providing spoken data from different settings. The second step to take towards providing appropriate norms for speech is to have a spoken corpus on its own presuppositions and not on presuppositions of writing. Although the earlier CANCODE project was compiled with the primary aim of exploring authentic spontaneous speech and investigating spoken grammar, it does not document spoken English by nonnative speakers. The speech recorded for CANCODE comprises speech from native speakers of English only, and the findings that arise from this corpus can only explain variation in the spoken language of native speakers, not in non-native speakers’ speech. Carter and McCarthy suggest the usage of CANCODE data by learners of English so learners of English can “see, hear and understand conversational English in a range of different contexts of use” (Carter and McCarthy 1997: 7). Of course, it would be ideal that those who are studying English and consider themselves learners could be exposed to authentic usage of spoken English from a variety of contexts, situations and settings, and undoubtedly, CANCODE data are of great interest; they reveal variation in native speaker usage of spoken English, which is important in studying spoken English. However, they exclude speech by speakers who speak it as a lingua franca. In this respect, they are not appropriate for speakers of English who will use it in lingua franca settings. For the purposes of ELF settings, the third essential preliminary is the availability of corpora that document usage of English in settings where it is used as a lingua franca. Target use for speakers in ELF settings is not native spoken usage that is culturally-loaded and that is full of idioms, but

176 5 Theoretical and practical implications rather effective speech. Presenting learners of English with the heavy idiomaticity present in native spoken English is not likely to provide them with more choices; it is only likely to create opacity in their speech. Prodromou discusses the concept of ‘authenticity’ and claims it is embedded in the context (Prodromou 1996: 372). When it is taken out of its context, it simply stops being authentic. Native speaker usage is not necessarily helpful for speakers in ELF settings, as also shown in the present study in the analyses of questions. Although rising question intonation is not always present in native speaker usage, it helps speakers in ELF settings register questions (section 4.1). The information and research findings that have been and will be generated by all the aforementioned corpora, provided that they present appropriate speech, will enable research to produce appropriate descriptions of the use of English as a lingua franca. A thorough documentation of spoken English from authentic settings will surely contribute to the understanding of a ‘spoken grammar’ and whether it is possible to set rules that reflect the nature of spoken English successfully. To date, there is no clear explanation or documentation of spoken grammar with the exception of a new corpus-based monograph, developing Brazil’s work (1995) (O’Grady, 2010). It is, however, the application of the research findings based on these corpora that is critical. The place to apply the research findings is the language classroom, which brings the need for a discussion of the teaching of English language. Before moving on to such practical implications (section 5.3.2), let us turn to what good English may mean and whether it is synonymous with correct English.

5.2.3. Good English The results of the present study and a discussion of the general notions of ELF situations bring to the fore another important notion, albeit a much more general one: Good English. What is good English? This question will naturally receive a variety of answers depending on a number of factors, e.g. the one answering the question, the setting that is relevant. Greenbaum discusses ‘good English’ by contrasting it to ‘correct English’ as follows (Greenbaum 1996: 17):

5 Theoretical and practical implications 177

Good English is sometimes equated with correct English, but the two concepts should be differentiated. Correct English is conformity to the norms of the standard language. Good English is good use of the resources available in the language. In that sense we can use a nonstandard dialect well and can use the standard language badly. By good English we may mean language used effectively or aesthetically; language that conveys clearly and appropriately what is intended and language that is pleasing to the listener or the reader.

First we may deal with whether English that does not conform to the norms is “pleasing to the listener or reader” or not and discuss this in connection with the survey in the present study (section 5.1). Some non-standard production may trigger certain ideologies in some people’s minds. Their responses may clash with what they do in interaction, as was the case in the present study. The respondents claim to find the omission of third person singular -s irritating although it did not cause disturbance in communication. This kind of contradiction is important, for it reveals people’s attitudes. Considering good English as the effective use of the available linguistic resources, and drawing a clear distinction between correct English and good English as suggested above, suggests that “good English” is not a notion that is determined necessarily by one’s level of proficiency. In other words, speakers who have lower levels of proficiency are also capable of speaking good English. This seems to corroborate the findings of the present study in the dimensions of form (section 4.1) and discourse (section 4.2). A large body of data showed in the present study that speakers who have lower levels of proficiency were able to use the language effectively. On the notion of effectiveness, with respect to ELF settings where speakers use English as the vehicular language, using the language effectively takes precedence over language complexity. “Conformity to the norms” will naturally be less critical for anyone who is trying to complete a task through the medium of that language. This is not to say that correctness is irrelevant. Naturally, grammatical accuracy is important; a sentence needs to be made up of the right constituents to be sensical. What is suggested here is that, in operating in ELF settings where speakers are from a range of levels of proficiency, both those who are native speakers or those who are highly proficient, and those who are non-native speakers with varying degrees of proficiency are challenged. Relevant to the dichotomy of native and non-native speakers is Ferguson’s discussion of academic writing. Ferguson plausibly argues that both

178 5 Theoretical and practical implications native and non-native speakers are “novices” in acquiring academic (writing) skills: With regard to the acquisition of this specialized competence, the native speaker and the non-native speaker both start out as novices, a position of parity that the native/non-native dichotomy obscures, but that is noted by some of the editors from Flowerdew’s (2001) study, who remark that many of the problems of Hong Kong/Chinese authors are shared by their native speaker counterparts (see also Swales 2004). This is not to say that native-speakerhood confers no advantage at all. It may do so with particular regard to intuitions of grammaticality [...]. The key dimensions of difference, then, are not so much native or non-native speaker status as expertise (novice or expert) and proficiency, the significance of the latter being underscored if one bears in mind the frequency with which so-called non-natives display greater facility in academic writing than natives, whose performance levels are in fact very variable. (Ferguson 2007: 28)

A similar argument can be made for academic speaking in general. In handling the everyday challenges of ELF settings, both native and non-native speakers are novices. It is argued here that being proficient in the language does not presuppose that one is also a pragmatically effective speaker. In settings where English is used as a lingua franca, communicative effectiveness takes precedence over language complexity. Any speaker may aim for better accuracy, fluency and language complexity, but when it comes to investing in a communicative situation, it is ways of achieving effectiveness that help speakers produce the desired outcome, i.e. be communicatively effective. The proficient/less proficient or the native/non-native speaker dichotomies, therefore, are not of primary relevance or utility to international settings. In the light of the above, the notion of good English in ELF settings appears most strongly associated with effectiveness, not with correctness or adherence to native speaker norms.

5.3. Practical implications The results of the present study also has some practical implications. Using another language than one’s own for high-stakes communication requires heavy investment in the communication process. Because of this, ELF settings are inherently challenging for all parties involved. From the results of the analyses, a number of implications emerge for, first, lecturers, with regard to the adjustment of lectures to lingua franca settings, for decision

5.3 Practical implications 179

makers as to providing lecturers and students with the help they might need in operating through the medium of a foreign language, and for those involved in the teaching of English. I will now discuss them in the same order here.

5.3.1. Comprehension-facilitating lecturing behavior How can we help students cope with English-medium lectures? The results here showed certain types of comprehension-facilitating behavior, which is what we will turn to now. Listening to a lecture is in one sense a much more complex task than working in groups, because speakers have little room for maneuver and little opportunity to make use of discursive strategies. Although more work has been done on lecturing in English than in student group-work sessions, it seems as if we can safely assume that lectures are riskier in terms of communicative effectiveness. At the same time, it seems as if it is only through studying interactive speech in lingua franca that we can get information on how meaning is negotiated and how communicative effectiveness is achieved. So investigating both monologic and dialogic speech produces the information we need on what type of communicative skills are helpful in getting the message across. Conveying information is clearly only one of the aims when giving a lecture. Researchers are becoming aware of the significance of lectures as speech events used not merely to convey information but also achieve a good rapport with the audience and convey attitudes and opinions. Nonnative speaker lecturers who teach non-native speaker students need information regarding how to deal with the challenges of the situation. There is to date little research-based advice for non-native lecturers teaching nonnative students, with the exception of Airey’s work on physics education (Airey 2009). The present project showed that the types of behavior outlined below are useful in increasing communicative effectiveness. The following tentative recommendations I am making here are primarily for lecturers who need to teach in what have now become ELF settings, but also for decisionmakers who can, through providing training, equip lecturers and students with the type of training they need to cope with the challenges of lingua franca settings.

180 5 Theoretical and practical implications The first recommendation deals with asking students questions and letting them raise questions. As simple as this may sound, as explained in detail in Chapter 4 and Chapter 5 here, questions are of specific importance. They are the only verbal real-time signals as to what is going on in communication and are invaluable in checking understanding. They are also important organizational devices. We could even say that questions are the only verbal features that can be used for checking the quality of comprehension in real time. In this sense, they serve the teaching-learning situation tremendously. It is, therefore, highly recommended for lecturers to ask questions several times during a lecture and give students the time they need to react, negotiate meaning and check understanding. Despite the common belief about students’ reluctance to ask questions, there is empirical evidence that they actually do want to ask and answer questions in lectures (Mulligan and Kirkpatrick 2000: 330, 331). The results here showed that lecturers asked many rhetorical questions. An average of 61 per cent of all the questions the lecturers asked were rhetorical. Undoubtedly, rhetorical questions are also helpful since they create a pause in the monologue and signal importance; however, actual questions will tell the lecturer more on whether his/her students have comprehended the subject matter. Generally speaking, asking and answering questions in lectures should be possible unless the lecture is a very large one. In very large lectures where there are at least one hundred students, interactivity becomes harder to achieve. Such large lectures tend to be intimidating for students to respond individually to questions. In situations of this sort, the lecturer can direct his/her questions to students and make use of technology when getting answers. Most lecture halls are now equipped with computers, and most universities with wireless Internet connections. The lecturer can easily get the answer to the question in the form of quick e-mails. Better yet, universities are increasingly getting equipped with electronic platforms, which allow the lecturer and students to send and receive instant messages. These instant messaging services could easily be used to facilitate the asking of questions during lectures from computer to computer16.

16. This study has investigated real-time communication and has not dealt with pre- and post- lecture activities and events prior to and subsequent to lectures. Even without using instant messaging, students can use these electronic platforms with ease to post their questions in advance, which can then be used by the lecturer in planning the lecture.

5.3 Practical implications 181

More costly and therefore perhaps less realistic for all universities to adopt are Audience Response Systems (ARS), which are increasingly being utilized at North American Universities for instant feedback purposes. ARSs increase interactivity with the help of a remote control and create real-time feedback opportunities. The remote controls connect to a main computer through the sensors in the room. After students have responded with the help of their remote controls, a bar chart is shown that both the lecturer and the students can see, indicating how many in the lecture had understood the issue in discussion. The advantages of this type of real-time feedback are increased student activity and interactivity (Banks 2006; Kift 2006: 92), more engaging lessons (Menon et al. 2004: 220), enhanced learning (Uhari et al. 2003: 3) and decreased crowd psychology (Menon et al. 2004: 220). There are positive results about increased numbers of responses to lecturers’ questions (Draper and Brown 2004: 89). When students need to produce an answer through a remote/handset, they are involved in mental processing and problem solving instead of “playing the role of the spectator” and waiting to see if the question will be answered by others (Draper and Brown 2004: 89). One disadvantage, and an important one, however, is the limitation in the nature of responses one can give through ARSs. It takes little time and effort to indicate understanding by clicking the remote; however, when it comes to giving detailed answers, this technology is yet limited and cannot initiate real-time verbal communication and high interactivity. The second recommendation is about the use of pragmatic strategies to enhance communication and increase interactivity. The importance of using pragmatic strategies (Björkman 2011; Mauranen 2005 and 2006b; Penz 2008) and the difficulties non-native speaker students have with lecture comprehension (Mulligan and Kirkpatrick 2000) have been discussed in the literature and covered in the earlier sections of this work. The interactivity that pragmatic strategies bring into the speech event and the proactive work that enhances understanding and prevents misunderstanding have been reported as very useful communicative behavior, and this has been supported by the results of the present study. In an experimental study (Morell 2004), significant changes were observed in student participation after the lecturers made their lectures more interactive and changed their lectures by using more macromarkers (i.e. starter, elicitation, accept, attitudinal, metastatement, conclusion as discussed in Murphy and Candlin 1979), questions and negotiation of meaning (Morell 2004: 335). In a more recent study from a very similar setting to the present setting, i.e. a technical university

182 5 Theoretical and practical implications in Finland, the lectures that were perceived as most comprehensible by students had a range of interactional features in them (Suvuniitty 2010). In the present study, while the group-work sessions had many examples of pragmatic strategies, lectures did not have so many instances and were largely monologic. Monologic events where the listener has very few opportunities if any to check his/her own understanding are where misunderstandings are most likely to occur. Therefore, it is highly recommended that lecturers in lingua franca settings create as many opportunities as possible for pragmatic strategies. The pragmatic strategies that have been investigated in this study are pragmatic comments to explain terms and concepts, to give details about the content of the task, to give information on the discourse structure and context, to make intent clear and to create common ground. Along with these strategies, backchanneling, reported as a useful strategy in spoken communication (Bjørge 2010) used to signal understanding, and repetition have been included in the investigation. Especially strategies, as in (323)– (326) below, and (225)-(236) in Chapter 4, looking ahead and signaling what is to come, as in (327)–(332), and (218)-(224)(both briefly discussed in Mulligan and Kirkpatrick 2000: 317), and helping the listener orient himself/herself by making the discourse structure clear are very helpful to the listener: (323) (324) (325) (326) (327) (328) (329) (330) (331) (332)

If we go back to the starting picture, gasification was one way to go from solid fuel … In biomass conversion we saw that… So multiplication is finished. We have actually been working on… I am now moving onto… …, which I will explain soon. I’ll show you what I mean. I will explain in a few second what green certificates are. I will try to tie this together by talking about infusion. and I will try to tie this together by talking about diffusion

Signaling significance is another good way of drawing the student’s attention to central issues around the subject area, and helping them distinguish what is important, as in (333)–(343) below:

5.3 Practical implications 183

(333) (334) (335) (336) (337) (338) (339) (340) (341) (342) (343)

It’s very important… This is the main discussion. One thing which is very important… That’s the ultimate goal. …, which is a very important part of catalysis engineering. It is noteworthy that… The reason why… That’s the reason… … not because of economics because of… There is a very easy explanation why… But there is a very important point:…

Finally, it may be recommended here that lecturers use Heads (344) and Tails (345), and Pre- (346) and Post- dislocation (347) for extra explicitness and the establishment of common ground, which aids communicative effectiveness: (344) (345) (346) (347)

The best part we have already calculated. We need to calculate carefully the margin efficiency. Actually the catalysts they could die. Well it is not so emission-free hydro-power.

The strategies mentioned in the present study are clearly strategies that any good lecturer should/would pay attention to. Advice given in other studies that investigated successful classroom discourse for non-native teaching assistant training given is in agreement with the advice provided in the present study: a. Naming processes b. Overtly marking major points, both to evaluate and reinforce student achievement c. Developing cohesion and continuity within and between classes by repetition and “linking talk” d. Explicitly organizing topics and marking topic change e. Stating the scope of the students’ responsibility f. Using questions in a timely fashion g. Using persuasive techniques (Rounds 1987:666). Another source of a more general nature on the importance of informative speaking (Samovar and Mills 1995: 284) lists repetition as an important feature of informative speech, as Rounds does in (c) above. In order to

184 5 Theoretical and practical implications contribute to clarity and deliver the intended information, those who prepare informative speech, e.g. lectures, are advised to consider definitions, reinforcement and emphasis, noting what is upcoming, referring to the familiar and summaries carefully among other things (Samovar and Mills 1995: 290–305). Among the seven items in Rounds’s list on communicatively competent teaching discourse, there are five strategies that have been found as probably increasing communicative effectiveness in the present study: a, b, c, d, and f, respectively, naming processes, overtly marking major points, developing cohesion and continuity by using linkers and repetition, organizing topics and marking topic change overtly and using questions (section 4.2.2). In addition, the four features the latter source mentions have all been found as strategies used by the lecturers at varying degrees in the present material (section 4.2.2). In addition to the lecturing behavior discussed above, there are visual aids that support the lecture. Most lecturers nowadays make use of slides when they give lectures, with the main aim of increasing the delivery and effectiveness of their lectures, and it is recommended in literature that all lecturers support the subject matter by using proper visual aids (Airey 2009; Airey and Linder 2006; Hellekjær 2009; Klaassen 2001). They do so also by providing their students with these slides, which they post on electronic platforms created for the course. Making these slides available helps students with the note-taking problems students face during lectures (Airey and Linder 2006; Hellekjær 2009). In fact, research comparing Englishmedium instruction with L1-medium instruction has shown that effective lecturing behavior and supporting lectures with proper visual aids were found to be more significant than the language used (Airey 2009; Airey and Linder 2006; Hellekjær 2009; Klaassen 2001). To make the most of these visual aids and increase effectiveness, the slides should be prepared in connection with the processes named clearly on the relevant slides, using bullets to overtly mark major points, putting sufficient linkers on the slides and making sure the topic change on the slides is in connection with the topic change in the lecture itself. If the questions the lecturer plans to ask are prepared in advance, they could be put on the slides, which would help students considerably in identifying important issues in connection with the subject matter.

5.3 Practical implications 185

5.3.2. Support for lecturers and students who need to operate in ELF settings Relevant to the effectiveness of lectures is support for lecturers and students who need to operate in ELF settings. It is important that awareness is raised on what should be the target language usage in ELF settings, which has been one of the implicit aims of the present study. Such awarenessraising can be achieved at the macro level by clearly-written language policies that include training for teachers and students who both need to be equipped with the skills needed to cope with the complexities of such settings, and at the micro level, by in-house training and courses that could be administered to both teachers and students. The proposal here is that lecturers who need to teach in ELF settings, regardless of their L1 backgrounds, attend courses and/or workshops on effective lecturing behavior that would increase communicative effectiveness in ELF settings and equip them with the skills required when teaching students with a wide variety of L1 backgrounds and proficiencies. There would be countless advantages in building this type of platform, among them, providing a forum for a detailed discussion on what target-like language usage should be for the specific setting/s. Such a course/workshop would provide the lecturer first with theory and background to the research in the field, then with practice, where they would put into practice the useful lecturing behavior that has been shown as effective in relevant research, and finally reporting to the group of lecturers, where they would discuss the effects of these applications. This type of training need not be as exhaustive and costly as one might initially expect. Even merely series of workshops would be a way to raise awareness and help those teaching in English. It is, however, a problem that some lecturers will be reluctant to attend such courses or workshops even when they are offered the opportunity. The reasons for their reluctance may be summarized as general lack of time and not acknowledging the need to get help in operating in Englishmedium education. An obvious solution to the time issue is to create room in the lecturers’ timetables for such professional development. When it comes to the lack of interest, some universities have made these courses or workshops obligatory. However, pedagogically, this is a questionable practice. It is important that lecturers understand the ramifications of lecturing in a language that is not their own. A recent study from a Swedish setting has shown that speakers have a 23 per cent slower speaking rate when they

186 5 Theoretical and practical implications give oral presentations in English compared to their production in Swedish, which affects information content in these presentations (Hincks 2010). The data come from oral presentations, which are prepared speech events; nevertheless, it could be certainly be argued that if speakers speak 23 per cent slower and say less even in prepared speech, the situation might be similar, or they might be even slower, in unprepared speech. It is as important to note that level of proficiency would not be the issue in training of this kind. Speakers who are more proficient need to convey information to those whose level of proficiency might be lower and not only vice versa. So whether or not a speaker is proficient is not decisive in this type of training. There is empirical support for this. A study carried out in Norwegian higher education showed that English-medium lecturers in Norway assumed that it was predominantly exchange students who had problems with English-medium education and not the Norwegians (Hellekjær 2009). However, the same study showed that there was no statistically significant difference between the Norwegian students and exchange students in terms of comprehension difficulties. The situation is similar in most Northern European countries, e.g. in the present setting Sweden. It must be conveyed to both lecturers and students that operate in ELF settings that it is also more proficient speakers who need to seek ways to communicate effectively with those who may not be equally proficient, and adaptation of one’s language to the situation and the speakers in that particular situation is crucial. The exchange students who are enrolled in studies abroad through the medium of English are required to take central exams and prove their levels of proficiency. Two such well-known and internationally-recognized tests are TOEFL and IELTS. The type of TOEFL that is offered depends on the country where the test will be taken. In the present setting, Sweden, both the Internet and paper-based TOEFL tests are offered. The main difference between these two formats is that the paper-based test does not test the candidate’s spoken skills. For this purpose, TOEFL has another test, Test of Spoken English (TSE), which has been designed to test the candidate’s spoken skills. The IELTS exam is offered in two formats, namely general or academic training. The test is made up of several sections, including a speaking section, namely, listening, general/academic reading, general/academic writing and speaking. These tests aim to give one sufficient information about the proficiency of a student; however, they do not necessarily provide us with sufficient information on the student’s speaking skills. The TOEFL paper-based test

5.3 Practical implications 187

comprises reading, listening and writing skills, and unless supported by the additional TSE test, it gives no information about a speaker’s spoken ability. In many countries that exchange students come from to the present setting, the paper-based test is offered and has a lower fee than the computer-based TOEFL. On this note the IELTS test appears to be more comprehensive since it includes a speaking section and one of the two formats of the test is geared towards academic studies. It is, however, evident from the tasks that they are based primarily on monologic speech with possibly some questions from the examiner. In this respect, they fail to fully reflect the communicative challenges a student faces in lingua franca settings in English-medium education. The tests would need to include dialogic speech with several parties present where there are information-gap activities and problem-solving tasks. This is not always possible for practical reasons with the resources available today. However, I highly recommend any university that calls itself international and that accepts the paper-based TOEFL test, that they require the Test of Spoken English (TSE) in addition, which will provide them with information on the candidate’s spoken production. Such a practice would make life so much easier for course administrators, and consequently students, who would be placed in the right groups. Local students in the present setting are required to have satisfactory skills in English, equivalent to English A17 for general admission. For local students who want to be enrolled in programs where English is the medium of instruction, the requirement is English B18. These admission requirements for local and exchange students are insufficient for testing the academic preparedness of a student candidate for studies in English, for they do not include a thorough examination of an individual’s language skills pertaining to speaking. The most pressing issue for both local and exchange students who are enrolled in English-medium programs is the training required to equip themselves with academic skills. The type of training they need is not primarily proficiency, but the academic skills one needs for studies in English. These skills include the explicitness strategies that aid comprehensibility, discussed in the form dimension of the present study, pragmatic strategies that are invaluable in increasing

17. In Swedish: Engelska A. It corresponds to the first level of studies in English at high school. 18. In Swedish: Engelska B. It corresponds to the second level of studies in English at high school.

188 5 Theoretical and practical implications the effectiveness of communication, discussed in the communicative effectiveness section of the present study, and general spoken skills with clear articulation and intonation, discussed in connection with asking questions here. Finally, there is awareness-raising for target usage. It is clear from the results of the present study that the target usage that is expected of students or what students expect from their lecturers does not need to be a native variety but effective usage of the language that allows for optimal information transfer. Jenkins touches upon the issue of admission tests and cites a representative of the new Pearson Test of English (PTE) Academic from an interview with the periodical, EL Gazette (September 2008: 10)

to create an international exam we started by hiring item writers from the UK, the US and Australia […] [b]ecause we are not using a single standard model of English, we can grade all non-native students on a single scale. The first thing we look for is comprehensibility – are they understandable to the native speaker? (Jenkins 2011)

It is naturally important to be able to be understood by a native speaker, and similarly, to understand a native speaker, especially if the student is to study in an English-speaking country. However, it runs short on judging how effective a speaker will be in lingua franca settings. Being understood by a native speaker is not enough for anyone who will need to use English in lingua franca settings. A wrong assumption, and a common one, is that being understood by non-native speakers is a subset of being understood by native speakers. Equally wrong is to assume that understanding a nonnative speaker is a subset of understanding a native speaker. There are certainly some overlaps, but also many differences in how native and nonnative speakers speak. Preparing students for their academic studies in English is surely a costly endeavor. The scheduling of such training is another complex issue since students come from a wide range of departments. However, the issue is pressing, and students in English-medium education will continue to have these needs. Examination boards need to realize the needs of those who will be speaking English to communicate with other non-native speakers, and universities that aim to be international need to invest in integrating effective language training into their education along with their studies, which would help students largely in handling the everyday challenges of English-medium education. International university English language re-

5.3 Practical implications 189

quirements need to be adjusted accordingly, without being “determined in accordance with entrance examinations grounded in native English” (Jenkins 2011: 927), for “it is a contradiction for any university anywhere that considers itself international to insist on national English language norms” (Jenkins 2011: 934). 5.3.3. Pedagogical applications: Issues for the language classroom We will now turn to the language classroom. I have not dealt with language instruction in this monograph. The data I used came from content courses, and the subjects did not include any student who was also enrolled in language courses. Nevertheless, the issues I have discussed so far have implications for language classes. In this section, I aim to address some pedagogical applications of the findings of the present study and ELF research findings. Some of the most important issues for the language classroom concern the needs and expectations of the specific group, comprehensibility as the priority in teaching English to those who will use English as a lingua franca, the use of realistic course materials and testing criteria for spoken production. In identifying the needs and expectations of a specific learner group, one first needs to return to the issue of norms and standards. The discussion on norms and standards in the previous section takes us to the pedagogical discussion of what norms and standards students should be presented with. In any debate on teaching English, we are reminded of the responsibility of prescribing a certain set of conventions as ‘correct’ usage since the default assumption is that teaching is prescriptive (Bex 2008: 229). This requires a description of standard English and automatically brings forward the question of what standard English is. This question is a very difficult one to answer, and I will suggest here that it is actually not possible to provide an answer that is appropriate for all learners. What can be done, however, is to distinguish between different types of learners and consider their needs and expectations. For example, there are learners who need to use English to speak to native speakers in countries where it is L1 and those who need to speak it in international settings mostly to communicate with other non-native speakers. As long as the teaching of English does not include its widest use in the world, it will be insufficient in providing learners with the knowledge and skills required to use it in its widest use. With regard to those who use English as a vehicular language on international grounds, the knowledge of how the language

190 5 Theoretical and practical implications is used primarily among its non-native speakers for different purposes in international settings will be more relevant than how it is used in Englishspeaking countries by its native speakers. So what ELT can do is to take steps towards meeting current demands instead of feeding on dated descriptions of English, and differentiating between different types of learners and their needs. After all, learners should be able to decide what type of English they need and wish to learn (Seidlhofer 2006). One should bear in mind that there is a wide spectrum of ELF settings in the world today (see Graddol’s list in Chapter 1), ranging in terms of the stakes they involve. Tourism perhaps is not always a high-stakes ELF situation whereas international safety or banking would involve high stakes, i.e. if miscommunication were to occur, it would have serious consequences for those involved. Some seem to suggest that ELF settings are generally less important than other so-called serious settings where communicative competence is important: English for them (next generation worldwide) is not simply the language to start conversations on a train or to place orders in a restaurant when travelling in a foreign country. It is the language of which they have to demonstrate a degree of mastery so as to win a place in education and employment in their own contexts and abroad. (Kuo 2006: 219)

There are of course ELF conversations on a platform while they are waiting for the train or while they are ordering in a restaurant; however, there are also very high-stakes settings, such as the setting investigated in the present study. It is precisely “winning a place in education” and working towards keeping the place they won that they want to do through the medium of English. Kuo claims in her article that ELF is not able to prepare learners for their future needs in international settings. She misses the point completely: It is precisely preparing speakers for international usage that ELF can do.19 It is by no means certain that being exposed to NS usage only can do this. In fact, several ELF research projects have had as their implicit aim as preparing individuals for international usage and equipping them with the skills they need to communicate effectively, such as the present study (Jenkins 2007: 112, 113). More research and proper descriptions from a variety of settings will undoubtedly provide ELT with invaluable information as to what English

19. See Jenkins 2007 for a critique of Kuo.

5.3 Practical implications 191

usage is like in different settings. Until then, realizing the needs and expectations of different types of learners and producing materials including features they will need to use frequently in their target settings will be important steps in the right direction. The transition from description to prescription should be made with caution, and this task should be undertaken only after thorough descriptions from a variety of settings have been made available. The second issue of importance here is prioritizing comprehensibility in language teaching. In settings where English is a vehicular language, comprehensibility must be the priority, followed by language complexity, provided that achieving complexity is one of the aims of the speakers in such settings at all. As I discussed earlier in this chapter (section 5.1), prioritizing comprehensibility in language teaching was first put forward as early as in the 1970s, with the functional hypothesis and the Functional Sentence Perspective. So, the issue is by no means a new one. The challenge has been to suggest ways of putting such theory into practice. What should then the language teacher actually do in the classroom and how should the teaching of English be modified to prioritize comprehensibility and provide the learner with a realistic model? One of the most recent fora where these issues have been brought into discussion has been ELF research. The importance of exposing learners to a wider range of English has been discussed in a number of studies (Seidlhofer 2001; Jenkins 2006a) alongside raising important but broader issues such as raising teachers’ and teacher educators’ awareness (Jenkins 2006a: 169), general “efforts to reduce the nativespeakerist element in some teaching materials” (Jenkins 2006a: 169), the merits of native and non-native teachers of English (Jenkins 2006a: 172), challenges to learner language theory (Jenkins 2006a: 166) and the importance of providing the learner with modern and broad-based descriptions of language. Preparing learners for the real world cannot be achieved through “inadequate and outdated descriptions of language”; it requires “models that can be applied to a variety of communicative goals” (Mauranen 2006c: 144). All these issues are naturally central to improving language teaching. However, ways of achieving what is seen as important, as discussed here, and how to put them into practice have been discussed much less, with the exception of the lingua franca core, which highlights the features one needs to produce correctly for intelligibility (Jenkins 2000), the inappropriateness of unilateral idiomaticity for ELF settings (Seidlhofer 2004: 220) and the importance of accommodation strategies with respect to their inclusion

192 5 Theoretical and practical implications in teaching (Jenkins 2000, 2006a: 174). The question, then, can be raised again: What types of modifications are suggested for the language classroom? The first modification should be made to course materials, which is the third issue here, in an attempt to produce realistic course materials, as dicussed in Jenkins (2006a: 169). This issue need not be as radical as some have suggested it to be. There need not be a set of acceptable features in the form dimension for learners to adhere to for TESOL practice to consider ELF usage. In this respect, the integration of ELF usage into the forms traditionally used in TESOL adds to the existing set of norms. To be more specific, the items to prioritize in the teaching of English, as they appear in the present study and in ELF literature, seem to be: • The inclusion of non-native speaker accents in listening comprehension materials: It is important that speakers who will operate in ELF settings are exposed to a variety of accents since this is precisely what they will need to do when they use English as a lingua franca. Of special importance are materials for listening comprehension including a variety of accents (Smith and Bisazza 1982; Smith 1982) and cases of disturbance where meaning is negotiated and where communicative strategies are employed (Watterson 2008: 402). In addition, authentic recordings can be turned into course materials where students test their listening comprehension and note-taking skills. Another positive outcome of this practice will be increased student awareness of what authentic English is in international settings. • The inclusion of pragmatic strategies in speaking and listening materials: It is clear in the findings of the present study (section 4.2.2) that pragmatic strategies and negotiating meaning play a big role in achieving communicative effectiveness in ELF settings (section 5.1). After learners are presented with materials of authentic usage from ELF settings (listening), role-plays and other communicative activities can be used to enable learners to practice these skills. • The inclusion of syntactic structures that help increase explicitness: In grammar teaching, it seems a useful practice to include Heads and tails/Pre- and post-dislocation as elements that occur frequently in spoken English and as elements that increase explicitness (Mauranen 2007). Exposing learners to authentic speech that includes these elements will help them see how they are used.

5.3 Practical implications 193

• Giving word order issues enough airing space in the classroom: A common issue emerging from both the form and communicativeness dimensions of the present study, and the questionnaire on perceived comprehensibility and irritation is the importance of word order. It appears that non-standardness in word order can impede communication and create irritation, which in turn may cause turbulence in communication. • The inclusion and prioritizing of materials practicing features whose absence leads to overt disturbance: In the present study, non-standard questions were different from all the other non-standard features since they were the only features that led to overt disturbance. Questions should be addressed thoroughly in language classrooms in communicative activities such as information-gap activities, or group-work activities where students work on a task together. Such practice would mirror real-life communicative situations. The production of realistic materials is an issue primarily for textbook writers but also for individual teachers who prepare course materials for their students. Especially in higher education settings where students will use English as a lingua franca, teaching materials must include English by speakers in lingua franca settings. This naturally will be enabled by proper descriptions of lingua franca settings, as discussed above. The view that prioritizes comprehensibility in teaching is not shared by all. It has been claimed that ELF research suggests all teaching be done only based on comprehensibility, because “what is needed for comprehension is all that is needed to be produced” (Kuo 2006: 216). This is by no means the case. What is suggested in the present study is that features that are critical in terms of communicative effectiveness must be prioritized, followed by features that are not equally vital for communicative purposes. The fourth issue here concerns testing. Interestingly, the natural order of teaching and then testing needs to be modified in dealing with exposing learners to authentic usage. Testing is not necessarily an issue that needs to be dealt with after addressing issues concerning teaching. In fact, testing is an equally important issue if not a more pressing one. ELT examination boards base their practices on the ownership of English predominantly by its native speakers. The same approach is taken by most teachers and teacher trainers (Jenkins 2006a: 172). If World Englishes and ELF re-

194 5 Theoretical and practical implications searchers together point to a need for pluralism and inclusivity (Bolton 2005: 78; Seidlhofer 2005: 170), practice needs to follow. So how should testing change in the light of comprehensibility-based teaching and the findings of the ELF literature so far (including the findings of the present study)?

One of the most pressing problems will be to find a way of incorporating a WEs-ELF perspective into testing (Canagarajah 2005a). Solving this problem will involve […] finding ways of identifying accommodation, so that candidates are able to adjust their English for the purposes of showing solidarity with, or promoting intelligibility for, an interlocutor, without the risk of being penalized because their resulting speech does not defer to native speaker norms. (Jenkins 2006a: 174)

Based on the results of the present study, I will suggest here that any set of criteria to evaluate the spoken production of a learner of English will need to consider the following carefully: • The acceptance of a variety of accents provided that intelligibility is achieved: It is important that native-speaker accents are not presented as the only acceptable accents to achieve the highest grades in the evaluation of a learner’s spoken production. The learner should be free to approximate his/her speech to a native variety if this is the learner’s aspiration, but this should not be a criterion. • The acceptance of, or tolerance to, morphosyntactic features that reduce redundancy: One of the types of non-standard morphosyntactic usage in the present study concern successful reductions of redundancy. An example of this type of usage is the non-marking of the plural on the noun. This is not to suggest that such non-standard forms should be taught to learners; however, not penalizing them on items that help communicative effectiveness, e.g. the unmarked plural, could be a practice worth considering for the future. • The acceptance of morphosyntactic features that help create extra explicitness: Another type of morphosyntactic non-standard usage in the present study is about features that create extra explicitness, e.g. preand post-dislocation. Although pre- and post-dislocation are considered incorrect in traditional grammar books (unlike Heads/Fronting and Tails), considering the communicative role they play in making the

5.3 Practical implications 195

message explicit to the listener, teachers do not necessarily need to correct such usage and penalize learners. Another example of such usage is repetition. Although repetition is generally considered undesirable (Lichtkoppler 2007: 39), it does contribute to the communicative effectiveness of spoken English by fulfilling important functions in ELF settings (Lichtkoppler 2007: 46) and is considered an accommodation strategy used for efficiency and cooperation purposes (Cogo and Dewey 2006: 70; the present study, section 4.2.2). Repetition does not necessarily signal disfluency. Instead of penalizing learners because of the repetition in their spoken production, teachers can pay more attention to the types of items that are repeated and for what purposes these items are repeated. In the present study, we have seen three types of repetition, only one of which is repetition due to disfluency (section 4.2.2). Repetition for emphasis both in dialogic and monologic speech, and other repetition in dialogic speech help learners achieve communicative effectiveness. In short, ELF speakers use the language creatively in ways that are communicatively effective. It would be unjust to penalize speakers for usage that is communicatively effective. As Jenkins put it, “It is unacceptable to label creativity as ‘L1 transfer’ (or more pejoratively, ‘L1 interference’) on the grounds that it results from contact between an individual’s L2 English and their L1, or from contact between L2 English speakers who have different L1s” (Jenkins 2007: 17). • The inclusion of different types of speech events in testing spoken production: Dialogic speech gives us more information with reference to disturbance, pragmatic strategies and how speakers negotiate meaning in interaction. If teachers aim to test a learner’s ability to negotiate meaning and achieve communicative effectiveness through appropriate strategies, they should assign learners dialogic tasks. This would enable teachers to test the pragmatic strategies that they have exposed their students to (the present section, see priorities in teaching). In most language teaching classrooms, however, learners are assigned monologic tasks, presentations being a popular one. It is not fully possible to test how meaning is negotiated in prepared speech, e.g. oral presentations. It is recommended that both monologic and dialogic speech events are used in the testing of a learner’s spoken production. This would be an attempt to mirror authentic ELF situations in a language classroom setting, and it would be useful in preparing the learner for dialogic speech events. !

196 5 Theoretical and practical implications Overall, in language teaching and testing, we need to prioritize what seems to be important for the specific group of learners at hand. Since it is never possible to teach the entire language, it is crucial to prioritize. Accuracy in production is another matter. It is certainly not irrelevant, but if it is effectiveness one is after, then it is not what matters the most. What I am suggesting then is, if we want to prepare speakers for ELF settings, we need to give them the tools by which they can achieve communicative effectiveness. Spending valuable classroom time on trying to achieve accuracy of grammatical structures which do not even seem to be necessary for successful communication might not be the best use of our, or students’ for that matter, time and effort. In this monograph, I have presented the results of a study of communicative effectiveness, and the ways in which the ELF speakers achieve this effectiveness were not always in line with native speaker usage. Native speaker model is surely a valuable model, especially for those who need to learn the language mostly to communicate with native speakers, but it is neither a realistic aim nor an attainable target for millions of learners of English in the world today. The best service we can do for learners of English then is to first raise their awareness that English is a living and therefore changing phenomenon, and then provide attainable goals and useful models for them that are well adapted to the reality of English today.

Chapter 6 Looking ahead

In this monograph, I aimed to provide an unprejudiced description and analysis of English as a lingua franca in an academic engineering setting. The study that forms the basis of this monograph has contributed to the knowledge in the field by providing a detailed account of ELF as the medium of instruction. The study elaborated on form, communicativeness and attitudes in a lingua franca setting. I have argued here that the results are likely to be representative of similar international universities and are therefore applicable to such academic settings. The choice of methods, i.e. a combination of extensive and intensive analyses, has yielded data that allowed generalizations of usage (extensive analyses) and provided specific information on frequency of the investigated features and strategies (intensive analyses). In addition, the inclusion of both monologic and dialogic speech enabled comparisons where appropriate. In this final chapter, I will discuss further implications of the results, namely implications for native speakers as well as an additional discussion on the internationalization of higher education and future language policy work, hence the title of the chapter. However, before I do so, at this juncture, it seems appropriate to summarize the work I presented in each chapter so far.

6.1. Summary In Chapter 1, I provided the reader with the background of English today, starting with notes on the original lingua franca, then moving on to a historical perspective from early uses of English in the nineteenth century to how it became the language of science and technology. Concerns were raised when English gained such a powerful position; however, its growth continued as a consequence of globalization, which showed itself in many domains, among them, higher education. The rest of the chapter dealt with scholars’ reactions to globalism in Europe and Sweden, and what the effects of such a widespread use of English could be for local languages and multilingualism. The chapter concluded by arguing that such widespread usage of English does not presuppose linguistic imperialism as long as its

198

6 Looking ahead

use by its non-native speakers is validated and legitimized. This requires a thorough description of the way English is used by its non-native speakers from different L1 backgrounds in international settings, when they use it as a vehicular language. English as a lingua franca is this vehicular language. In Chapter 2, I aimed to give a detailed account of the research carried out in ELF until the time of writing, with the main sections organized in terms of normative elements. The studies in pragmatics from ELF settings were carried out without norms. The very nature of pragmatics as a field makes it very hard if not impossible to apply norms to usage. There are, however, other fields where there are clear normative elements, i.e. phonology, phraseology and morphosyntax. Phonology is a fairly closed system, which easily lends itself to description and the use of norms. Phraseology and morphosyntax are not closed systems like phonology; nevertheless, there are normative elements that have been applied to them. The chapter gave a review of prominent studies in these fields with respect to ELF. I have chosen to review the studies that are of highest relevance to the present monograph rather than summarizing all ELF studies so far. Finally, attitude and irritation studies that form the basis of the survey in the present study are elaborated on. I started Chapter 3 by describing Swedish higher education as it is at the time of writing with the most recent developments, such as the new tuition fees for incoming students. Rather early in the chapter, I gave the research questions and hypotheses, pointing to the need to carry out the present investigation. Following this, I discussed the methodological considerations in detail and explained the methods used in the several stages of the study, from its starting point to the final analyses. I used a combination of intensive and extensive analyses to allow for both generalizability and detailed analyses. The ten stages of the study were subject and speech event selection, recording and listening with a protocol, selecting and categorizing speech events, identifying non-standard usage, categorizing non-standard usage, selecting features for further study, checking for overt disturbance, non-standard morphosyntactic usage vs. standard morphosyntactic usage, analysis at the discourse level and finally, a survey on perceived communicativeness and attitudes. In Chapter 4, I presented the results from all the sections of the investigation, starting with the results in the form. The results in the form dimension of the study confirmed that form alone does not cause much overt disturbance. Among all the non-standard usage, only non-standard question

6.1 Summary

199

formulation caused overt disturbance in communication. The results showed that there were three cues to registering a question: The interrogative adverb/pronoun (in Wh-questions), syntax and question intonation. In the cases analyzed by means of Praat, when syntax was non-standard and failed to act as a cue, question intonation aided the communication process. When both syntax was non-standard and intonation was not rising, there was disturbance. The findings also seemed to indicate that, regardless of the way native speakers make use of question intonation, rising intonation seemed to help speakers in lingua franca settings. The types of overt disturbance that were likely to have been caused by extra-linguistic factors were dealt with separately, and examples of such cases were provided in the chapter. The findings in the communicativeness dimension showed, first of all, that form could not be considered on its own without any anchoring to other dimensions such as the discourse level. One of the very probable reasons why there were so few cases of overt disturbance was the proactive work carried out by the speakers through pragmatic strategies. It is highly likely that this type of proactive work prevents a large part of the potential disturbance here, and there is support for this in the literature (section 4.2.2). This indicated that without a sufficiently frequent use of pragmatic strategies, lectures are potential minefields where students do not get the opportunity to negotiate meaning and check their understanding. The last section of Chapter 4 provided the reader with the results of the survey on perceived communicativeness and attitudes. The overlap between the non-standard formulations students claimed were not comprehensible and the ones they were irritated by indicated that the survey had been effective in measuring attitudes. Attitudes can play a significant role in the communication process since negative attitudes can adversely affect communication. I revisited the research questions and the hypotheses in Chapter 5, followed by two main sections on the discussion of the findings and the theoretical and practical implications. The theoretical implications revolve around the status of ELF in general, elaborating on four of the most debated questions raised about research in ELF are answered, i.e. whether ELF speakers can be regarded as learners, whether ELF usage is learner language, whether ELF usage is sui generis and whether ELF is a variety (section 5.2.1). The answers to these questions have been obvious to ELF researchers for long, but the answers must be supported by more empirical data, which I have done in the present investigation. A fifth and very im-

200

6 Looking ahead

portant question, namely where ELF speech should be sought, is intended primarily for researchers who wish to undertake studies of ELF usage, with the aim of conveying to them that ELF usage can be found in ELF settings and not in English language classrooms where the aim is to learn the language, which makes accuracy more important than it is in ordinary ELF conversations. Finally, I touched upon what the notion of ‘good English’ may mean to those in lingua franca. In the last section of the chapter, I discussed the practical implications of the findings for those who operate in ELF settings. Among these are lecturers and students as well as language teachers. The implications for these groups are discussed in the given order. First, recommendations were made with regard to effective-lecturing behavior and ways of helping students in handling the challenges of English-medium education where they need to use ELF. The next discussion focused on ways of supporting lecturers and students, followed by a final section on issues to be considered in language classrooms. With this book, I hope to have contributed to the understanding of ELF. Possibly the most important contribution of the book is in the area of defining the effective speaker and the notion of communicativeness. My aim has been to provide a detailed account on the notion of the effective speaker in academic settings, showing clearly that it is pragmatic ability and not necessarily high proficiency that makes a speaker effective. From the results, the notion of effectiveness in academic ELF settings emerges as being determined primarily by pragmatic ability (function) and less by proficiency (form). This suggests that, in similar ELF settings, being pragmatically fluent speakers is much more effective than simply being proficient at the language. This finding, along with the other results of the study, are of relevance, primarily to content teachers and students who operate in ELF settings, but also to decision and policy and other high-level decision makers with the responsibility to prepare language policies and curricula, and also to language teachers who provide ESP/EAP training in ELF settings. In this monograph, I have recommended the following actions for these groups: •

Content teachers need to equip themselves with the strategies needed to achieve effective communication and structure their lectures accordingly. They need to be made aware that effective lecturing skills are not in direct proportion to being highly proficient in the language.

6.1 Summary

201

•

Students need to learn about ways of becoming effective communicators. They need to be made aware that this can be achieved regardless of their L1 backgrounds.

•

Policy makers need to broaden their horizons on the needs of content teachers and students in these settings. They need to be realistic when it comes to their expectations from both teachers and students.

•

Decision makers need to seek ways to provide content teachers and students with the training they need.

•

Curricula planners need to realize that time needs to be spared for the training of content teachers and students who need to cope with the everyday realities of ELF settings.

•

ESP/EAP teachers need to set realistic goals for those who operate in lingua franca settings. First come functionality and the strategies one needs for functionality, then language complexity. It is important to set spoken norms and standards to spoken English and not continue with evaluating speech against written norms and standards. We have long been demanding that language teaching conforms to spoken language, but our image of spoken language keeps changing. Language teaching must be kept up-to-date with this changing image of spoken language.

The results of the study in the present monograph, and the results of ELF research in general, have implications for native speakers of English and for internationalization and language policy practices. I will discuss these in the following two sections.

6.2. Notes for native speakers There are additional implications of the results here, and results of ELF research in general, for native speakers, which I have chosen to discuss separately here. Issues regarding the assumed native speaker authority has been discussed thoroughly before in ELF literature (e.g. Jenkins 2007: 36,

202

6 Looking ahead

37; Seidlhofer 2011: 33–38). Among relevant issues are the ownership of English (e.g. Haberland 2011; Seidlhofer and Jenkins 2003; Wee 2002; Widdowson 1994, 2003: 43), the status and role of non-native speaker teachers (e.g. Llurda 200420, 2009) and making non-native speaker teachers aware of their non-native assets, enabling them to use their assets in teaching (Seidlhofer 2003). It is no news to ELF researchers that native speakers may need to adjust themselves and their English to be able to communicate with non-native speakers effectively, but this no longer seems to be limited to a small group of researchers any more. Today, issues around native speakerism have started finding space in major media outlets as well, such as The New York Times. Subsequent to an essay written by Lawrence Summers, the former president of Harvard University, the NY Times ’Room for Debate’ section published a series of responses, in the form of short articles. Lawrence had said in his essay that, because English had become a global language, there was little need for American students to learn foreign languages to prepare themselves for international experience. This caused some controversy, and brought about discussions on multilingualism and the state of being the native speaker of the global lingua franca English. In his response with the title ‘Advice for native speakers’, Michael Erard, the author of Babel No More, included the following about native speakers in international settings and English as a lingua franca: Whether or not you think learning a language other than English is valuable, it's true that English has become the language of international communication. But that doesn't let native English speakers off the hook. In order for them to really benefit from the status of English as a global lingua franca, they still have linguistic investments to make. (Erard 2012)

20. Llurda uses the term English as an International Language (EIL).

6.2 Notes for native speakers 203

In his article, Erard also included some adjustments native speakers would need to make while communicating in international settings:

So if you're a native speaker of English [and that's your only language] accommodating your ear and your speech to those users of the global lingua franca is a cultural and linguistic skill. It requires practice; you don't automatically get it by virtue of being a native speaker. You have to learn how to hear around accents, word choices and grammatical patterns. You have to learn to suspend your judgments of what may seem like deviations and errors, because in a particular variety of English, those things may be perfectly acceptable. You also have to understand how other people's mother tongues, educational systems and community histories influence the English they speak. (Erard 2012)

What Erard says is in agreement with the findings of the study that forms the basis of the present monograph. The speakers in the academic setting described here use English as a lingua franca effectively. Among the subjects, there are speakers who are very advanced, and others less so. They come from nineteen different L1 backgrounds (Chapter 3). Yet, despite their diverse background (e.g. L1s, level of proficiency, cultural background), they know how to adjust their own linguistic patterns appropriately to get the job done. In such international settings, they seem to learn very quickly– if they did not know it already– which linguistic patterns to use and which ones not to, how to express themselves and how not to, and how to repair if disturbance occurs. None of the subjects here happened to be native speakers with the exception of one North American student.21 They have a wide range of non-standardnesses in their speech. Yet overt disturbance occurs very seldom. The speakers can complete the tasks assigned to them through the medium of ELF without serious difficulty. The academic setting described here is representative of any international academic setting. To be able to communicate effectively in such a setting, a native speaker would need to behave accordingly and make the necessary adjustments. These adjustments would not necessarily be simplifying their speech patterns and speaking ‘easy English’ as one might think, since there are some highly proficient ELF speakers. These adjustments

21. Native speakers were not excluded from the corpus in the present study. Only one subject happened to be a native speaker in the courses and classes I visited.

204

6 Looking ahead

would entail however adjusting one’s ears to new phonological patterns (rather than imitating native speaker patterns) (Jenkins 2000) and “hear around accents” as Erard says (see Jenkins 2000: 227 on pronunciation learning for native speakers of English). In addition to accents, native speakers would need to adjust themselves to new “word choices” (see e.g. Metsä-Ketelä 2006; Pitzl 2009) and “grammatical patterns” (see e.g. Björkman 2008a, 2008b; Breiteneder 2005; Cogo and Dewey 2006, 2011; Hülmbauer 2009; Hüttner 2009; Klimpfinger 2009). As Erard says, native speakers would need to “hear around accents, word choices and grammatical patterns”. Most importantly, they would need to “suspend their judgments”, as Erard adds very rightly, because what one might dismiss as errors quickly may be perfectly acceptable variants of the language in its sociolinguistic setting. None of this is received “by virtue of being a native speaker”. It is acquired with practice and effort, just the way a foreign language is acquired. In fact, it seems like native speaker students in international higher education settings “may be at more risk in the future”, unless they manage to adjust to these international settings, as described here (Jenkins, Cogo, and Dewey 2012: 302). Once one accepts the fact that native speakers also need to practice communicating in ELF settings and put hard work into achieving communicative effectiveness, there is perhaps less authority for native speakers to assume. Similarly, there perhaps is not much need to worry about native speakers having the upper hand (but see Van Parijs 2011 and section 1.3.2 here). It has been suggested that linguistic injustice arises when those who have English as their mother tongue get the lingua franca for free with no effort while non-native speakers of English need to acquire it by spending much time, energy and money (e.g. Van Parijs 2011). This is of course true when it comes to accuracy and fluency. However, if one thinks native speakers are always at an advantage because they are more accurate and fluent than non-native speakers, then one does not take into consideration the fact that native speakers need to adjust to ELF settings. So, native speakers are not at an advantange in ELF settings only because of their native-speaker status. This is true for the type of ELF setting as well. In academic settings, for example, one needs to use academic English with its conventions. Academic English is no one’s native language (Bourdieu and Passeron 1994: 8). Academics who need to communicate through the medium of English, native or non-native, need to strive for academic literacy, which must be acquired with much serious practice, time and effort.

6.2 Notes for native speakers 205

The discussion here with reference to native-speaker does not only show that native speakers also need to work at communicating effectively in ELF settings, but it also indicates, once again, that the native/non-native dichotomy is not a satisfactory criterion when judging linguistic competence.

6.3. The internationalization of higher education and language policy practices The results here have implications for the internationalization of higher education and consequently, language policy work. To start with internationalization of higher education, we first need to reach an understanding of ‘internationalization’. The dictionary definition of the word is simply “existing, occurring, or carried on between nations” (Oxford dictionaries). But what does internationalization mean in the context of higher education and what is required for a university to call itself international? Scholarly interest in the notion of international university has been focused primarily on internationalization being synonymous to Englishmedium universities (see also section 1.3). Naturally, such concerns have been raised in countries with smaller languages, Denmark/Danish being an example. To Preisler, ‘international university’ “does not equate with the uniformity of across-the-board use of English language and cultural heritage, but rather with the practice and utilization of cultural and linguistic diversity, even in the face of Anglophone dominance” (Preisler 2011: xiii). In the introduction of a volume of collected works on language and learning in the international university, Preisler suggests that internationalization “should not be dealt with in simplistic terms of traditional polarity between, on the one hand, local practices and the use of national language; and, on the other hand, the invasion (in EFL countries) of an international lingua franca, English…” (2011: xiv). Haberland and Mortensen make similar points in the introduction to their journal special issue on the topic ‘Language and the international university’: […] we believe that it is important not to overestimate the role of English and to acknowledge that there is more to university internationalization than mere Anglicization. Globalizing processes have indeed led to a situation where English has attained an unprecedented degree of globality. But an unprecedented degree of globality does not mean total globality. English is not spoken in every corner of the world, just in more places than any

206

6 Looking ahead

other language ever before. Estimates about the total number of speakers of English worldwide vary, but even the least conservative estimate of 1.5 billion English speakers falls short of the total figure for the world population. (Haberland and Mortensen 2012: 1, 2, original emphasis)

The view represented by Preisler, Haberland and Mortensen is shared by a number of scholars, as summarized in section 1.3 (e.g. Ljosland 2011; Phillipson 2006) who are concerned about English threatening smaller languages and multilingualism. I will argue here that English, when it is used as a lingua franca, does not necessarily constitute a threat to the local language or an obstacle in the way of multilingualism. There is empirical support for this from a university setting in Sweden, showing that the local language comes into play when needed. Söderlundh’s investigation (also see section 1.3.3) of English-medium courses shows clearly that Swedish is used when there is a need to opt for it (2010: 222): Overall, the study shows that Swedish holds a strong position on the courses observed. In quantitative terms, the language occurs rarely in wholegroup teaching, but it nevertheless permeates the course environment as a whole. To generalise somewhat, English as a language of instruction means, in practice, that lectures and common course activities take place in English. In other study activities, the choice of language is guided by basic socio-linguistic factors, and Swedish is used spontaneously by lecturers and students in interaction with other Swedish speakers. (Söderlundh 2010: 222)

It is clear from Söderlundh’s study that no matter what the global decisions are, local needs prevail and the local language is used along with English. The setting Söderlundh investigated is a typical international university setting. In this regard, there is little or no reason why her findings should not be representative of other ELF higher education settings. If speakers can make local choices and speak other languages than the common lingua franca, then there is also little reason to worry about a so-called English as a lingua franca “invasion”, as Preisler calls it (2011: xiv). English itself should not pose a threat to the local languages or lingua francas, because it simply is not used as a vernacular: English has spread in continental Europe, and in post-colonial Africa and Asia typically as a by-product of (other) primary globalized ventures, which can be accomplished in any other major language. Moreover, it is functioning in these territories primarily as a lingua franca, not as a vernacular. Therefore, it does not endanger their local vernaculars, because it

6.3 Language policy practices

207

does not compete with them, even while it is being used as the medium of higher education for some subjects and their scholars are encouraged to publish in English. […] The myth of English as a “killer language” par excellence, so much repeated in the linguistics literature on language endangerment, is certainly not supported by the facts of language practice (Mufwene 2006: 124, emphasis added)

The possible endangerment of a language as a lingua franca should not affect its use as a vernacular, as English is “spreading as an economically powerful lingua franca, hardly ever as a vernacular or in any way that would place it in competition with the indigenous vernaculars or lingua francas, whose domains of usage do not overlap with its” (Mufwene 2006: 125). In this sense, there is no reason why English as a lingua franca alone should impede linguistic or cultural diversity. It is perfectly possible for English and other languages to “co-exist […] in a composite repertoire of linguistic resources” (Seidlhofer 2011: 153). This is possible, however, only if English is liberated from a mere ENL status. It is the lingua franca status of English that allows for such a co-existence. And when one registers that ELF is English used “in its own right”, and that it has “interdependent status as a complementary component in a plurilinguistic repertoire” (Seidlhofer 2011: 153), one sees that ELF is not a threat to multilingualism. ELF is under the umbrella of multilingualism. English, when used as a lingua franca, “is no longer viewed as connected to the culture of the traditional English-dominant countries” (Baker 2009: 570). If “English-native-speaker pronunciation, grammar, and vocabulary choice are inappropriate in lingua franca settings”, which they are (see literature review in Chapter 2), then “native-speaker cultural assumptions and frames of reference are equally out of place” (Baker 2009: 567, 568). In fact, ELF can be a tool to express culture and identity (Baker 2009), unlike the claims made in early ELF research that ELF is neutral in terms of culture and identity (House 2003a; Meierkord 2002; as discussed in Baker 2009: 586). Baker’s results support Jenkins (2007), that speakers regard ELF as a platform to express and possibly create “cultures and identifications” (Baker: ibid). Time, effort and resources are better invested in finding ways into effectiveness when English is used as a lingua franca and in helping those who will operate in ELF settings. Of great relevance to this matter are language policy documents. Language policy documents should mirror ground-level reality, and this reality is language practice.

208

6 Looking ahead

It is not possible to refer to specific language policy documents here without compromising the anonymity of the present work and violating ethical rules (see section 3.3), so the criticism and/or recommendations I will make here will have to be of a general nature. What I will discuss below apply to the language policy documents mainly in Scandinavia, as it is Scandinavia that is of immediate relevance to the present monograph. The first issue that has been problematic in some language policy documents is their protectionist nature regarding the local languages in question. Unfortunately, since the first language policy documents started to appear in Scandinavia, they have been treated, and subsequently referred to mostly as documents written to provide guidelines for the protection of the local language from English. The common view has been that the presence of a language policy document at a university manifests the steps and measures it has taken towards protecting the local language. The policy documents in Scandinavia are generally detailed with regard to how, where and to what extent the local language should be used, thereby making an effort to ensure the use of the local language in as many domains as possible within the university. Naturally, it is a good deed to use the local language where possible and take the necessary measures towards keeping it as one of the academic languages available. The issue that has been largely overlooked is how, where and to what extent, and most importantly what type of English should be used. In general, a very marginal section of these policy documents focus on where and how English should be used, and more problematically, no policy documents with the exception of two mention the use of English as a lingua franca at all, or specify what type of English is considered appropriate and therefore acceptable for use in the university setting in question. A third issue I regard problematic is the way multilingualism is discussed in language policy documents. A number of universities have set themselves the goal to become multilingual universities. What is perhaps too idealistic, however, is to expect several languages to be used in settings where this might be unrealistic, e.g. a classroom setting when the aim is mutual comprehension. If the aim is for everyone to be able to communicate and share information, and by doing so completing the tasks assigned, then there must be a lingua franca that is understood by everyone, even though it might be at varying degrees. English is this lingua franca, and “no other language, whether at European level or worldwide, gives any sign of rivaling English as a universal lingua franca, and the snowball effect cur-

6.3 Language policy practices

209

rently unfolding is such that no language will ever do” (Van Parijs 2011: 22). Speakers are still free to speak the language of their choice, depending on with whom they want to communicate and for what purposes, and as Söderlundh (2012) showed, they seem to be able to do so. However, expecting a lecturer to deliver subject matter in more than two languages or using more than two languages in a testing situation is unrealistic. So, while multilingualism is certainly a goal to work towards, I believe realistic goals must be set as to what extent and in what contexts one can aim for the use of more than one or two languages. There are several ways in which universities can create opportunities for several languages to be used. A first step is to make room for elective language courses in the curriculum, both general and for specific purposes. This is crucial even at technical universities that might not see languages as central to their main goals. Certain domains might have a dominant language if that domain has an area of expertise in a certain geographical area. Vehicle engineering is an example of such a domain; France has had considerable expertise in this domain, and for this reason among others, engineering students have a reason to study French. Chinese has become an important language of scientific and technological development, so again, engineering students have reasons to study Chinese. It certainly takes time, effort and resources to learn to use a language effectively (see Van Parijs on estimating the cost of language learning, in Van Parijs 2011: 71), but even a low degree of proficiency may open doors for future study or work purposes. Universities can also strive to have strong ties with centers of excellence in other countries, and work on student (in this case, especially outgoing students rather than incoming students) and staff exchanges. To sum up what I have discussed here on the topic of multilingualism in relation to language policy, while it is certainly desirable to be competent in several languages, one should set realistic goals as to in which settings and to what degree other languages can be used. Language policies need to consider ground-level reality when setting goals for language use in the international university. Finally, relevant to what I have discussed above is the relationship between ground-level reality and language policy rhetoric. It is also a common (mis)conception that language policy documents should aim predominantly to outline the ‘desired scenario’ and that it should be up to the departments of that university to consider the ways in which this scenario can be applied. There should be general guidelines as to the applicability of this designed scenario. Language policy rhetoric must take into account what is

210

6 Looking ahead

practiced and practicable in university settings and design applicable solutions rather than focusing on utopian aims. Spolsky makes a distinction between language practices, i.e. what people actually do, language ideology, i.e. the beliefs about language use, and finally, any effort to modify language practice (Spolsky 2004: 5). Language policy documents in Scandinavia have focused on the third component, i.e. efforts to modify language practice rather than focusing more on language practices and considering how people use the language to cater for their needs. In fact, it is language practices that are true language policies (Spolsky 2009: 5). As Mufwene maintains, very rightly in my opinion, languages “are tools which enable their speakers to adapt to ever-changing ecologies rather than from the perspective now dominant in linguistics, that speakers should serve their languages and preserve their integrity” (Mufwene 2006). If we consider this, then instead of focusing on how we can dictate people to use one language over another, we should focus on what language choices they make for what purposes. English is a useful tool as a lingua franca, which enables those who speak it to perform a variety of tasks, depending on their needs. Language policy documents, through the help of relevant research, should observe how speakers get their needs catered for, and what choices they make, rather than making top-down decisions without a careful consideration of language practices.

6.4. ELF: Final remarks The widest and the most important use of English today is its usage as a lingua franca. The number of speakers who need to learn English to use it as a lingua franca is only going to increase, which brings the need to have thorough descriptions and analyses of ELF usage from different settings. The academic world is only one of these settings, albeit one of the most influential ones since it is through academia that research and knowledge are disseminated. If we wish to prepare speakers, both lecturers and students, for the activities they need to carry out in English in their everyday contexts, we need detailed descriptions of English that mirror its real use (Mauranen 2006c: 148). It is true that work on academic ELF is now established as a sub-field of ELF with very interesting research projects that have already been carried out and are going on at the time of writing (e.g. ELFA projects); however, we are far from done. As long as the internationalization of higher education continues, and every piece of evidence tells us

6.4 Final remarks

211

it will, there will be need for more research to understand the dynamics of such academic settings. It is my true hope that this study has taken us further in realizing the importance of function in academic speech. I am happy to be ending the present monograph on a positive note. The field of ELF is thriving with numerous interesting projects and individual studies, a large number of PhD studies from a range of countries is underway, there have been five international conferences on ELF so far, and radio programs in prestigious channels (BBC Radio 4), the Journal of English as a Lingua Franca (JELF) has just had its first issue released (at the time of writing), and the Developments in English as a Lingua Franca (DELF) book series is now in place. There is every reason to think that relevant descriptions of English as a lingua franca will continue to be made, which will be invaluable to its non-native speakers who will use it in ELF settings, to its native speakers on how to communicate in ELF settings, to those who need to teach English for international purposes, to textbook writers as to what they should present as the prescription and what weight they should give to functional and redundant features of the language, and last but not least, to the researchers in the field as to what other gaps to fill.

Appendices

Appendix 1: Sample lecture transcription

LECTURE: LF-040906 L1: SWEDISH NO OF SPEAKERS: THREE S1: SWEDISH, THE LECTURER S2: STUDENT, SWEDISH S3: EXCHANGE STUDENT, SPANISH NO OF PARTICIPANTS: 20 NATURE OF STUDENT BODY: MIXED RECORDING DURATION: 44:35 (ONE SESSION) DEPARTMENT: INFORMATION AND COMMUNICATION TECHNOLOGY <S1> should we continue? where were we we have executed two , no that’s wrong we haven’t executed anything yet but we have issued two complete i iterations of this this small loop , nothing has been executed really because the memory unit hasn’t delivered load data . and if you want to check this up in the book there’s a slightly different account of tomasola’s algorithm describing the same program with more like tables rather than figures and the previous slide shows one such table just like this part in the book that i am referencing , ok so we will start the third iteration of the loop and a third load instruction and there is again a free load buffer tagged number three and we will write that tag number into the tag storage near floating point register at zero so nothing sensational yet now we fetched the third , instance of the multiplication instruction and we can’t do this there are no free resources the multiplier has two reservation stations , there is no more pipeliner in the multiplier it cannot accept the third , source operant pair for later execution so we must stop here this instruction will stick in the instruction queue and since instructions er can only leave the instruction queue in program order later instructions will stop behind this multiplication instruction , so things stall for a while here until memory returns , with well i suspect it will be the first load value that we requested from address xx corresponding to the first iteration of the loop so we now sit for a while wait for memory, now we have waited for memory , the memory unit delivers data , which data where does it come from well it comes from address xx which is stored in a load buffer tagged with tag number one so the memory units (xx) and tag number one together do the data with the amount of tags we have in this , processor in this floating unit four bits are quite enough, one,

Sample lecture transcription 213

two, three and so on up to an including eleven four bits is enough to encode that so we can think of this as four additional bits , on the data base ok so we have data plus four additional bits signifying this is tag number one what happens , all units check those four bits against whatever tag value they have in their tag storage erm i think i should close the blinds /aa/ i couldn’t . yeah so the reservation stations here compare the tag number appearing on a bus with the tag numbers they have stored in the reservation stations the store buffers also compare tag numbers stored compared to tag number appearing on a bus and the floating point registers also compare tag numbers , stored with the tag number appearing on a bus and there’s only one match this reservation station has tag number one stored and that’s the same number as that appearing on the bus so this reservation station enables its right signal and reads data of the bus and stores it in the reservation station the other locations just sinently silenty ignore the value on the common data base it’s not for them so now we have two values in this reservation station so the multiplier can start calculation and in order to decide the way my slides should be presented i have to make some assumptions and my assumption now is that memory is slowest and (xx) floating point multiplication is not so slow so i assume that this multiplication will finish before the next read operation from memory so the next next the next thing that will happen will not be the memory unit delivering more data but rather the multiplier delivering a product erm yeah i have been a little sloppy with my slides i can see that tatatam that shouldn’t be there yeah so multiplication is finished , and the multiplier sends out the result onto the common data base tagged with tag number ten because that was the tag owned by the reservation station where these two source values came from , , you’ve already guessed this probably everybody compares tag number ten from the common data base with whatever number they have stored in their tag storage fields , ten does not equal three ten does not equal eleven so on however the store buffer here finds that the tag on the common data base matches tag number ten that it has stored , so the first store buffer will save data from the base into its data field and then this store buffer contains a value and then address so the memory unit can start executing this store operation , saving the result of the first multiplication to memory , hasty well . a few minutes ago i said that this multiplication had to stall there was no free reservation station left by the multiplier this condition no longer holds there is one perfectly free reservation station so you can issue this third multiplication instruction . into this reservation station , copying , the tag from floating point register zero and the value same as always from floating point register two and then the result will be written into floating point register four , so the processor overrides the previous tag value with the tag value for this reservation station as you can probably see, register f zero will receive no intermediate values loaded from memory except the last one and register f four will receive no intermediate values , from multiplication but the last one so apart

214

Appendices

from executing instructions potentially out of order the algorithm has renamed the very few floating point registers so that the processor can use a much larger set of registers consisting of , storage at various reservation stations for the functional units and at load and store buffers for the memory unit . it is noteworthy that instructions are taken from the instruction queue in program order and then they are potentially executed out of order in the execution part of the floating point unit and they are also completed out of order there is no reorder buffer in this floating point unit , meaning of course that this floating point unit had imprecise exceptions if a floating point instruction would cause an exception, then it woul would be impossible in general to say exactly which instance of the instruction if it is in a loop or exactly which instruction in a more general case that caused the exception , and if it can decide that this instruction caused the exception it might very well happen that several ins instructions ahead of the accepting instruction has executed and completed and updated its destination registers and it might also very well happen possibly at the same time that some or several instructions before the accepting exceptional instruction has not completed and will be aborted but you have now seen one way of er handling renaming an out-of-order execution you will have to mentally add a reorder buffer if you like to get precise exceptions precise exceptions means that it is possible to single out one instruction such that that instruction and all previous instructions have completed and no following instruction have , been issued or have been started , and that means that the program can be restarted from the instruction following the one i pointed out , there is an equivalent definition that states that er the instruction itself should not have executed but all previous instructions only and er , er , it is of practical importance only which one selects the one where the exception instruction is pointed out it has not been executed completely because it was it caused an exception or if one is to point out the instruction before the exceptional one then we have completed all instructions after the one we point out , yeah , mhm so that’s dynamic instruction scheduling , i would like to move on to branch prediction but presumably you have questions or some kind of matter you would like to discuss regarding instruction scheduling <S2> i have a question you showed the picture of the pentium four <S1> yes <S2> erm , the err what is it called alltså efter instruktionen , jag kommer inte ihåg22 <S1> this picture <S2> yeah , reorder buffer <S1> reorder [buffer] <S2> [where is that] , is that in this picture

22. I mean after the instruction I don’t remember (author’s translation).

Sample lecture transcription 215

<S1> where is the reorder buffer mhm it’s not there <S2> er , where should it be <S1> this this ah diagram looks like one of those where the reorder buffer is combined with the instruction window i’ll show you what i mean so if the instruction window and reorder buffer is combined into a single structure as in this illustration then instructions would , execute controlled by the combined structure and then the result would return into the combined structure to be committed in order what about the pentium four , this is the part of the diagram where resources are allocated and where there are some kind of scheduling from floating point instructions for integer instructions so so i would place the instruction reordering here but i can’t be sure <S2> so in that case the reordering is done before the integer, or what , cause there is no feedback on the x after that (xx) <S1> mhm well there is no feedback of hej there is no feedback of results er but , there may be control lines not shown that makes sure that the register files are updated in the in program order mhm i can’t say that it’s there but i would place it there that’s that’s the best of my knowledge any more questions on instruction scheduling ok so moving onto branch prediction the idea is that most conditional branches are skewed so that they are usually either taken or not taken and er if we could find out which for a particular branch we could use that to speed up program execution by executing instructions ahead of the branch terminology , if there is a branch instruction somewhere and then the next instruction in memory in sequential address order , is executed then we fall through the branch the branch is not taken if if however the branch updates the program counter to some non-sequential value then we take the branch and the simplest way to do dynamic branch prediction is this we will need all of the program counter perhaps save for two bits at the end if we always have aligned 32-bit addresses and the byte-address memory , but generally all of the program counter will be used to reference the instruction cache , we can use the same address to reference a prediction table which will perhaps have one bit taken or not taken and is perhaps because it’s not the usual case but it’s the simplest case , so , all of the instruction address goes to the instruction cache that’s around 32 bits and prediction table should have the same address that’s 32 bits and one bit of information for each address that will be two raised to the power of 32 bits that’s er that’s nearly a gigabyte mhm we can’t have a gigabyte of prediction table inside the processor no way so we ignore more than half of the bits of the instruction address and use the least significant bits ten bits perhaps to index this prediction table and then the table becomes a thousand bits total, which is completely manageable and manufactureable well , if the prediction table is not as large as main memory then there will be aliasing so that , any questions no if two instructions are exactly a thousand and twenty four entries apart in the prediction table they will use the same prediction bit mhm so we could add a tag the instruction

216

Appendices

cache does that the instruction cache will deliver (xx) codes to the instruction decoder it is important that incorrect (xx) codes are never delivered to the processor because then it will go away and execute instructions , mhm that should never exist so er there’s a tag that checks that er we get precisely the instruction from precisely the address we specify even if we get it from the instruction cache and not from main memory for the prediction table we need not have this , how come , well the worst thing that can happen is this case we are to be we are to execute this branch instruction but we’re looking at the prediction bit corresponding to this instruction because they use the same prediction bit in the table what can happen well the prediction can be the same no problem thank you or the prediction could be the other way so we get the wrong prediction what happens well we have to take the mispredict penalty of some cycles , and that’s all there’s no chance that we will get incorrect program execution because we reuse entries in the prediction table if we did similar things in the instruction cache or in the data cache we would get illegal program results like in this case we will only get a few cycles extra of execution time . a few cycles extra of execution time mhm that sounds ominous but on the other hand , this is a one bit per entry table if we were to add a tag we would add something like twenty bits on every entry multiplying the chip area of the table twenty fold , so we would have 20 kilo bits of table rather than one kilo bit and then we would have to reduce the size of the instruction cache or the data cache or both to find room for this prediction table and reducing cache size will almost immediately translate into lower performance because of more cache misses and more references to the slow main memory so this occasional er incorrect prediction that gives us a few extra cycle might very well be worth . having a slightly larger instruction cache and a slightly larger data cache for whatever we use the extra chip area for there is a snag with one bit prediction its memory is so short and , this little program is the easiest way to describe how one bit prediction can be wrong on the other hand many programs that you execute daily such as text editors they do rather rarely actually do loops (xx) like this but it’s an easy way to explain what can happen we start the first loop and no interesting things is in the are in the product prediction table we start the second loop there’s no interesting things but after a few iterations we are spinning around in the innermost loop and we are predictly correct predicting correctly every time and then the innermost loop finishes and that will be a misprediction because we will predict another iteration of the innermost loop , but that will not happen and then we will move to the outermost loop and we will correctly predict that there will be another iteration of the outer loop and then we will go into the inner loop again when we come here to the end of the inner loop we have changed the prediction because last time , the branch er wasn’t taken not up to here at least so we predict that the inner loop will not execute another iteration which will be a misprediction , so for each time we finish the inner loop we will get two mispredictions one when we believe the loop will continue but it doesn’t and then the next one when we believe that the loop will not continue but we’ll start it another round so it

Sample lecture transcription 217

will continue that’s the second one . and we will not expect the processor to find out exactly how many iterations will happen and will never make a misprediction but it would be OK if it mispredicted just once , when the inner loop finished and then kept the prediction that we will go around in the inner loop so it would predict correctly the next time we enter that loop , and that can be done using two bit prediction which requires two misprediction before it changes its opinion of what it should predict and this is a stake diagram , for every entry in the prediction table we would have two bits maybe coded just like this , to specify where where in this stake diagram er the branch corresponding to that location in the prediction table er should be considered to be . so if we have the inner loop again , that loop branch , if we take the branch to continue the loop we will say it has been taken and then taken and then taken and taken and taken for quite a while so we predict taken and then we come to the end of the inner loop where the loop branch is not taken but we don’t change our mind about the prediction the processor still believes that the branch will be taken this is just a temporary exception from the rule and then we go out into the outer loop and back starting a new round of the inner loop and the branch at the end of the inner loop will be taken this time and so we’re back here . having just one misprediction as we wanted in order to understand this diagram er you might want to have this loop preferably written in fortran because fortran has (xx) statements not just the semi-colon the semi-colon will be the same for both loops but the (xx) statements in fortran are similar languages will be different so there will be a next statement with the inner loop and then another next statement for the outer loop or you could do the same thing with curly braces if you , you like c and then you could see that er er at the end of the inner loop we will be here and the branch will not be taken once and then we come to the end of the inner loop it will be taken and taken again for the number of iterations so we’ll have only one misprediction when we believe it will be taken but we are at the end of one round of the inner loop so it will be not taken one time there is another no should i move on? <S2> yeah <S1> there is another stake diagram that is possible for two bit prediction which is called a saturate encounter algorithm . if we have the same loop nest and we believe no and we have the inner loop taken . we have the branch at the end of the inner loop taken when we continue to loop we will be at state number three in this example and then once the branch at the end of the inner loop is not taken so we go back here but we still predict taken because all states to the right of the limit predict taken , and then the next time the loop er the loop branch is taken so we come back here <S3> er question <S3> <S1> mhm <S3> this is the same but two different pictures or <S1> this is not the same but it is two different pictures , there is a limiting case that discerns the saturate encounter from the diagram in the previous

218

Appendices

picture , when two mispredictions actually happen for some other program not related to the loop nest we’ve been discussing for some other program we have a loop that is taken taken taken and taken again and then it’s not taken we still predict that it is taken we we believe that this is a temporary er change but then it is not taken again and then we believe that it will not be taken any more of course . now if the branch should be taken now it will requires it will require two mispredictions again before we change the prediction with the counter however , we have a branch that is taken and taken and taken and taken and then it’s not taken it’s not taken but in this case if the branch is now taken after exactly two mispredictions we will change our minds about the prediction immediately , so there is a change there is different behavior for this graph and the previous one , and er . some processor designers prefer the saturate encounter and some prefer er the other diagram , yeah <S3> so using only state one and two in that diagram would would correspond to one big prediction <S1> yeah er , it might happen that starting at some state or another the predictor will arrive here and then the branch is not taken so it changes its mind and then the branch is taken so the mind is changed back the prediction is changed back so it oscillates between predict not taken and predict taken and predict incorrectly every time it might happen on the other hand there is a similar situation for this state diagram where a branch is predicted incorrectly twice in a row , not taken not taken and then we predict it to be not taken and then it’s mispredicted again twice in a row because now it’s taken and taken so we change our mind and think ok it will be predict taken and then it’s not taken not taken so anomalous cases can happen for this diagram too , just like for the saturate encounter , but er the argument is valid that the saturate encounter does not always require two mispredictions to change the predicted er change the prediction yes . what to do with a prediction er . sorry well of course continue fetching and executing instructions at the predicted branch target or sequentially if the branch is predicted not taken , er before we take a break i might say a few words about the possibility to fetch both the instruction at the branch target and the next sequential instruction , that’s considered a bad idea and why is it a bad idea well , branches occur frequently so it will , very quickly happen that , the processor will have to branch another time . here , and branches are in general heavily skewed so if we can predict this branch with some reasonable accuracy we can be for instance , eighty per cent sure that that it the branch will be taken then it’s much better to use eighty per cent no to use a hundred per cent of the processor’s resources to pursue the path that is eighty per cent correct than to use fifty per cent of the processor’s resources on either path because if we use if we try to execute both paths we are guaranteed to waste fifty per cent of the resources we spend on that execution , and it will multiply so somewhere here we will be wasting seventy five per cent of the resources , yeah , now it’s time for a break

Appendix 2: Sample group-work transcription GROUP-WORK: BODA-050207 NO OF SPEAKERS: FIVE S1: SWEDISH (L2) L1: ARABIC S2: FRENCH (EXCHANGE STUDENT) S3: SWEDISH S4: GERMAN (EXCHANGE STUDENT) S5: SWEDISH RECORDING DURATION: 40:57 (ONE SESSION) DEPARTMENT: AERONAUTICAL AND VEHICLE ENGINEERING <S1> ok er welcome to the third session , and err we have to discuss our problem and we will start with the first one , er . why the case of the typical session , er is it in general that it is not possible to have analytical flutter results <S2> it’s my question and it was er raised page <S3> 53 <S2> yeah <S3> 53 <S1> 53 <S3> in the e [xample] or <S2> [example] the typical section . the second second sentence <S1> that’s what i asked myself <S3> mhm <S1> maybe [it’s] <S3> [(xx)] <S1> [to do] that <S4> [@@] <S3> no , maybe mode is quite arbitrary (xx) unconvenient actually flutter <S4> I had the same question my in my mind when i read that because it’s <S3> there is no reason <S2> it means that we can only have flutter results] with computers <S4> [yeah]

23. These sections are sections from the textbooks and therefore not speakers’ own production. No feature from these sections has been included in the analyses or the results in the present study.

220

Appendices

<S2> because it’s right that in part in the in the formal shelters we had some applications for the typical sections and we had we get some analytical results but here we have no analytical results but , i don’t know <S3> it’s very strange it’s very strange <S2> or we can continue in <S3> @@ <S1> @@ <S1> actually this is the second one strange <S3> @@ <S1> while the flutter velocity is lower than the divergence velocity what does that physical phenomena that occur <S2> mhm first i am i am not convinced that the flutter velocity is always lower than the divergence speed because i er (xx) the code for the flutter and it is err and when when you put a mass er a concentrating mass somewhere in the wing sometimes it appears that the flutter velocity is greater greater than the divergence speed , so it is not always er <S1> but it should be less , lower <S2> it’s lower if you do nothing but <S4> but where is written or you know why does the question come up why the flutter’s velocity is lower than the divergence velocity you see <S3> you see <S2> it’s not my question <S3> it’s not your question <S2> no <S4> it’s x’s24 question i mean er erm how you asked why but how did you (fx) or is it somewhere written in here that it’s <S2> is it written that it’s <SU> (xx) <S1> er yeah er this is an er remember the course the (xx) course <S2> in your <S3> ok <S4> the <S2> but i think it’s right because when you when you are in a plane you can see that the wing is is vibrating it’s not going so it’s right that <S3> mhm <S2> the flutter speed is less than divergence speed <S3> mhm <S2> in some cases if you do something with the wing it can be greater

24. The student’s name has been removed.

Sample group-work transcription

221

<S3> it’s not i don’t know flutter is somehow initial and (stable) (more) and the and the divergence is completely unstable <S1> yeah <S3> so i don’t know i i agree with you <S4> it should be <S3> that it should be lower <S2> but why <S3> why <S1, S3, S4, S5> @@ <S1> this is the (xx) <S3> yeah you see flutter is , on the real part everything is zero it’s flutter , that is not really unstable not really stable to be (xx) so , let me see now label <S4> @@ <S3> @@ <S4> i mean er this you said that flutter is some kind of phase shift <S2> mhm <S4> and er i think this kind of phase shift is something which is occurs earlier than complete divergence because it’s er the divergence aero aero the aerodynamic forces have to erm have to decrease the the stiffness of the wing somewhere the stiffness of the whole system so i think this is a very in this case the frequency somehow was zero something like that so er increasing the air speed i would say decreases the , the frequency , decreases the frequency decreases decreases and then if it’s zero the frequency then it’s divergence or something like that <S3> mhm is it <S4> or if no i would say maybe not zero but if the , if we can’t obtain any , any frequency any more because if K is zero or the stiffness is zero there is no no restoring force so you can’t have any <S3> yeah yeah yeah <S4> any frequency so this is somehow the lowest frequency you can imagine @@ so i i would say something like that you know because ok the wing is there is no airflow so the wing is like stiff then you increase the airflow and it becomes less stiff less stiff so the frequency the (air) frequency of the wing somehow goes down i think that divergence something like then it’s gone i mean the frequency is gone @@ so it’s somehow the , the last last of these things you can reach <S2> but we saw in the wing X that when their speed was increasing the vibrations were increasing also <S4> the vibrations but not the not the frequency maybe the frequency but i think the frequency did not i mean if it becomes stronger the wing has to be faster and vibrating but the frequency doesn’t have to go up , no if if the way if the , if it becomes you know if it starts like that <S2> mhm

222

Appendices

<S4> and then it goes like wooooww wooowwww then it has to move faster on on this part but the all over time doesn’t get shorter you know the all over time from from here to here of all oscillation i would say this <S2> xx <S4> i couldn’t see it by looking but i @@ yeah but i would guess so because he always argumented like if we increase the air speed the stiffness of the wing decreases so if the stiffness decreases i would say the frequency increa er decreases as well it’s my interpretation i am not sure <S3> divergence is a it’s it’s a steady state , flutter is , it’s non-stable instead of , is it really possible to <S4> to compare that i don’t know it’s just <S3> maybe suddenly the er the er unsteady motion , will trigger the er . ah i give up <S4> @@ <S3> @@ i think it’s hard to compare these two things because we , we look at them so differently <S3> yeah it’s confusing (xx) <S4> we try to separate them so that we can co compute them easily but then it’s hard to to compare them <S3> mhm <S4> actually err i can i think if we can turn this question to the question is the flutter velocity always lower than the divergence velocity , because we i think we can’t even state that <S3> very tricky questions in this group [@@] <S4> yes [@@] <S1, S2, S5> [@@] <S4> i think somehow we fucked ourselves <S1, S2, S3, S5> [@@] <S3> actually just <S4> i think we can just go on <S1> ok <S4> i think we can’t really bring it to the end [this question] <S3> [mhm] <S4> but it’s , anyway , i think we can just have some ideas but in the end we won’t get a result <S1> ok , question three <S2> (xx) the links of my question <S2> <S1, S3, S4, S5> @@ <S1, S3, S4, S5> <S1> i already , @@ is there a way to increase the flutter speed , in the lecture notes for the third computer session we are advised to introduce a concentrated mass on the wing i (wrote) the flutter go through to introduce the mass and i noted that putting a sufficient mass towards the leading edge made the flutter speed is larger than the divergence speed , if all i did is right er what are the physics behind the experiment , and if it’s wrong, this is the

Sample group-work transcription

223

mass should be towards the trailing edge my question remains what are the physics behind , the idea of introducing constant trailing mass so you mean that if you could immense on the trailing edge you would get a larger <S2> yeah yeah if i am if i am [right] <S1> [on the trailing] or <S2> if i am right the mass should be in the leading edge but if i am wrong it should be er put on the trailing edge but in both case why er er what is the physics behind this idea of putting a mass in front of the (xx) <S4> mhm <S4> <S2> (xx) and why does it increase the <S3> so if you put err if you put sufficient mass towards the leading edge the flutter speed larger <S4> would be larger <S2> if if your your your if your origin original wing and the flutter speed i formed was 15 m/sec <S3> ok <S2> and the divergence speed was about 20 and if i put er i don’t remember the exact mass but if i put a mass towards the trailing edge and towards the leading edge <S4> leading [edge] <S2> [the] the flutter speed is something like 13 m/sec <S1> 30 <S2> and the divergence speed is 28 29 m so it becomes larger but the mass i put is very large it’s it’s something like one kilo or less so it’s big <S3> ok <S4> (xx) in this <S1, S2, S3, S4, S5> @@ <S3> wow <S2> but it is the condition to have er a flutter speed er larger than divergence speed but i think in in for aircraft it’s er it’s er , never have guessed that you have divergence speed great er less than the flutter speed so erm i i put very big mass er which is half the mass of the total wing but i think in a real real case you can’t do that because you don’t want to have flutter speed er larger than the divergence speed but we we <S3> mhm <S2> are we are asking in the computer lab to to to find the the the minimum mass, which make the flutter speed greater than the diverge divergence speed and i did that and i found half the mass of the wing from that we have to put it in the (xx) but i don’t know why <S3> ok <S4> mhm <S2> that’s why i find it from this reason

224

Appendices

<S1> with with putting a mass do you mean that you made the wing more stiff <S4> no you just put a mass without any stiffness <S1> without any stiffness <S4> yeah you just you know just just like a something you you glue on it or you you stick on it or something <S1> ok ok ok <S2> (it) verifies the position of gravity between the <S1> yeah yeah [yeah] <S1> <S2> [the g j] alpha <S1> depends on if you shift it back (xx) <S3> so you move the center of gravity (move) forward <S1> mhm <S3> er <S2> so i tried and i found this results but i i didn’t manage to understand why . and maybe maybe the mass should be in the trailing edge but if it is in the trailing edge i i still don’t understand what what are the physics <S3> i don’t know i haven’t really found what behind the physics (xx) <S2> we can take question four <S1, 2, 3, 4, 5> @@ <S4> yeah i think it must have something to do with the er properties of the inertia forces which are with the phase shift of the inertia forces , they are always er are they 100 180 degrees behind the behind this you know behind the er acceleration and the acceleration goes up, the inertia forces points down , and er so i think it has it got something to do with that you know the wing is er . if it’s in the front of the wing then it it contributes in another another situation to you know it’s like er for example if the wing is just going up just accelerating accelerating up upwards and er then there is a bigger mass in the front this one will slow down and , so this is something which erm <S2> in fact i think it will decrease the angle of attack or something <S4> exactly that’s what i mean it decreases the angle of attack at the moment when it goes up so it’s somehow erm in this situation at x counter counterwise to this one so it tries to stabilize it at this moment you know <S3> mhm <S4> it accelerates up because of the aerodynamic forces which are coming from downwards somehow you know they they act upwards so if this one decreases then in the same time you know contemporarily then it’s really good for that you know and if it’s here <S3> increase <S4> exactly the (xx) i it’s just an idea <S3> mhm

Sample group-work transcription

225

<S4> yeah i think it has got something to do with it because it’s er , er the inertia forces are always 180 degrees shifted to the to the acceleration , but if you take a look at er erm the stiffness force and restoring force due due to stiffness they always count at 180 degrees shifted to the to they way to the deflection so it’s something else er erm these forces are always somehow a little too late you know what i mean if , this one already this the inertia forces already stop the the motion of due to to a to the acceleration so but there is another problem because if if , if the acceleration stops these will will err tend to to keep the motion going so it’s something , else which is a problem it’s more complicated than what i just wrote i think because you have to look at the frequency (xx) you have to somehow to compare when it acts and which direction and you have to have a look at the others as well as the stiffness and the stuff so i think it’s complex with <S3> the question laboration was , do you you have to explain why (xx) <S2> yes you are <S1> mhm <S2> it’s written can you explain the physics behind your condition <S3> (the answer is) [@@] . yeah that’s one possible explanation <S1, S2, S4, S5> [@@] <S3> erm <S4> yeah i think it would be quite interesting to see diagram of the like the side curves , and then you put in the side curve for the for the different forces as well and you can somehow see which one has got which effect i think <S3> mhm <S4> and then you should have to do er to think about the fact that we change the position of these so i think it has got something to do with this one somehow decreasing the angle of attack when it’s going up and the other way around you know so i think you have to take this one and a (xx) and an <S3> if it’s let’s see on the (out) flutter then it should . ah @@ yeah i go for this. gravitation of . the decrease of <S1> ok <S3> [what] <S1> it’s [complex] <S4> so we just gonna pass this question or <S1> yeah , question number four , how should the angle vectors (xx) (the intrepid) if they are complex valued <S2> phase shift it was written <S3> where <S2> the physical significance of the complex valued angle vector , is that the interaction with the , aerodynamic forces brings in phase shift , between the different degrees of freedom given (xx) , page 50

226

Appendices

<S4> i’ve marked it but i didn’t really understand it <S2> yeah <S4> i thought i would have i thought i understood it but i didn’t i didn’t really think that it’s about that one <S3> (xx) <S4> but you’re right yeah <S3> yeah but yeah but yeah . mhm [and er] <S2> [if you] if you run the run your code your flutter code you will you will have a , a picture and the screen and you will see them in the motion of the wing and you have two degrees of freedom fh and F alpha and when will you run it you will see that if h is doing that alpha is will do the same but with a phase shift if h is doing that it is difficult to to if h is doing that alpha will do that <S4> yeah <S3> ok <S2> so , (do) you understand what i mean , there is a there is a delay between the <S1> ok ok <S2> two degrees (xx) that’s my interpretation , i am not sure <S4> yeah i think you’re right <S3> it may be also , ah @@ <S4> and i think you can you can see that the value has become complex if you look at these <S2> mhm <S4> because erm if it’s not complex then it’s just er then you just get an amplitude for certain time and it’s always the same like that . this is this is no phase shift and you can erm and , this is non-complex , so if you take the phase shift into account it becomes like that and it’s complex (xx) <S2> without without (xx) of that <S4> ok <S2> or this means that . with five it means that you’re (xx) <S1> (xx) <S3> (xx) something from the er , the real part <S2> could you repeat this <S3> @@ , what , (xx) something from the real part (xx) is it magnitude <S2> yeah , the real part is magnitude and the imaginary part is the , so [the] <S3> [the the] <S2> the real part it means that for instance the <S3> [is it dis displacement] <S2> [do we have er] lots of displacement in er h and no displacement in

Sample group-work transcription

227

<S3> ok @@ <S5> this was the same question <S1> yeah <S2> four and five <S1> number six . the reduced frequency is obviously of significance but how should the value of the reduced frequency be interpreted and what conclusions can be drawn from it <S3> mhm yeah that’s my question i was thinking about er , it was said that it is the same as significance if er (use of) frequency as the (radius) number how to say i wonder if it is there is any er numbers where we can assume a steady state or (do) you know what i mean . or if it’s <S4> i don’t know i think i asked somehow something like that in the lecture , like i asked what kind of frequencies that if it doesn’t have any particle <S3> mhm <S4> but he er he he said it is a frequency , dimension is pretty easy he said @@ but he didn’t really answer er erm what it physically means , so <S3> yeah but what i understand it is it is the time it takes for one (xx) to travel (xx) <S4> yeah that’s what he showed yes <S3> and er compare it to the er freq yeah yeah the exactly . but . yeah , i mean is it for some , some rate is it . like i don’t know if if the rate is (xx) is it ok to assume , a steady state is it ok to does this er er unsteady motion er , computations are the really (xx) <S4> (xx) yeah <S3> great @@ <S4> yeah he said something like that yeah if something like if the particle travels slower than that happens then it’s somehow steady then <S3> it doesn’t feel that <S4> and you just erm won’t know if there’s a a certain number <S3> mhm <S4> like there you should take steady and you should [i think that’s something we have to pass on] <S3> [@@ it’s just laugh (xx)] <S4> ok <S3> yeah let’s <S1> i have difficulty to er to well understand the algorithm pk (and that’s it) can we discuss about that , then we have question eight , almost the same question what is the principle of pk method <S2> (this part) i don’t understand the idea is to find . the the idea of er of the code is you choose a k <S1> k a k

228

Appendices

<S2> and then you find p <S1> p , pk <S2> but er . you you are not interested in all the ps you want er the ps which are er which are subs(xx) that’s the imaginity imaginary part of p is equal to k <S1> ok <S2> ok so , you take one k , you are , so first when you we’ll do that later when you you could and you (serve) that for different case you are in dimension two so we will find each time two (xx) frequency so and you will see that when you will run the program you will have graphs like that i don’t know why but [yeah like that] <S1> [@@] <S2> [@@] so this is the first one , this is p1 , this is p2 . and er what you what you want is to find the p1 and p2 such that the imaginary part of p1 and p2 are equal to k so that’s why you draw this this part <S1> k <S2> so the the two solutions are this one and this one and the first part of the algorithm is to find the the region when you when you (xx) because when you when you apply the bisection method you have to to find difference to domain (xx) easier solution , so the first part of this algorithm is to find the domain where you have you your solution so , we will do that er er ima at at the beginning you have k which is er as a number it should be <S3> very small [zero] <S1> [zero] <S4> [zero] <S2> zero point zero one , and you find (this one), and you find first er the first first igon vector igon value is here <S4> it’s got the imaginary part up there <S2> and then you have to combine if er you have to see if the if the <S4> imaginary part , of the (xx) <S2> it’s like , erm , i’ll do like that and you run the code first for k is zero point zero one and then you run the code for instance for k is zero point one and you will find two solutions this one and this one and you have to see if if there is a crossing and if there is a crossing it means that you have formed your domain because this is the this solution and to see if there’s a crossing you have , yeah there isn’t much is (explains) so for k is zero point zero , one you have that the image the imaginary part of p1 , here is larger than k1 , because this this this distance is more than this distance ok <S1> mhm <S2> or you can see that like , you can see also that your point is er <S1> higher <S2> above than this er this (xx) <S1> mhm <S2> so in this case you have your imaginary you er erm erm

Sample group-work transcription

229

<S4> imaginary part <S2> imaginary part , which is length and then you continue and at the moment you will have a crossing which means that here in our case at k is zero point one we have imaginary part p1 which is less here which is which is less than k k1 , but it’s not the same k1 it here it is <S1> this is it <S2> zero point zero one and here it is zero point one so it means that you have crossed er this straight line and it means that you have formed your domain and your domain is you can say that k1 is between zero point zero one and zero point one so you have <SU> COUGH <S2> you have formed the domain and then you can apply the bisection (xx) which consists in divided in dividing new domain er by two (xx) so you can do that from the first step if you want i will do another , another sketch we can imagine that zero point one you can imagine that zero point zero one is here and zero point one , is here and what you want to find is that so the idea is you have the , the lower boundary is is this one zero point zero one and the upper boundary is zero point one <S1> mhm <S2> ok , so the first thing is that you divide your domain by two so k is zero point zero one plus two point one divided by two which is zero point fifty five <S1> mhm <S2> so we can do something like that [so you are] <S1> [zero point one] minus zero point zero one (xx) <S2> no <S1> no <S3> mhm <S2> you have to you have to er to take the the middle of your domain and taking the middle is that <S1> mhm <S3> mhm , is it <S2> you will you will understand [when i] when i draw it <S1> [ok ok ok] <S2> so here you take the mid part the middle of this domain which is here , ok <S1> mhm <S2> and you see then you have to adapt your er your er your domain you have to er you see that now you have two domains , and in one domain you have no solution and in one domain you have one solution and you have to find in which domain you have the solution so er here er if so we are here now and you see that here in this part the imagi the imaginary part of p1 is more important , is [greater] <S4> [bigger] <S4>

230

Appendices

<S2> than k1 it means that here you have the imagi the imaginary part you have imp er imaginary part of p1 which is also greater than p1 so here and here you have in in both case imaginary part of p1 which is greater than k k so it means that you have no (xx) so this domain is not valid so you are in the other domain and you adapt , you update your your domain so now your k lower is this one zero point zero fifteen and the the the upper limit of your domain remains <S1> the same one <S2> and then you divide [by two er] <S1> [and so and so and so ok] <S2> and find the solution <S1> fine <S3> these are <S4> here something written that the newton thing is much faster have you tried that <S2> er you have if you want you can try it you have er a (xx) function <S4> have you tried it <S2> i won’t try it because you have only two two (xx) newton of your program and it will do so comp computationary it’s yes it’s more efficient but you’ll (learn) nothing because you use you use (xx) function , it’s not very interesting i think . the (xx) programming is already done so you (don’t need to) apply (xx) function , or you can write your newton code if you want , we did we did it in [(xx)] <S3> [(xx)] <S2> so it’s not interesting for us to read it <S4> yeah <S2> but we didn’t do <S4> but you could probably (xx) the program <S2> yes you can but we didn’t deal with bisection so it’s it’s good to (xx) so i think we we didn’t do the bisection and it’s not asked to do in the in the [(xx)] <S4> [no no no no] <S2> (xx) it is not asked to do newton . i think it’s more simple but once we’ve already seen <S3> yeah it’s faster <S4> this one is more or less you know obvious what to do here , i mean this is more or less what you tend to do and you (xx) so i mean if you if you see it you will say like ok we just take the bound two boundaries and then we take the half of it and look what’s happening here so it’s somehow what you tend to do in the first so i think that’s why we are supposed to do it like that <S3> mhm <S4> ok so i would say this is answered <S1> mhm ok [question] [number] <SU> []

Sample group-work transcription

231

<S2> [question] eight , it’s it was about thirteen <S1> figure thirteen <S2> which page <S4> and this is very much like er <S1> fifteen <S4> automatic control <SU> (xx) <S1> automatic control yeah , it’s a , regular problem <S4> when you’re supposed to to draw the (xx) <S3> mhm <S4> i think it looks really much like that <S3> yeah yeah (we say so) <S4> @@ . i know who asked <S1> but the difference er the different between the these two <S3> ah ok <S1> what is the different , that you have more stable , which one is <S4> i would say it’s like we start at the point where the air speed is zero and if the air speed is zero you just get erm the normal vibration (xx) the (xx) frequency of the wing <S3> mhm <S4> as we did it as as you can show like that yeah <S3> mhm <S4> and that’s pure or almost pure erm vibration without any damping . we we don’t (xx) <S3> mhm <S4> so er then you get only imaginary parts like you get two different omegas two different frequencies one here one here only these two not this one and you can make a cross here and here , and if you now increase the air speed . then you can find out that one of these two (angles) will just go in the other direction ok you treat it a little it will be here you increase it a little just once and it will be here and here for example <S3> mhm <S4> you increase it a little more and it will be here and here a little more here and here so when it’ll be here and here , and if this one which is that point , then you found one one angle which , tends to be unstable because i don’t know if you remember from automatic control <S3> mhm <S4> everything which is in that (half plane) is stable <S3> stable <S4> everything which is here is unstable and everything everything which is here on this line is <S1> då är det klart

Appendix 3: Sentences used in the questionnaire Sentence 1 There is discriminization against women engineers in our society. Sentence 2 By spending 10% more on material, we can forsify the construction of the building. Sentence 3 It is important to try to levelize the volatility of the current market. Sentence 4 It is more easy to create a matrix B which creates the conditions we want. Sentence 5 It is much more clearer than you think. Sentence 6 Actually in Tokyo it is a more big problem. Sentence 7 How many energy is needed? Sentence 8 Each people will write his/her section. Sentence 9 Typically you want to have five kilogram of oxygen. Sentence 10 So there are two way of stating the same theory. Sentence 11 In many many case you can gasify it. Sentence 12 The job will be a lot easier if you have a extremely efficient compressor. Sentence 13 You can use it in the different ways. Sentence 14 From these figures, you can have idea what reasonable speed runner size is. Sentence 15 There is a further method which are sensitive to porosity in rocks. Sentence 16 However, the runner blades was not that developed. Sentence 17 I will talk about how a turbine operate in the system.

Sentences used in the questionnaire 233

Sentence 18 My idea is to explain how this board is working. Sentence 19 A Francis turbine is using the whole turbine equation. Sentence 20 How much rain are you getting per year? Sentence 21 And the plates get heat up very quickly. Sentence 22 I think it’s a rather huge project that built underground. Sentence 23 It can be happened that sometimes everything fails despite the formula. Sentence 24 Why is not good to combust directly? Sentence 25 Why the function looks like that? Sentence 26 But the drawback here it is not very easy to extract hydrogen from water. Sentence 27 The supercapacitors I don’t know much about them. Sentence 28 Well it is not so emission-free hydropower. Sentence 29 Here you see how does it look like. Sentence 30 Still we have to find out what sources do we have on this bus. Sentence 31 You get a feeling how is the cost developing for wind farms. Sentence 32 It is a not very good generator. Sentence 33 This point is supposed to not move. Sentence 34 It looks not good.

Syntactic

Interaction pattern

Time: 00:00

Activity type

Department:

Speech event: Lecture Group-work

Appendix 4: The observation protocol

D

ND

Phonological

D

ND

Lexical

Date: D

ND

References Adolphs, Svenja 2005 ‘I don’t think I should learn all this’— a longitudinal view of attitudes towards ‘native speaker’ English. In The Globalisation of English and the English Language Classroom, Claus Gnutzmann, and Frauke Intemann (eds.), 119–131. Tübingen: Gunter Narr. Aickin, Joseph 1967 The English Grammar. London: The Scolar Press Limited. Airey, John 2009 Science, language, and literacy: Case studies of learning in Swedish university physics. Unpublished PhD thesis. Uppsala University. Airey, John, and Cedric Linder 2006 Language and the experience of learning university physics in Sweden. European Journal of Physics 27 (3): 553–560. Albrechtsen, Dorte, Birgit Henriksen, and Claus Færch 1980 Native speaker reactions to learners’ spoken interlanguage. Language Learning 30 (2): 365–395. Allen, Bryce, Jian Qin, and W. F. Lancaster 1994 Persuasive communities: A longitudinal analysis of references in the philosophical transactions of the Royal Society, 1665–1990. Social Studies of Science 24 (2): 279–310. Alptekin, Cem 2007 Readers Respond. Teaching ELF as a language in its own right: Communication or prescriptivism? ELT Journal 61 (3): 267–268. Altbach, Philip 2004 Globalisation and the university: Myths and realities in an equal world. Tertiary Education and Management 10 (3): 3–25. Ammon, Ulrich 2001 English as a future language of teaching at German universities? A question of difficult consequences, posed by the decline of German as a language of science. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 343–361. Berlin: Walter de Gruyter. Aston, Guy 1993 Notes on the interlanguage of comity. In Interlanguage Pragmatics, Gabrielle Kasper, and Shoshana Blum-Kulka (eds.), 224–250. New York, NY: Oxford University Press. Audacity http://audacity.sourceforge.net/download/ Baker, Will 2009 The cultures of English as a lingua franca. TESOL Quarterly 43 (4): 567–592.

236

References

Bakhtin, Mikhail 1981 The Dialogic Imagination: Four Essays by M. M. Bakhtin. Michael Holquist (ed.). Austin, TX: University of Texas Press. Banks, David A. 2006 Audience Response Systems in Higher Education: Applications and Cases. Hershey, PA: Information Science Publishing. BASE. British Academic Spoken English corpus. http://www2.warwick.ac.uk/fac/soc/al/research/collect/base/ Berg, E. Catherine, Francis M. Hult, and Kendall A. King 2001 Shaping the climate for language shift? English in Sweden’s élite domains. World Englishes 20 (3): 305–319. Bex, Tony 2008 ‘Standard’ English, discourse grammars and English language teaching. In Standards and Norms in the English Language, Miriam A. Locher, and Jürg Strässler (eds.), 221–239. Berlin: Mouton de Gruyter. Biber, Douglas 2006 University Language: A Corpus-based Study of Spoken and Written Registers. Amsterdam: John Benjamins. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Fine gan 1999 Longman Grammar of Spoken and Written English. London: Longman. Bizzell, Patricia 1982 Cognition, convention, and certainty: What we need to know about writing. PRE/TEXT 3: 213–241. Bjørge, Anne Kari 2010 Conflict or cooperation: The use of backchanneling in ELF negotiations. English for Specific Purposes, 29(3): 191–203. Björkman, Beyza 2008 a ‘So where we are?’ Spoken lingua franca English at a technical university in Sweden”. English Today 24 (2): 35–41. 2008b English as the lingua franca of engineering: The morphosyntax ofacademic speech events. Nordic Journal of English Studies 7 (3): 103–122. 2009 From code to discourse in spoken ELF. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 225–252. Newcastle: Cambridge Scholars Publishing. 2011 Pragmatic strategies in English as an academic lingua franca: Ways of achieving communicative effectiveness. Journal of Pragmatics 43 (4): 950–964. 2012a Questions in academic ELF interaction. Journal of English as a Lingua Franca 1(1): 93–119. 2012b Review of Linguistic Justice for Europe and for the World, Philippe Van Parijs. International Studies in the Philosophy of Science 26.

References 237 Bloomfield, Leonard 1933 Language. New York, NY: Henry Holt. Boersma, Paul, and David Weenink 2005 Praat: Doing Phonetics by Computer. http://www.praat.org/ Bolinger, Dwight W. 1978 Intonation across languages. In Universals of Human Language, Phonology 2, Joseph Harold Greenberg, Charles A. Ferguson and Edith A. Moravcik (eds.), 471–524. Stanford, CA: Stanford University Press. 1989 Intonation and Its Uses. Palo Alto, CA: Stanford University Press. 1998 Intonation in American English. In Intonation Systems: A Survey of Twenty Languages, Daniel Hirst, and Albert Di Cristo (eds.), 45–55. Cambridge: Cambridge University Press. Bolton, Kingsley 2005 Symposium on World Englishes (Part II). Where WE stands: Approaches, issues, and debate in World Englishes. World Englishes 24 (1): 69–83. Bolton, Kingsley, and Maria Kuteeva 2012 English as an academic language at a Swedish university: Parallel language use and the ‘threat’ of English. Journal of Multilingual and Multicultural Development 33 (5): 429–447. Bourdieu, Pierre 1991 Linguistics and Symbolic Power. Cambridge, MA: Harvard University Press. Bourdieu, Pierre, and Passeron, J.-C. 1994 Introduction: Language and the relationship to language in the teaching situation. In Academic discourse, Pierre Bourdieu, J.-C. Passeron, and Monique de Saint Martin, 1–34. Cambridge: Polity Press. Brazil, David 1995 A Grammar of Speech. Oxford: Oxford University Press. Breiteneder, Angelika 2005 The naturalness of English as a European lingua franca: The case of the ‘third person -s’. Vienna English Working Papers 14: 3–26. Brenn-White, Megan, and Edwin van Rest 2012 English-Taught Master's Programs in Europe: New Findings on Supply and Demand. Institute of International Education. Brock-Utne, Birgit 2001 The growth of English for academic communication in the Nordic Countries. International Review of Education 47 (3–4): 221–233. Bruthiaux, Paul 2003 Squaring the circles: Issues in modeling English worldwide. International Journal of Applied Linguistics 13 (2): 159–178. Burt, Marina K. 1975 Error analysis in the adult EFL classroom. TESOL Quarterly 9 (1): 53–63.

238

References

Canagarajah, Suresh 2005 Changing communicative needs, shifting pedagogical priorities, revised assessment objectives. Paper presented at featured symposium on The Assessment of World Englishes, 14th World Congress of Applied Linguistics, Madison, WI. 2006 Negotiating the local Language as a lingua franca. Annual Review of Applied Linguistics 26: 197–218. Carter, Ronald, and Michael McCarthy 1995 Grammar and the spoken language. Applied Linguistics 16 (2): 141– 158. 1997 Exploring Spoken English. Cambridge: Cambridge University Press. 2006 Cambridge Grammar of English. Cambridge: Cambridge University Press. Caudery, Tim, Margrethe Petersen and Philip Shaw 2008 The motivations of exchange students at Scandinavian universities. In Students, Staff and Academic Mobility in Higher Education, Mike Byram, and Fred Dervin (eds.), 114–130. Newcastle: Cambridge Scholars Publishing. Cenoz, Jasone 2006 Across the assessment gap: Challenges for practice and research. In Bridging the Assessment Gap in English-Medium Higher Education Robert Wilkinson, and Vera Zegers (eds.), 281–291. Nijmegen: AKS-Verlag Bochum. Charles, Maggie, Diane Pecorari and Susan Hunston 2009 Academic Writing: At the Interface of Corpus and Discourse. London: Continuum. Cogo, Alessia 2008 English as a Lingua Franca: Form Follows Function. English Today 24 (3): 58–61. 2009 Accommodating difference in ELF conversations. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 254–274. Newcastle: Cambridge Scholars Publishing. 2010 Strategic use and perceptions of English as a lingua franca. Poznan Studies in Contemporary Linguistics 46 (3): 295–312. Cogo, Alessia, and Martin Dewey 2006 Efficiency in ELF communication: From pragmatic motives to lexico-grammatical innovation. Nordic Journal of English Studies 5 (2): 59–93. 2011 Analysing English as a lingua franca: A corpus-driven investigation. London: Continuum. Coleman, James A 2006 English-medium teaching in European higher education. Language Teaching 39 (1): 1–14. Corré, Alan 2005 A glossary of lingua franca. http://minds.wisconsin.edu/bitstream/

References 239 Crystal, David 1995 The Cambridge Encyclopedia of the English language. Cambridge: Cambridge University Press. Dalton-Puffer, Christiane, Gunther Kaltenboeck, and Ute Smit 1997 Learner Attitudes and L2 Pronunciation in Austria. World Englishes 16 (1): 115–128. Darquennes, Jeroen, and Peter Nelde 2006 German as a lingua franca. Annual Review of Applied Linguistics 26: 61–77. Decke-Cornill, Helene 2003 ‘We would have to invent the language we are supposed to teach’: The issue of English as a lingua franca in language education in Germany. In Context and Culture in Language Teaching and Learning, Mike Byram, and Peter Grundy (eds.), 59–71. Clevedon: Multilingual Matters. Delamare, Trish 1996 The importance of interlanguage errors with respect to stereotyping by native speakers in their judgements of second language learners’ performance. System 24 (3): 279–297. Demont-Heinrich, Christof 2005 Language and national identity in the era of globalization: The case of English in Switzerland. Journal of Communication Inquiry 29 (1): 66–84. Deterding, David, and Andy Kirkpatrick 2006 Intelligibility and an emerging ASEAN English lingua franca. World Englishes 25 (3): 391–410. Dewey, Martin 2007 English as a lingua franca and globalization: An interconnected perspective. International Journal of Applied Linguistics 17 (3): 332– 354. Draper, Stephen W., and M. I. Brown 2004 Increasing interactivity in lectures using an electronic voting system. Journal of Computer Assisted Learning 20: 81–94. Dörnyei, Zoltan 2001 Teaching and Research Motivation. London: Longman. 2007 Research Methods in Applied Linguistics. Oxford: Oxford University Press. Dürmüller, Urs 2001 The presence of English at Swiss universities. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 363–389. Berlin: Walter de Gruyter. ELFA A Corpus of English as a Lingua Franca in Academic Settings. http://www.eng.helsinki.fi/elfa/elfacorpus.htm.

240

References

Erard, Michael 2012 Advice for native speakers. The New York Times. Room for debate. http://www.nytimes.com/roomfordebate/2012/01/29/ Erling, Elizabeth J. 2002 ‘I learn English since ten years’: The global English debate and the German university classroom”. English Today 18 (2): 8–13. 2004 Globalization, English and the German university classroom: A socio-linguistic profile of students of English at the Freie Universität Berlin. Unpublished PhD thesis. University of Edinburgh. Erling, Elizabeth J., and Tom Bartlett 2006 Making English their own: The use of ELF among students of English at the FUB. Nordic Journal of English Studies 5 (2): 9–40. European commission 2009 The Bologna Process: Towards the European Higher Education Area. http://ec.europa.eu/education/higher-education/doc1290_en.htm Fahmy, Jane Jackson, and Linda Bilton 1990 Listening and Note-taking in Higher Education. Ericdb Document No. ED366189. Washington DC: Education Resources Information Center (ERIC) Fayer, John M., and Emily Krasinski Native and nonnative judgements of intelligibility and irritation. Language Learning 37 (3): 313–326. Ferguson, Gibson 2007 The global spread of English, scientific communication and ESP: Questions of equity, access and domain Loss. Ibérica 13: 7–38. Firbas, Jan 1976 A study in the functional perspective of the English and the Slavonic interrogative sentence. Brno Studies in English 12: 9–56. 1992 Functional Sentence Perspective in Written and Spoken Communication. Cambridge: Cambridge University Press. Firth, Alan 1990 ‘Lingua franca’ negotiations: Towards an interactional approach. World Englishes 9 (3): 269–280. 1996 The discursive accomplishment of normality: On ‘lingua franca’ English and conversation analysis”. Journal of Pragmatics 26 (2): 237–259. 2009a Doing not being a foreign language learner: English as a lingua franca in the workplace and (some) implications for SLA. International Review of Applied Linguistics 47: 127–156. 2009b The lingua franca factor. Intercultural Pragmatics 6 (2): 147–170. Firth, Alan, and Johannes Wagner 1997 On discourse, communication and (some) fundamental concepts in SLA research. The Modern Language Journal 81: 285– 300.

References 241 Flowerdew, John 1994 Research of relevance to second language lecture comprehension: An overview. In Academic Listening: Research Perspectives, John Flowerdew (ed.), 9–23. Cambridge: Cambridge University Press. Fowler, William C. 1859 English Grammar: The English Language in its Elements and Forms with a History of its Origin and Development. New York: Harper and Brothers. Friedrich, Patricia 2000 English in Brazil: Function and attitudes. World Englishes 19 (2): 215–223. Galloway, Vicki B. 1980 Perceptions of the communicative efforts of American students of Spanish. Modern Language Journal 64 (4): 428–433. Garfield, Eugene 1989 The English language: The lingua franca of international science. The Scientist 3 (10): 344–345. Gazzola, Michele 2012 The linguistic implications of academic performance indicators: General trends and case study. International Journal of the Sociology of Language 216: 131–156. Gerber, Ans, Johann Engelbrecht, Ansie Harding, and John Rogan 2005 The influence of second language teaching on undergraduate mathematics performance. Mathematics Education Research Journal 17 (3): 3–21. Giora, Rachel 1985 What’s a coherent text? In Text Connexity, Text Coherence: Aspects, Methods, Results, Emel Sözer, E. (ed.), 16–35. Hamburg: Buske. 1988 On the informativeness requirement. Journal of Pragmatics 12 (5/6): 547–565. [Reprinted in Cognitive Aspects of Language Use, Asa Kasher (ed.), 1989. 63–96. Amsterdam: North Holland. Reprinted in Asa Kasher, A. (ed.) 1998. Pragmatics: Critical Concepts 5. 468– 486. London: Routledge]. Graddol, David 1997 The Future of English. London: British Council. Gramkow Andresen, Karsten 1993 Lingua franca discourse: An investigation of the use of English in an international business context. Unpublished M.A. thesis. Aalborg University. Grau, Maike 2005 English as a global language: What do future teachers have to say? In The Globalization of English and the English Language Classroom, Claus Gnutzmann, and Frauke Intemann (eds.), 261–274. Tübingen: Gunter Narr.

242

References

Greenbaum, Sidney 1996 The Oxford English Grammar. Oxford: Oxford University Press. Guilloteaux, Marie- Josè 2007 Motivating language learners: A classroom-oriented investigation of teachers’ motivational practices and students’ motivation. Unpublished PhD thesis. University of Nottingham. Gunnarsson, Britt-Louise 2001 Swedish, English, French or German – the language situation at Swedish universities. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 287–316. Berlin: Mouton de Gruyter. Gunnarsson, Britt-Louise, and Katarina. Öhman 1997 Det internationaliserade universitetet. En studie av bruket av engelska och andra främmande språk vid Uppsala Universitet [The Internationalized University. A Study of the Use of English and Other Foreign Languages at Uppsala University] Uppsala: Department of Nordic Languages, Uppsala University. Gustafsson, Bengt, Göran Hermerén, and Bo Petersson 2006 Good Research Practice — What Is It? Views, Guidelines and Examples. Stockholm: Vetenskapsrådet (Swedish Research Council). Guy, Gregory Riordan 1981 Linguistic variation in Brazilian Portuguese: Aspects of the phonology, syntax, and language history. Unpublished doctoral dissertation. University of Pennsylvania. Haarman, Harald, and Eugene Holman 2001 The dominance of English as a language of science in Finland and its role for the transition to network society. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 229–261. Berlin: Walter de Gruyter. Haberland, Hartmut 2011 Ownership and maintenance of a language in transnational use: Should we leave our lingua franca alone? Journal of Pragmatics 43 (4): 937–949. Haberland, Hartmut, and Janus Mortensen 2012 Language variety, language hierarchy and language choice in the international university. International Journal of the Sociology of Language 216, 1–6. Hakala, Henrik 2007 ‘Almost as annoying as the Yank; better accent, though’: Attitudes and conceptions of Finnish students toward accents of English. Unpublished M.A. thesis. University of Helsinki. Hellekjær, Glenn Ole 2009 Language matters: Assessing lecture comprehension in Norwegian

References 243 English-medium higher education. In Language Use in Languageand-Content Integrated Learning (CLIL), Ute Smit, Tarja Nikula, and Christiane Dalton-Puffer (eds.), Amsterdam: John Benjamins. Hellekjær, Glenn Ole, and O. and Marit R. Westergaard 2003 An exploratory survey of content learning through English at Nordic universities. In Multilingual Approaches in University Education, Charles van Leeuwen and Robert Wilkinson (eds.), 65–80. Nijmegen: Valkhof Pers. Hincks, Rebecca 2010 Speaking rate and information content in English lingua franca presentations. English for Specific Purposes 29 (1): 4–18. Hirst, Daniel 1998 Intonation in British English. In Intonation Systems: A Survey of Twenty Languages, Daniel Hirst, and Albert Di Cristo (eds.), 56–77. Cambridge: Cambridge University Press. Hirst, Daniel, and Albert Di Cristo 1998 A Survey of Intonation Systems. In Intonation Systems: A Survey of Twenty Languages, Daniel Hirst, and Albert Di Cristo (eds.), 1–44. Cambridge: Cambridge University Press. House, Juliane 1996 Developing pragmatic fluency in English as a foreign language. Studies in Second Language Acquisition 18 (2): 225–252. 1999 Misunderstanding in intercultural communication: Interactions in English as a lingua franca and the myth of mutual intelligibility. In Teaching and Learning English as a Global Language, Claus Gnutzmann (ed.), 133–160. Tübingen: Stauffenburg. 2002a Communicating in English as a lingua franca. In EUROSLA Yearbook 2, Susan, Foster-Cohen, Tanja Ruthenberg, and Marie L. Poschen (eds.), 243–261. Amsterdam: John Benjamins. 2002b Developing pragmatic competence in English as a lingua franca. In Lingua Franca Communication, Karlfried Knapp, and Christiane Meierkord (eds.), 245–267. Frankfurt: Peter Lang. 2003a English as a lingua franca: A threat to multilingualism? Journal of Sociolinguistics 7 (4): 556–578. 2003b Teaching and learning pragmatic fluency in a foreign language: The case of English as a lingua franca. In Pragmatic Competence and Foreign Language Teaching, Alicia Martínez Flor, Esther Usó Juan and Ana Fernández Guerra (eds.), 133–159. Castellò de la Plana: Publicacions de la Universitat Jaume I. 2009 Introduction: The pragmatics of English as a lingua franca. Intercultural Pragmatics 6 (2): 141–145. Hülmbauer, Cornelia 2007 ‘You moved, aren’t? The relationship between lexico-grammatical correctness and communicative effectiveness in English as a lingua

244

References

2009

franca. Vienna English Working Papers 16 (2): 3–35. ‘We don’t take the right way. We just take the way that we think you will understand’– The shifting relationship of correctness and effectiveness in ELF communication. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 323– 347. Newcastle: Cambridge Scholars Publishing.

Hüttner, Julia 2009 Fluent speakers –fluent interactions: On the creation of (co)-fluency in English as a lingua franca. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 274–297. Newcastle: Cambridge Scholars Publishing. Hult, Francis 2003 English on the streets of Sweden: An ecolinguistic view of two cities and a language policy. Working Papers in Educational Linguistics 19 (1): 43–63. 2007 Multilingual language policy and English language teaching in Sweden. University of Pennsylvania. Unpublished PhD thesis. Hultfors, Per 1986 Reactions to Non-Native English: Native English-Speakers’ Assessments of Errors in the Use of English Made by Non-Native Users of the Language. 2 vols. Stockholm Studies in English 67 & 71. Stockholm: Almqvist & Wiksell International. Hyltenstam, Kenneth 1999 Svenskan i minoritetsperspektiv [Swedish from a minorities point of view]. In Sveriges sju inhemska språk — ett minoritetsperspektiv, [Sweden’s seven local languages– a minorities point of view] Kenneth Hyltenstam (ed.), 205–240. Lund: Studentlitteratur. Hynninen, Niina 2011 The practice of ‘mediation’ in English as a lingua franca interaction. Journal of Pragmatics 43 (4): 965–977 Inoue, Fumio 2001 English as a language of science in Japan: From corpus planning to status planning. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 446–471. Berlin: Walter de Gruyter. James, Allan R. 2000 English as a European lingua franca: Current realities and existing dichotomies. In English in Europe: The Acquisition of a Third Language, Jasone Cenoz, and Ulrike Jessner (eds.), 22–38. Clevedon: Multilingual Matters. Jenkins, Jennifer 2000 The Phonology of English as an International Language: New Models, New Norms, New Goals. Oxford: Oxford University Press. 2002 A sociolinguistically based, empirically researched pronunciation

References 245 syllabus for English as an international language. Applied Linguistics 23 (1): 83–103. 2003 World Englishes: A Resource Book for Students. London: Routledge. 2006a Current perspectives on teaching World Englishes and English as a lingua franca. TESOL Quarterly 40 (1): 157–181. 2006b Points of view and blind spots: ELF and SLA. International Journal of Applied Linguistics 16 (2): 137–162. 2007 English as a Lingua Franca: Attitude and Identity. Oxford: Oxford University Press. 2009 English as a Lingua Franca: Interpretations and attitudes. World Englishes 28 (2): 200–207. 2011 Accommodating (to) ELF in the International University. Journal of Pragmatics. Special issue: The Pragmatics of English as a Lingua Franca 43 (4): 926–936. Jenkins, Jennifer, Alessia Cogo, and Martin Dewey 2011 Recent developments in research into English as a lingua franca. Language Teaching 44 (3): 281–315. Jenkins, Jennifer, Marco Modiano and Barbara Seidlhofer 2002 Euro-English”. English Today 17 (4): 13–19. Johansson, Stig 1975 A Methodological Study of the Communicative Effect of Learners’ Errors. In Papers in Contrastive Linguistics and Language Testing, Stig Johansson (ed.), Lund: Gleerup. 1978 Studies in Error Gravity: Native Reactions to Errors Produced by Swedish Learners of English. Gothenburg Studies in English 44. Gothenburg: University of Gothenburg. Johnson, Richard 1969 (1706) Grammatical Commentaries. Menston: Scolar Press. Josephson, Olle 2004 Engelskan i 2000-talets Sverige [English in Sweden in the twentyfirst century]. In Engelskan i sverige. Språkval i utbildning, arbete och kulturliv [English in Sweden: Language choice in education, at workplace and culture]. Svenska språknämnden, 7–24. Kachru, Braj 1982 Models for non-native Englishes. In The Other Tongue, Braj Kachru (ed.), 31–57. Urbana, IL: University of Illinois Press. 1985 Standards, codification and sociolinguistic realism: The English language in the outer circle. In English in the World: Teaching and Learning the Language and Literatures, Randolph Quirk, and Henry G. Widdowson (eds.), 11–30. Cambridge: Cambridge University Press. Kaplan, Robert B. 2001 English — the accidental language of science? In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 3–27. Berlin: Walter De Gruyter.

246

References

Karlgren, Jussi, and Preben Hansen 2003 Cross-language relevance assessment and task context. Paper presented at the Advances in Cross-Language Information Retrieval. Third Workshop of the Cross-Language Evaluation Forum, CLEF, Rome, Italy. Kaur, Jagdish 2009 Pre-empting problems of understanding in English as a lingua franca. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 107–124. Newcastle: Cambridge Scholars Publishing. 2011 'Doing being a language expert': The case of the ELF speaker. In Latest Trends in ELF Research, Alasdair Archibald, Alessia Cogo, and Jennifer Jenkins (eds.), Cambridge Scholars Publishing, 5375. Kawai, Yuko 2007 Japanese nationalism and the global spread of English: An analysis of Japanese governmental and public discourses on English. Language and Intercultural Communication 7 (1): 37–55. Kift, Sally 2006 Using an Audience Response System to Enhance Student Engagement in Large Group Orientation: A Law Faculty Case Study. In David A. Banks (ed.), Audience Response Systems in Higher Education: Applications and Cases, 80–96. Hershey, PA: Information Science Publishing. Kiparsky, Paul 1972 Explanation in Phonology. In Goals of Linguistic Theory, Stanley Peters (ed.), 189–227. Englewood Cliffs, NJ: Prentice-Hall. Kirkpatrick, Andy 2008 English as the official language of the Association of Southeast Asian Nations (ASEAN): Features and strategies. English Today 24 (2): 27–34. 2010 Researching English as a lingua franca in Asia: The Asian Corpus of English (ACE) project. Asian Englishes 31 (1): 4–18. 2012 English as an Asian Lingua Franca: The ‘Lingua Franca Approach’ and implications for language education policy. Journal of English as a Lingua Franca 1 (1): 121–139. Kivistö, Anne 2005 Accents of English as a lingua franca: A study of Finnish textbooks. Unpublished MA thesis. University of Tampere. Klaassen, Renate 2001 The international university curriculum: Challenges in Englishmedium engineering education. Unpublished PhD thesis. Delft University of Technology. Klimpfinger, Theresa 2009 ’She’s mixing the two languages together’– forms and functions of

References 247 code-switching in English as a lingua franca. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 348–371. Newcastle: Cambridge Scholars Publishing. Knapp, Annelie 2011 Using English as a lingua franca for (mis-)managing conflict in an international university context: An example from a course in engineering. Journal of Pragmatics 43 (4): 978–990. Kruseman, Nieuwenhuijzen Arie 2003 Conceiving a bilingual university: Challenges and issues. In Multilingual Approaches in University Education: Challenges and Practices, Charles van Leeuwen, and Robert Wilkinson (eds.), 7–11. Nijmegen: Valkhof Pers. Kryuchkova, Tatjana B. 2001 Russian as intermediary language in Europe. In Sociolinguistica. Verkehrssprachen in Europa – außer English. Lingua Francas in Europe – except English. Langues véhiculaires en Europe – sans l’anglais, Ulrich Ammon, U., Klaus J. Mattheier and Peter H. Nelde (eds.), 96–119. Tübingen: Max Niemeyer Verlag. Kuo, I-Chun (Vicky) 2006 Addressing the issue of teaching English as a lingua franca. ELT Journal 60 (3): 213–221. Labov, William 1966 The Social Stratification of English in New York City. Washington DC: Center for Applied Linguistics. Labrie, Normand, and Carsten Quell 1997 Your language, my language or English? The potential language choice in communication among nationals of the European Union. World Englishes 16 (1): 3–26. Lappalainen, Sari 2001 English as lingua franca. Towards fluent communication. Unpublished MA thesis. University of Turku. Leech, Geoffrey 2000 Grammars of spoken English: New outcomes of corpus-oriented research. Language Learning 50 (4): 75–724. Lesznyák, Ágnes 2002 From chaos to the smallest common denominator. Topic management in English lingua franca communication. In Lingua Franca Communication, Karlfried Knapp, and Chistiane Meierkord (eds.), 163–193. Frankfurt: Peter Lang. Levis, John M. 1999 The intonation and meaning of normal yes/no questions. World Englishes 18 (3): 373–380. Lichtkoppler, Julia 2007 ‘Male. Male.‘ – ’Male?‘ – ’The Sex is Male.‘ The Role of Repetition

248

References in English as a Lingua Franca Conversations”. Vienna English Working Papers 16 (1): 39–65.

Lindberg, Bo 1984 De lärdes modersmål. Latin, humanism och vetenskap i 1700- talets Sverige [The language of the educated: Latin, humanism and science in the eighteenth century Sweden]. Gothenburg Studies in the History of Science and Ideas 5. Gothenburg: Acta Universitatis Gothoburgensis. Lindsey, Geoffrey A. 1985 Intonation and interrogation: Tonal structure and the expression of a pragmatic function in English and other languages. Unpublished PhD thesis. University of California. Littlewood, William T. 1981 Language variation and second language acquisition theory. Applied Linguistics 11 (2): 150–158. Ljosland, Ragnhild 2007 English in Norwegian academia: A step towards diglossia? World Englishes 26 (4): 395–410. 2008 Lingua franca, prestisjespråk og forestilt fellesskap: om Engelsk som akademisk språk i Norge. Et kasusstudium i bred kontekst. [Lingua Franca, Prestige and Imagined Communities: On English as an Academic Language in Norway. A Case Study in its Broader Context]. Unpublished PhD thesis. NTNU (Norwegian University of Science and Technology). 2011 English as an academic lingua franca: Language policies and multilingual practices in a Norwegian university. Journal of Pragmatics 43 (4): 991–1004. Llurda, Enric 2004 Non-native-speaker teachers and English as an international language. International Journal of Applied Linguistics 14 (3): 314–323. 2009 Attitudes towards English as an international language: The pervasiveness of native models among L2 users and teachers. In English as an international language: Perspectives and pedagogical issues, Farzad Sharifian (ed.), 119–135. Multilingual matters. Bristol: UK. Ludwig, Jeanette 1982 Native-speaker judgments of second-language learners’ efforts at communication: A review. The Modern Language Journal 66 (3): 274–283. Matsuda, Aya 2003 The ownership of English in Japanese secondary schools. World Englishes 22 (4): 483–496. Mauranen, Anna 2003 The corpus of English as lingua franca in academic settings. TESOL Quarterly 37 (3): 513–527.

References 249 2005

English as a lingua franca: An unknown language? In Identity, Community, Discourse: English in Intercultural Settings, Giuseppina Cortese, and Anna Duszak (eds.), 269–292. Frankfurt: Peter Lang. 2006a A rich domain of ELF – the ELFA corpus of academic discourse. Nordic Journal of English Studies 5 (2): 145–159. 2006b Signaling and preventing misunderstanding in English as lingua franca communication. International Journal of the Sociology of Language 177: 123– 150. 2006c Spoken discourse, academics and global English: A corpus perspective. In Spoken English, TESOL and applied linguistics: Challenges for theory and practice, Rebecca Hughes (ed.), 143–158. London: Palgrave Macmillan. 2007 Hybrid voices: English as the lingua franca of academics. In Language and Discipline Perspectives on Academic Discourse, Kjersti Flottum (ed.), 243– 259. Newcastle: Cambridge Scholars Publishing. 2009 Chunking in ELF: Expressions for managing interaction. Journal of Intercultural Pragmatics 6 (2): 217–233. 2012 Exploring ELF. Academic English shaped by non-native spekers. Cambridge: Cambridge University Press. Maybaum, Rebecca, and John M. Swales 2006 Pre-and post-dislocations in MICASE”. MICASE Kibbitzer 11. http://lw.lsa.umich.edu/eli/micase/Kibbitzer/Dislocation.htm McKay, Sandre Lee 2002 Teaching English as an International Language: Rethinking Goals and Approaches. Oxford: Oxford University Press. Medgyes, Péter, and Mónika László 2001 The foreign language competence of Hungarian scholars: Ten years later. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 261–286. Berlin: Mouton de Gruyter. Meierkord, Christiane 1996 Englisch als Medium der interkulturellen Kommunikation. Untersuchungen zum non-native-/non-native speaker-Diskurs. Frankfurt: Peter Lang. 2000 Interpreting successful lingua franca interaction. An analysis of nonnative-/non-native small talk conversations in English. In Neuere Entwicklungen in Der Gesprächsforschung. Sonderausgabe Von Linguistic Online 5 (1). Karin Pittner, and Annika Fetzer (eds.) 2002 ‘Language stripped bare’ or ‘Linguistic masala’? Culture in lingua franca conversation”. In Lingua Franca Communication, Karfried Knapp, and Christiane Meierkord (eds.), 109–133. Frankfurt: Peter Lang. 2004 Syntactic variation in interactions across International Englishes. English World-Wide 25 (1): 109–132.

250

References

2005

Black South African Englishes – towards a variationist account. Erfurt Electronic Studies in English, (EESE). http://webdoc.sub.gwdg.de/edoc/ia/eese/artic25/meierk/12005.html 2006 Lingua franca communication past and present. International Journal of the Sociology of Language 177: 9–30. Melchers, Gunnel, and Philip Shaw 2003 World Englishes. London: Arnold. Menon, Anil S., Shannon Moffett, Melissa Enriquez, Miriam M. Martinez, Dev Parvati, and Todd Grappone 2004 Audience Response Made Easy: Using Personal Digital Assistants as a Classroom Polling Tool. Journal of the American Medical Informatics Association 11 (3): 217–220. Mesthrie, Rajend, and M. Rakesh Bhatt 2008 World Englishes: The Study of New Linguistic Varieties. Cambridge: Cambridge University Press. Mesthrie, Rajend, Joan Swann, Ana Deumert, and William L. Leap 2001 Introducing Sociolinguistics. Edinburgh: Edinburgh University Press. Metsä-Ketelä, Maria 2006 ’Words are more or less superfluous’: The case of more or less in academic lingua franca English”. Nordic Journal of English Studies 5 (2): 117–143. Mitchell, Linda C. 2006 Grammar wars: Seventeenth and eighteenth century England. In The Handbook of World Englishes, Kachru, Braj B., Yamuna Kachru and Cecil L. Nelson (eds.), 475–496. Oxford: Blackwell. Mollin, Sandra 2006 Euro-English: Assessing Variety Status. Tübingen: Gunter Narr. Morell, Tessa 2004 Interactive lecture discourse for university EFL Students. English for Specific Purposes 23 (3): 325–338. Mufwene, Salikoko 2006 Language endangerment: An embarressment for linguistics. In Proceedings from the parasessions of the forty-second meeting of the Chicago Linguistic Society 42 (2), Jacquilne Bunting, Sapna Desai, Robert Peachey, Christopher Straughn, Zuzana Tomkovà 42 (2): 111–140. Mulligan, Denise, and Andy Kirkpatrick 2000 How much do they understand? Lectures, students and comprehension. Higher Education Research and Development 19 (3): 311–335. Munro, Murray J., and Tracey M. Derwing 1999 Foreign accent, comprehensibility, and intelligibility in the speech of second language learners. Language Learning 49 (1): 285–310.

References 251 Murray, Heather, and Silvia Dingwall 2001 The Dominance of English at European Univer- sities: Switzerland and Sweden Compared. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 85–112. Berlin: Walter de Gruyter. Murphy, Dermot F., and Christopher Candlin 1979 Engineering listening discourse and listening comprehension. Practical Papers in Language Education 2: 1–79. Mühleisen, Susanne 2003 Towards global diglossia? In The Politics of English as a World Language: New Horizons in Postcolonial Cultural Studies, Christian Mair (ed.), 107–118. Amsterdam: Rodopi. Nastansky, Heinz-L. 2004 National strategy in the internationalisation of higher education: The German perspective. In Integrating Content and Language: Meeting the Challenge of a Multilingual Higher Education, Robert Wilkinson (ed.), 49–54. Maastricht: Universitaire Pers. Neville-Barton, Pip Neville, and Bill Barton 2005 The Relationship between English Language and Mathematics Learning for Non-native Speakers. Wellington, New Zealand: Teaching and Learning Research Initiative. http://www.tlri.org.nz/pdfs/13909.pdf Nixon, Graham 1972. Measuring corpus adequacy. Archivum Linguisticum nsIII.101– 104. Noorman, Ilona 2008 ENL, ESL, EFL, EIL or ELF: On the Idea of Teaching the Lingua Franca Core at Dutch Secondary Schools. Unpublished MA thesis. Utrecht University. O’Grady, Gerard 2010 A Grammar of Spoken English Discourse: The Intonation of Increments. Continuum Studies in Theoretical Linguistics. Olsson, Margareta 1977 Intelligibility. An evaluation of some features of English produced by Swedish 14-year-olds. Gothenburg Studies in English 40. Gothenburg: Acta Universitatis Gothoburgensis. Osgood , Charles E., George J. Suci, and Percy Tannenbaum 1957 The Measurement of Meaning. Urbana, IL: University of Illinois Press. Osimk, Ruth 2009 Decoding sounds: An experimental approach to intelligibility in ELF. Vienna English Working Papers 18, 64–89. Park, Joseph Sung-Yul. and Lionel Wee 2009 The Three Circles Redux: A Market-theoretic Perspective on World Englishes. Applied Linguistics 30 (3): 389–406.

252

References

Parkvall, Mikael 2005 Foreword. In A glossary of lingua franca, Alan Corré (ed.), http://minds.wisconsin.edu/bitstream/item/3920/edition3/foreword.ht ml Penz, Hermine 2008 “What do we mean by that?” – ELF in Intercultural Project Work. Paper presented at the ESSE conference. University of Aarhus, Denmark, August 22–26. Petersen, Margrethe, and Philip Shaw 2002 Language and disciplinary differences in a biliterate context. World Englishes 21 (3): 357–374. Pérez Aguilar M., and Macia Arnó E. 2002 Metadiscourse in lecture comprehension: Does it really help foreign language learners? Atlantis 24 (1): 3–21. Phillipson, Robert 1998 Globalizing English: Are linguistic human rights an alternative to linguistic imperialism? Language Sciences 20 (1): 111–112. 1999 Voice in global English: Unheard chords in Crystal, loud and clear. Applied Linguistics 20 (2): 265–276. 2006a English, a cuckoo in the European higher education nest of languages? European Journal of English Studies 10 (1): 13–32. 2006b Figuring out the Englishisation of Europe. In Reconfiguring Europe: The Contribution of Applied Linguistics, Constant Leung, and Jennifer Jenkins (eds.): 65–86. London: Equinox. 2008 Lingua franca or lingua frankensteinia? English in European integration and globalization. World Englishes 27 (2): 250–284. Pickering, Lucy 2006 Current research on intelligibility in English as a lingua franca. Annual Review of Applied Linguistics 26: 219–233. 2009 Intonation as a pragmatic resource in ELF interaction. Intercultural Pragmatics 6 (2): 235–255. Pickering, Lucy, and Jason Litzenberg 2011 Intonation as a pragmatic resource in ELF interaction, revisited. In Latest Trends in ELF research, Alasdair Archibald, Alessia Cogo, and Jennifer Jenkins (eds.), Newcastle Upon Tyne: Cambridge Scholars Press. Pitzl, Marie- Luise 2005 Non-understanding in English as a lingua franca: Examples from a business context. Vienna English Working Papers 14 (2): 50–71. 2009 ’We should not wake up any dogs’: Idiom and metaphor in ELF. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 298–322. Newcastle: Cambridge Scholars Publishing.

References 253 Planken, Brigitte 2005 Managing rapport in lingua franca sales negotiations: A comparison of professional and aspiring negotiators. English for Specific Purposes 24 (4): 381–400. Preisler, Bent Introduction. In Language and Learning in the International Univer2011 sity: From English Uniformity to Diversity and Hybridity, Bent Preisler, Ida Klitgård, and Anne Fabricius (eds.), viii–3. Bristol: Multilingual Matters. Prodromou, Luke 1996 Correspondence from Luke Prodromou. ELT Journal 50 (4): 371– 373. 2003 Idiomaticity and the Native Speaker. English Today 19: 42–48. 2007a Bumping into creative idiomaticity. English Today 23 (1): 14–25. 2007b Is ELF a variety of English? English Today 23 (2): 47–53. 2008 English as a Lingua Franca: A Corpus-based Approach. London: Continuum. Pullin-Stark, Patricia 2009 No joke–This is serious! Power, solidarity and humour in business English as a lingua franca (BELF). In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 152– 177. Newcastle: Cambridge Scholars Publishing. Pölzl, Ulrike, and Barbara Seidlhofer 2006 In and on their own terms: The ‘habitat factor’ in English as a lingua franca interactions. International Journal of the Sociology of Language 177: 151–176. Quirk, Radolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik 1985 A Comprehensive Grammar of the English Language. London: Longman. Rajadurai, Joanne 2003 Revisiting the concentric circles: Conceptual and sociolinguistic considerations. Asian EFL Journal 7 (4): 111–130. 2007 Intelligibility studies: A consideration of empirical and ideological issues. World Englishes 26 (1): 87–98. Ranta, Elina 2006 The ‘attractive’ progressive — Why use the -ing form in English as a lingua franca? Nordic Journal of English Studies 5 (2): 95–116. 2009 Syntactic features in spoken ELF — Learner language or spoken grammar? In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 84–107. Newcastle: Cambridge Scholars Publishing. Reichardt, Charles S., and Thomas D. Cook 1979 Beyond qualitative versus quantitative methods. In Qualitative and Quantitative Methods in Evaluation Research Cook, Thomas D.

254

References

Cook, and Charles S. Reichardt (eds.), 7–32. Newbury Park, CA: Sage. Risager, Karin 2012 Language hierarchies at the international university. International Journal of the Sociology of Language 216: 111–130. Rodrik, Dani 1997 Has Globalization Gone Too Far? Washington D.C.: Institute for International Economics. 1999 The New Global Economy and the Developing Countries: Making Openness Work. Washington D.C.: Overseas Development Council. Rounds, Patricia 1987 Characterizing successful classroom discourse for NNS teaching assistant training. TESOL Quarterly 21 (4): 643–671 Salminen, Louhiela Leena, Mirjaliisa Charles, and Anna Kankaanranta 2005 English as a lingua franca in Nordic corporate mergers: Two case companies. English for Specific Purposes 24 (4): 401–421. Salö, Linus 2010 Engelska eller svenska? En kartläggning av språksituationen inom högre utbildning och forskning [English or Swedish? A survey of the language situation in higher education and research]. Stockholm: Språkrådet. Samovar, Larry A., and Jack Mills 1995 Oral Communication: Speaking across Cultures. Boston, MA: McGraw Hill. Sayer, Andrew 1992 Method in Social Science. London: Routledge. Seidlhofer, Barbara 1999 Double standards. Teacher education in the expanding circle. World Englishes 18 (2): 233–245 2001 Closing a conceptual gap: The case for a description of English as a lingua franca. International Journal of Applied Linguistics 11 (2): 133–158. 2002 The shape of things to come? Some basic questions about English as a lingua franca. In Lingua Franca Communication, Karlfried Knapp, and Christiane Meierkord (eds.), 269–302. Frankfurt: Peter Lang. 2003 A Concept of International English and Related Issues: From ‘Real English’ to ‘Realistic English’. Strasbourg: Council of Europe. http://www.esev.ipv.pt/servicos/upload%5Cma%5C292%5CSeidlhof erEN.pdf 2004 Research perspectives on teaching English as a lingua franca. Annual Review of Applied Linguistics 24 (1): 209–239. 2005 English as a lingua franca. ELT Journal 59 (4): 339–341. 2006 English as a lingua franca in the expanding circle: What it isn’t. In English in the World: Global Rules, Global Roles, Rani Rubdy, and

References 255 Mario Saraceni (eds.), 40–50. London: Continuum. Common ground and different realities: World Englishes and English as a lingua franca. World Englishes 28 (2): 236–245. 2009b Accommodation and the idiom principle in English as a Lingua Franca. Intercultural Pragmatics 6 (2): 195–215. 2011 Understanding English as a Lingua Franca. Oxford: Oxford University Press. Seidlhofer, Barbara, Angelika Breiteneder, and Marie-Luise Pitzl 2006 English as a lingua franca in Europe: Challenges for applied linguistics. Annual Review of Applied Linguistics 26: 3–34. Seidlhofer, Barbara, and Jennifer Jenkins 2003 English as a lingua franca and the politics of property. In Christian Mair (ed.), The politics of English as a world language: New horizons in postcolonial cultural studies. NY Rodopi. 139–157. Selinker, Larry 1972 Interlanguage. International Review of Applied Linguistics in Language Teaching 10 (3): 209–231. 1992 Rediscovering Interlanguage. London: Longman. Shaw, Philip 2005 The languages of international publication in economics in 1900. In Text I Arbete/Text at Work, Ulla Melander-Marttala, Hary Näslund, Ingegerd Bäcklund, and Ulla Brestam (eds.), 340–348. Institutionen för Nordiska Språk, Uppsala. 2008 Engelska som lingua franca och som internationellt vetenskapsspråk [English as a lingua franca and English as the language of international science]. In Jansson, E. (ed.), Vetenskapsengelska – med svensk kvalitet? [Scientific English–with Swedish quality?], 21–34. Stockholm: Språkrådet. Shaw, Philip, Tim Caudery, and Margrethe Petersen 2009 Students on exchange in Scandinavia: Motivation, interaction, ELF development. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 178–200. Newcastle: Cambridge Scholars Publishing. Shim, Rosa Jinyoung 2002 Changing attitudes towards teaching English as a world language in Korea. Journal of Asian Pacific Communication 12 (1): 143–158. Sifakis, Nicos 2007 The education of reachers of English as a lingua franca: A transformative perspective. International Journal of Applied Linguistics 17 (3): 355–375. 2009 Challenges in teaching ELF in the periphery: The Greek context. ELT Journal 63 (3): 230–237. Sifakis, Nicos C., and Areti-Maria Sugari 2005 Pronunciation issues and EIL pedagogy in the periphery: A survey of 2009a

256

References

Greek state school teachers’ beliefs”. TESOL Quarterly 39 (3): 467– 488. Simonsen, Dag F. 2005 Over the Fence — and into English? Reflections on adolescents, academics, linguistic development and language policy in Norway in the early 2000s. In The Consequences of Mobility, Ben Preisler, Anne Fabricius, Hartmut Haberland, Susanne Kjaerbeck, and Karin Risager (eds.), 249–271. Roskilde: Roskilde University, Department. of Language and Culture. Simpson, Rita C., Sarah L. Briggs, Janine Ovens, and John M. Swales 2002 The Michigan Corpus of Academic Spoken English. Ann Arbor, MI: The Regents of the University of Michigan. Skutnabb-Kangas, Tove, and Robert Phillipson 2000 The world came to Sweden — but did language rights? Current Issues in Language and Society 7 (1): 70–86. Smit, Ute 2009 Emic evaluations and interactive processes in a classroom community. In English as a Lingua Franca: Studies and Findings, Anna Mauranen, and Elina Ranta (eds.), 200–225. Newcastle: Cambridge Scholars Publishing. 2010 English as a lingua franca in higher education. A longitudinal study of classroom discourse. Mouton de Gruyter Series: Trends in Applied Linguistics. Smith, Larry E. 1982 Readings in English as an International Language. Pergamon: London. Smith, Larry E., and John A. Bisazza 1982 The comprehensibility of three varieties of English for college students in seven countries. Language Learning 32 (2): 259–269. Smith, Larry. E., and Cecil L. Nelson 1985a International intelligibility of English: Directions and resources. World Englishes 4 (3): 333–342. 1985b World Englishes and issues of intelligibility. In The Handbook of World Englishes, Braj B. Kachru, and Cecil L. Nelson (eds.), 428– 446. Oxford: Blackwell. Smitterberg, Erik 2005 The Progressive in 19th-Century English. A Process of Integration. Amsterdam: Rodopi. 2007 Spot the Error: A Problem-based Workbook on English Grammar and Usage. Lund: Studentlitteratur. Sobkowiak, Wlodzimierz 2005 Why not LFC?” In English Pronunciation Models: A Changing Scene, Katarzyna Dziubalska-Ko!acyzyk, and Joanna Przedlacka (eds.), 131-151. Frankfurt: Peter Lang.

References 257 Spada, Nina, and Maria Fröhlich 1995 COLT Communicative Orientation of Language Teaching Observation Scheme: Coding Conventions and Applications. Sydney: Macquarie University, National Centre for English Language Teaching and Research (NCELTR). Spolsky, Bernard 2004 Language Policy. Cambridge, UK: Cambridge University Press. 2009 Language Management. Cambridge, UK: Cambridge University Press. Stiglitz, Joseph E. 2002 Globalization and Its Discontent. New York, NY: Norton Publisher. Suvuniitty, Jaana 2010 Lecturers’ questions and student perception of lecture comprehension“ Helsinki English Studies 6. Swales, John M. 1990 Genre Analysis: English in Academic and Research Settings. Cambridge: Cambridge University Press. 1997 English as tyrannosaurus rex. World Englishes 16 (3): 373–382. 2002 Integrated and fragmented worlds: EAP materials and corpus linguistics. In Academic Discourse, John Flowerdew (ed.), 150–164. London: Longman. Swan, Michael 1980 Practical English Usage. Oxford: Oxford University Press. Swedish Agency for Networks and Cooperation in Higher Education. http://www.nshu.se/english.html Swedish Deparment of Education. God sed i forskningen. 1999. (SOU 1999:4) Stockholm: Utbildningsdepartementet. http://www.regeringen.se/sb/d/108/a/23647 Swedish National Agency for Higher Education www.hsv.se Szpyra-Kozlowska, Jolanta 2003 The lingua franca core and the Polish learner. Neofilologia V: 193– 210. Söderlundh, Hedda 2004 Lika bra på engelska? En undersökning av hur studenter i Sverige förstår kurslitteratur på svenska respektive engelska [As good in English? An investigation of students’ comprehension of coursebook literature in Swedish vs English]. Språk & Stil 14: 137–165. 2005 Svenska är lättare att fatta [Swedish is easier to understand]. Språkvård 2: 29–33. 2010 Internationella universitet – lokala språkval. Om bruket av talad svenska i engelskspråkiga kursmiljöer. [International universities – local language choices. On spoken Swedish in English-medium course environments.] Skrifter utgivna av Institutionen för nordiska språk vid Uppsala universitet 83. Uppsala.

258

References

2012

Global policies and local norms: Sociolinguistic awareness and language choice at an international university. International Journal of the Sociology of Language 216: 87–109. Tella, Seppo, Anne Räsänen, and Anne Vähäpassi (eds.) 1999 Teaching Through Foreign Language: From Tool to Empowering Mediator. Helsinki: Publications of Higher Education Evaluation Council 5. Timmis, Ivor 2002 Native speaker norms and international English. ELT Journal 56 (3): 240–249. Tislevoll, Jan R. 2001 Norsk språkpolitikk og domenetap for norsk språk. Rapport til Nordisk ministerråd. Oslo. Truchot, Claude 2001 The Languages of Science in France: Public Debate and Language Policies. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 319–328. Berlin: Walter de Gruyter. 2002 Key Aspects of the Use of English in Europe”. Language Policy Divi- sion, Council of Europe, Strasbourg. Online at http://www.Coe.int/T/E/Cultural_Cooperation/education/Languages/ Language_Policy/Policy_development_activities/Studies/List.Asp.). Uhari, Matti, Marjo Renko, and Hannu Soini 2003 Experiences of using an interactive Audience Response System in lectures. BMC Medical Education 3 (12): 1–6. Ultan, Russell 1978 Some general characteristics of interrogative systems. In Universals of Human Language, 4, Syntax, Joseph H. Greenberg (ed.), 211–249. Stanford: Stanford University Press. van Heuven, Vincent J., and Judith Haan 2000 Phonetic correlates of statement versus question intonation in Dutch. In Intonation: Analysis, Modelling and Technology, Antonis Botinis (ed.), 119–144. Dordrecht: Kluwer. van Heuven, Vincent J., and Klaske van Leyden 2003 A contrastive acoustical investigation of Orkney and Shetland intonation. In Proceedings of the 15th International Congress of Phonetic Sciences, Sole, Maria-Josep Sole, Daniel Recasens, and Ronaldo J. Romero (eds.), 805–808. Barcelona. van Heuven, Vincent J., and Ellen van Zanten 2005 Speech rate as a secondary prosodic characteristic of polarity questions in three languages. Speech Communication 47: 87–99. van Leeuwen, Charles, and Robert Wilkinson 2003 Introduction to ‘Multilingual approaches in university education: Challenges and practices’. In Multilingual Approaches in University

References 259 Education: Challenges and Practices, Charles van Leeuwen, and Robert Wilkinson (eds.), 11–17. Nijmegen: Valkhof Pers. van Leeuwen, Charles 2003 Feasibility of policy in university language teaching today. In Multilingual Approaches in University Education: Challenges and Practices, Charles van Leeuwen, and Robert Wilkinson (eds.), 19– 45. Nijmegen: Valkhof Pers. Van Parijs, Philippe 2011 Linguistic justice for Europé and for the World. Oxford: Oxford University Press. Vinke, A. A. Diana 1995 English as the medium of instruction in Dutch engineering education. Unpublished PhD thesis. Delft University of Technology. VOICE. Vienna-Oxford International Corpus of English. www.univie.ac.at/voice. Walker, Robin 2001a Pronunciation priorities, the lingua franca core, and monolingual groups. Speak Out! Newsletter of the IATEFL Pronunciation Special Interest Group 28: 4–9. 2001b Pronunciation for International Intelligibility”. English Teaching Professional 21.10–13. 2005 Using student-produced recordings with monolingual groups to provide effective, individualized pronunciation practice. TESOL Quarterly 39: 550–557. 2010 Teaching the Pronunciation of English as a Lingua Franca. Oxford: Oxford University Press. Watterson, Matthew 2008 Repair of non-understanding in English in international communication. World Englishes 27 (3/4): 378–406. Wee, Lionel 2002 When English is Not a Mother Tongue: Linguistic Ownership and the Eurasian Community in Singapore. Journal of Multilingual and Multicultural Development 23 (4): 282–295. Widdowson, Henry G. 1994 The ownership of English. TESOL Quarterly 28 (2): 377–389. 1997 EIL, ESL, EFL: Global issues and local interests. World Englishes 16 (1): 135–146. 2003 Defining Issues in English Language Teaching. Oxford: Oxford University Press. Willemyns, Roland 2001 English in linguistic research in Belgium. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 329–342. Berlin: Walter de Gruyter.

260

References

Williams, Jessica 1990 Another look at Yes/No questions: Native speakers and non-native speakers. Applied Linguistics 11 (2): 159–182. Wilson, David 2002 The Englishisation of academe: A Finnish perspective. Jyväskylä: University of Jyväskylä Language Centre. Wong, Jean, and David Olsher 2000 Reflections on conversation analysis and nonnative speaker talk: An interview with Emanuel A. Schegloff. Issues in Applied Linguistics 11 (1): 111–128. WrELFA Written Academic ELF corpus project. http://www.helsinki.fi/englanti/elfa/wrelfa.html Wright, Sue 2006 French as a lingua franca. Annual Review of Applied Linguistics 26: 35–60. Wu, William Y., Dennis W. K. Chan and Björn H. Jernudd 2001 English in science communication in Hong Kong: Educational research output. In The Dominance of English as a Language of Science: Effects on Other Languages and Language Communities, Ulrich Ammon (ed.), 193–205. Berlin: Walter de Gruyter. Wächter, Bern, and Friedhelm Maiworm 2008 English-taught Programmes in European Higher Education. Bonn: Lemmens. Young, Richard F. 1991 Variation in Interlanguage Morphology. Frankfurt: Peter Lang. Zacharias, Nugrahenny T. 2005 Teachers’ beliefs about internationally-published materials: A survey of tertiary English teachers in Indonesia. RELC Journal 36 (1): 23– 37.

Index admission, 29, 147, 162, 187, 188 analytic comparative, 92, 94, 95, 140, 141, 148, 149, 159 article usage articles, 82, 84, 94, 95, 149 Asian Corpus of English (ACE), 49, 246 assertion, 151 attitudes, 15, 37, 45, 46, 51, 56, 57, 64, 79, 138, 139, 147, 158, 159, 166, 177, 179, 197, 198, 199, 235, 241, 245, 255 Bologna Declaration, 14, 19 CANCODE, 1, 174, 175 CLIL, 1, 15, 243 Common European Framework, 162 commonalities, 32, 33, 49, 50, 51, 62, 92, 138, 140, 141, 142, 143, 145, 149, 164, 168, 169, 170 comparative and superlatives, 140, 141, 149, 159 contact languages, 28, 163 countability, 93, 94, 95, 96, 97, 140, 150 creoles, 28, 50, 147, 148, 149, 150, 168, 172 description, 4, 13, 30, 42, 43, 53, 55, 59, 63, 65, 79, 160, 174, 189, 191, 197, 198, 254 diglossia, 13, 18, 19, 20, 22, 23, 24, 248, 251 disfluency, 123, 131, 195 disturbance, 63, 73, 77, 78, 98, 101, 110, 111, 118, 119, 144, 153, 155, 156, 159, 177, 192, 193, 195, 199, 203 domain domains, 2, 12, 22, 23, 24, 25, 59, 209, 228, 229, 230, 240, 249 domain loss, 24

effectiveness, viii, 28, 29, 30, 32, 34, 35, 38, 41, 42, 50, 51, 53, 65, 98, 123, 143, 155, 156, 158, 162, 163, 173, 177, 178, 179, 183, 184, 185, 188, 192, 193, 194, 195, 196, 200, 204, 207, 236, 243, 244 ELFA, 1, 38, 39, 43, 48, 49, 51, 72, 73, 132, 175, 210, 239, 249 engineering, ix, 6, 17, 26, 37, 51, 61, 70, 80, 82, 88, 169, 183, 197, 209, 236, 246, 247, 259 ERASMUS, 1, 50 ethics ethical, 71 exchange, 14, 37, 61, 69, 78, 128, 136, 164, 172, 186, 187, 238, 255 explicitness, 39, 50, 53, 105, 123, 132, 143, 145, 149, 150, 159, 168, 173, 183, 187, 192, 194 extensive analyses, 66, 77, 93, 197, 198 extra-linguistic elements, 111, 114 fluency, 35, 36, 157, 178, 204, 243, 244 functional sentence perspective, 150, 152 General American (GA), 45 globalization, 4, 9, 13, 15, 16, 17, 23, 31, 59, 197, 239, 252 good English correct English, 160, 176, 177, 178, 200 higher education, viii, 2, 13, 14, 15, 16, 17, 18, 19, 20, 22, 26, 27, 29, 50, 59, 186, 193, 197, 198, 204, 205, 206, 207, 210, 238, 243, 251, 252, 254, 256 identity, 21, 22, 36, 57, 167, 207, 239 idiomaticity, 46, 47, 176, 191, 253 IELTS, 1, 186, 187

262

Index

indirectness, 91 intelligibility, 43, 45, 64, 170, 191, 194, 240, 243, 250, 251, 252, 256 intensive analyses, 66, 67, 73, 76, 93, 94, 197 interactivity, 72, 180, 181, 239 interlanguage, 54, 235, 239 interrogative adverb, 89, 90, 101, 110, 111, 152, 153 interrogative pronoun, 104 interrogativity, 100, 101, 111, 155 intonation, 40, 88, 101, 102, 103, 104, 105, 107, 108, 109, 110, 111, 115, 116, 117, 151, 153, 154, 155, 159, 188, 247, 258 irritation, 32, 54, 55, 59, 63, 64, 79, 80, 138, 139, 140, 141, 145, 158, 159, 193, 198, 240 language policy, 10, 12, 27, 29, 43, 62, 197, 201, 205, 207, 208, 209, 244, 256 Latin, 7, 8, 25, 31, 248 learner language, 54, 55, 56, 77, 88, 147, 148, 149, 150, 160, 163, 164, 165, 168, 191, 199 learners, 15, 41, 46, 57, 153, 154, 160, 161, 162, 163, 164, 165, 170, 172, 173, 175, 176, 189, 190, 191, 192, 193, 194, 195, 196, 199, 235, 239, 242, 248, 250, 252 lecturing, 8, 179, 184, 185, 200 levels of proficiency, 32, 33, 162, 168, 177, 186 local languages, 11, 18, 21, 197, 206, 208, 244 metadiscourse, 39 MICASE, 1, 39, 48, 51, 63, 72, 132, 133, 174, 175, 249 misunderstanding, 34, 38, 77, 155, 156, 181, 249 morphology, 74, 92, 93 native speakers, 3, 4, 5, 21, 28, 31, 32, 33, 35, 41, 42, 43, 46, 47, 48, 54, 56, 57, 74, 80, 111, 122, 132,

158, 161, 166, 167, 173, 175, 177, 178, 188, 189, 193, 196, 197, 198, 199, 201, 202, 203, 204, 205, 211, 239, 240, 260 negation, 91, 92, 140, 142, 148, 149, 159 non-understanding, 38, 136, 259 normative elements, 32, 33, 43, 59, 198 not marking the plural on the noun plural marking, 82, 149 noun phrase NP, 82, 84, 85, 86, 93 objectivity, 78 observation scheme, 71 overt disturbance, 62, 63, 77, 78, 98, 110, 111, 113, 114, 115, 119, 122, 143, 144, 146, 151, 153, 159, 168, 193, 198, 199, 203 passive voice, 88, 140, 146, 150, 159 pedagogical, 44, 189, 238, 248 perceived communicativeness, 79, 80, 138, 140, 145, 147, 158, 166, 198, 199 pidgins, 28, 50 pitch, 40, 45, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 112, 113, 114, 115, 116, 117, 118 pragmatic strategies, 42, 53, 60, 62, 78, 79, 98, 123, 124, 125, 137, 144, 145, 155, 156, 157, 163, 168, 181, 182, 187, 192, 195, 199 pragmatics, 31, 32, 33, 35, 40, 41, 42, 43, 52, 59, 167, 168, 198, 243 prescription, 43, 191, 211 question intonation, 74, 98, 99, 101, 105, 107, 108, 109, 110, 111, 113, 114, 115, 122, 144, 153, 154, 155, 176, 199, 258 real-time signals, 119, 180 received pronunciation (RP), 45 recording recordings, 68, 69, 70, 71, 72, 76, 78, 79, 80, 113, 123, 173, 198

Index redundancy, 53, 143, 149, 150, 163, 168, 173, 194 science and technology, 2, 6, 7, 8, 9, 10, 11, 12, 13, 14, 31, 59, 197 seminars, 72, 78, 123 SLA second language acquisition, 63, 167, 240, 245 spoken grammar, 174, 175, 176, 253 subject-verb agreement subject-verb disagreement, 84, 86, 140, 145, 149, 158, 159 teachers teaching, 15, 26, 29, 42, 57, 75, 139, 158, 163, 185, 191, 193,

263

195, 200, 201, 202, 241, 242, 248, 256, 260 tense and aspect, 51, 87, 88, 150 TESOL, 1, 192, 235, 237, 245, 248, 249, 254, 256, 259 TOEFL, 1, 186 total duration, 110, 152 transcripts, 131 word order, 90, 91, 98, 101, 103, 139, 140, 145, 149, 150, 153, 155, 158, 159, 175, 193 Yes/No questions, 89, 90, 99, 101, 110, 122, 151, 153, 154, 155, 260