Universidade de Aveiro Departamento

de Línguas e Culturas

1999

Susan Jean English for Science and

Howcroft Technology

a Computer Corpus-based
Analysis of English Science and
Technology Texts for Application
in Higher Education.

Dissertação apresentada à Universidade

de Aveiro para cumprimento dos
requisitos necessários à obtenção do
grau de Doutor em Linguística, ramo
Linguística Inglesa, realizada sob a
orientação científica do Dr. Anthony
Barker, Professor Associada do
Departamento de Línguas e Culturas da
Universidade de Aveiro
 The Jury

President Reitor da Universidade de Aveiro

Vogais
Prof. Dr. Anthony David Barker
Professor Associado do Departamento de Línguas e Culturas da
Universidade de Aveiro

Prof. Dr. Carlos Alberto Agapito Galaricha

Professor Associado do Departamento de Comunicação e Arte da
Universidade de Aveiro

Profª. Drª. Maria Isabel Ponte Gonçalves de Salazar Casanova

Professora Auxiliar da Faculdade de Letras da Universidade de Lisboa

Profª. Drª. Belinda May Sousa Maia

Professora Auxiliar da Faculdade de Letras da Universidade de Porto

Prof. Dr. António Augusto de Freitas Gonçalves Moreira

Professor Auxiliar do Departamento de Didáctica e Tecnologia Educativa
da Universidade de Aveiro

iii
 Acknowledgements

First of all I would like to thank my supervisor Prof. Dr. Anthony

Barker for having given me such an immense amount of his valuable
time and for his detailed advice on my work without which it would not
be as it is.
Special thanks go to Prof. Dr. Belinda Maia of the University of
Porto who found the time in a very busy schedule to read my work, I
greatly appreciated her encouragement and very valuable advice on
computer corpora.
I am indebted to Prof. Dr. António Moreira for his patience and vast
knowledge, directing me to new sources of information, advising me on
statistics and generally having an answer to the many questions I have
put to him over the years I have been doing this work.
I should like to thank the students of the Ano Comum and my
colleagues in the University of Aveiro for their support. Particular thanks
must go to Dr. Silas Granjo who did so much to help me at the beginning
with the technical side of the computer work.
I should also like to thank everyone in the Department of Línguas e
Culturas in the University of Aveiro who helped to make this work
possible.
Last but not least, my sincere thanks to my family and friends for
their patience, advice and support. One friend and colleague in particular
stands out, Gill Moreira, for her tolerance, humour and good sense which
helped me to reach the end of this project.

v
 Abstract

This thesis presents two analyses: first the analysis of computer

corpora from undergraduate textbooks to isolate the (American) English
language of science and technology they present; secondly an analysis of
the English language competence of undergraduates starting their
university studies in science and technology. These two analyses are
contrasted in order to apply the results to the design of an English
language syllabus for first year undergraduates.
A frequency and range word list was produced using a large
baseline corpus to contrast with the main corpora taken from physics
and chemistry textbooks on the students’ bibliographies as a resource for
syllabus design. Secondly, four corpora, two main and two sub-corpora
produced from the physics and chemistry textbooks on the bibliographies
of the undergraduates were analysed using Biber’s (1988) algorithms and
functions for variation across speech and writing.
The student intake was tested over five years and the results of
those tests analysed. It was found that there was considerable variation
in the students’ levels of language competence. However, there was a
close correlation between the students’ competence and the number of
years they had studied English in secondary school. Nevertheless there
were students with extremely advanced competence and some with little
or no competence in English amongst the undergraduates.
Comprehension of scientific texts was generally found to correlate with
more advanced competence and more years of study.
The frequency and range word list showed the contexts which are
appropriate for materials to be used with these students and
demonstrated variation from many of the accepted views of the language
of science and technology. The computer corpora analyses varied from
Biber’s academic prose category. The sub-corpora demonstrated greatest
variation which is believed to be as a result of specific cultural and/or
literary material in the analogies used in the textbooks.
The heavy load of cultural background knowledge which the reader
would need in order to work with the textbooks adequately was also
found in the exercises the students were supposed to use for practice on
the topic presented in the chapter. This and the interpretation of visuals
in the textbooks were considered to be two principle factors that needed
to be emphasised in a syllabus for first year undergraduates. However,
given the time constraints on language teaching for science and
technology students, a methodology which would lead to greater student
autonomy is suggested using computer corpus-based studies - data-
vii
driven learning and computer-supported distance communications and
learning.

viii
 Resumo
Esta tese apresenta duas análises: primeiro uma análise de corpora
computadorizados, criados a partir de livros dos estudantes de
licenciaturas, para isolar a linguagem Inglesa (Americana) das ciências e
tecnologias que apresentam; segundo uma análise dos conhecimentos da
língua Inglesa que estes alunos apresentam ao iniciar os seus estudos
universitários em ciências e tecnologias. Estas duas análises são postas
em contraste para se aplicar os resultados obtidos ao desenho de um
programa de língua Inglesa para os alunos do primeiro ano.
Foi criada uma lista com a abrangência e a frequência das palavras
de um corpus de larga base, para ser contrastada com os principais
corpora compilados dos livros de física e química constantes das
bibliografias dos estudantes, como uma fonte para o desenho de
programas. Seguidamente, quatro corpora, dois principais e dois
subordinados, produzidos a partir dos livros de física e química referidos
nas bibliografias dos estudantes, foram analisados usando os algoritmos
e funções de Biber (1988) para variações entre linguagem falada e escrita.
Durante cinco anos, à entrada para a Universidade, os estudantes
foram submetidos a testes e os resultados analisados. Constatou-se que
havia variações consideráveis no nível de conhecimentos da língua por
parte dos estudantes. Contudo, havia uma correlação apertada entre as
competências dos estudantes e o número de anos que tinham estudado
Inglês nas escolas secundárias. Todavia, havia estudantes com
competências extremamente avançadas e outros com competências
reduzidas, ou quase nulas, em Inglês. A compreensão de textos
científicos estava geralmente correlacionada com os níveis mais
avançados de competências e maior número de anos de estudo.
A lista com a abrangência e a frequência das palavras mostrou os
contextos apropriados dos materiais a utilizar com estes estudantes e
demonstrou que havia diferenças em relação a muitos dos pontos de
vista aceites em relação à linguagem das ciências e tecnologias. A análise
dos corpora computadorizados varia das categorias da linguagem da
prosa académica de Biber. Os corpora subordinados mostram uma maior
variação, que se julga ser devida a materiais específicos, culturais e/ou
literário, usados nas analogias dos livros de estudo.
O grande peso dos conhecimentos de fundo de que os estudantes
necessitam para trabalhar adequadamente com os livros de estudo foi,
também, encontrado nos exercícios que necessitam de fazer para
praticarem o que está referido nos tópicos dos capítulos. Isto, juntamente
com a interpretação das imagens dos livros, foram considerados os dois
principais factores a precisarem de ser relevados no programa para o
primeiro ano dos estudantes. Contudo, atendendo às restrições de tempo
ix
para o ensino de línguas a estudante de ciências e tecnologias, a
metodologia que conduziria a maior autonomia dos alunos será baseada
na utilização de corpora computadorizados (data-driven learning) e
aprendizagem à distância assistida por computador.

x
 CONTENTS

Jury..................................................................................................................... iii

Acknowledgements............................................................................................ v

Abstract............................................................................................................. vii

Resumo .............................................................................................................. ix

Contents............................................................................................................. xi

Index of Figures............................................................................................... xvi

Index of Tables ............................................................................................... xvii

Abbreviations used ....................................................................................... xviii

Chapter 1 Introduction............................................... 3
1.1 Science and Technology Education ............................................11
1.2 Lifelong Education.....................................................................15
1.3 The Impact of New Technology ..................................................21
1.4 The Dominance of English in Science and Technology................26
1.5 The Situation in Portugal ...........................................................28
1.6 Science and Technology Undergraduates and English ................32
1.7 The Ano Comum..............................................................................34
1.8 Appropriate Text Types ..............................................................36
1.9 The Corpora ..............................................................................37
1.10 CD ROM Material ....................................................................38
1.11 The Syllabus ...........................................................................42
1.12 The Research...........................................................................43
1.13 Methodology ............................................................................44

Chapter 2 Historical and Theoretical Background to ESP

.................................................................................. 49
2.1 English for Special Purposes................................................51
2.1.1 Phrasebooks ...........................................................................52
2.2 The Register Analysis Approach ..........................................54
2.2.1 European Languages for Special Purposes ..............................55
2.2.2 Methodologies.........................................................................56
2.2.3 Scientific Specificity ................................................................57
2.2.4 Syllabus Implications .............................................................60
xi
 2.2.5 Publications and Coursebooks based on Register Analysis ......61
2.2.6 Criticism of Register Analysis..................................................62
2.2.7 The Impact of Modern Technology on Register Analysis...........64
2.2.8 Variation Studies ....................................................................66
2.2.9 Recent Studies........................................................................67
2.3 The Discourse Analysis Approach ........................................69
2.3.1 Definition ...............................................................................69
2.3.2 The American School ..............................................................70
2.3.3 The Prague School ..................................................................70
2.3.4 The British School ..................................................................72
2.3.5 Systemic Functional Grammar ...............................................73
2.3.6 Rhetorical Moves ....................................................................76
2.3.7 Organisational Features..........................................................77
2.3.8 Discourse Rules......................................................................78
2.3.9 Coursebooks based on Discourse Analysis ..............................80
2.3.10 Criticism of Discourse Analysis .............................................80
2.3.11 Educational Structures.........................................................83
2.3.12 Student Competence.............................................................84
2.3.13 Register and Genre Theory or Variation Studies ....................85
2.3.14 Discourse Analysis and Computers.......................................87
2.4 The Needs Analysis Approach ..............................................88
2.4.1 Needs and ESP .......................................................................88
2.4.2 The Development of Needs Analysis ........................................88
2.4.3 Needs and Syllabus Design .....................................................89
2.5 The Corpus Analysis Approach ............................................93
2.5.1 English Corpora Development.................................................93
2.5.2 Corpora Use ...........................................................................95
2.5.3 The Birmingham Corpus COBUILD.........................................98
2.5.4 The Lancaster-Oslo-Bergen Corpus.......................................100
2.5.5 The Brown Corpus ...............................................................102
2.5.6 The London-Lund Corpus .....................................................104
2.5.7 The British National Corpus .................................................106
2.5.8 The Longman/Lancaster Corpus...........................................107
2.5.9 Other Corpora ......................................................................108
2.5.10 EFL Student Corpora ..........................................................110
2.5.11 Specialised Corpora ............................................................110
2.5.12 Concordances .....................................................................111
2.5.13 Undergraduate Textbook Corpora .......................................112

Chapter 3 Research Methodology ........................... 117

3.1 Frequency and Range Word List ........................................120
3.1.1 Contrastive Analysis .............................................................121
3.1.2 Context.................................................................................122
3.1.3 Collocations..........................................................................123
xii
 3.1.4 The Baseline Corpus.............................................................125
3.1.5 The Level of the Material in the Corpora................................126
3.1.6 Previous studies and Text-Types ...........................................128
3.1.7 What are words?...................................................................129
3.1.8 Other Features of the Text and Corpus .................................132
3.1.8.1 American Words and Spellings...........................................133
3.1.8.2 Abbreviations.....................................................................133
3.1.8.3 Pronunciation Conventions ................................................134
3.1.8.4 Latin and Greek Influence..................................................134
3.1.8.5 Word Preferences ...............................................................136
3.1.9 Optical Character Recognition ..............................................137
3.1.9.1 Typographics .....................................................................138
3.1.9.2 Titles, Subtitles, Summaries and Conclusions....................139
3.1.9.3 Formulae, Numbers, Equations and Tables........................140
3.1.9.4 Diagrams and Drawings.....................................................141
3.1.10 Comparison with other published data ...............................142
3.2 Needs Analysis ..................................................................143
3.2.1 The Students’ Level of English ..............................................143
3.3 Biber’s Methodology of Variation Studies and Corpora
Analyses ..............................................................................151

Chapter 4 Test Results for New Students ................. 155

4.1.1 Student Numbers ............................................................ 155
4.1.2 The Preliminary Test ....................................................... 155
4.1.3 Test Results 1993/94...................................................... 158
4.1.4 Test Results 1994/95...................................................... 162
4.1.5 Test Results 1995/96...................................................... 166
4.1.6 Test Results 1996/97...................................................... 168
4.1.7 Test Results 1997/98...................................................... 170
4.2 Needs Analysis by University Department................... 175
4.3 Constraints ................................................................ 177

Chapter 5 Scientific English for Undergraduate Learners

................................................................................ 183
5.1 Analysis of Results ............................................................183
5.1.1 The Baseline Corpus.............................................................183
5.1.2 Range and Context ...............................................................188
5.1.3 American Words and Spellings..............................................190
5.1.4 Abbreviations .......................................................................191
5.1.5 Pronunciation Conventions...................................................191
5.1.6 Plurals from Latin and Greek ................................................192
5.1.7 Word Preferences ..................................................................194
xiii
 5.2 Other Features of the Text ................................................196
5.2.1 Typographics ........................................................................196
5.2.2 Titles, Subtitles, Summaries and Conclusions.......................198
5.2.3 Formulae, Numbers, Equations and Tables...........................199
5.2.4 Diagrams and Drawings .......................................................203
5.3 The Undergraduate Textbooks ...........................................205
5.3.2 The Physics and Chemistry Algorithms and Functions compared
with Biber’s Academic Prose....................................................208
5.3.3 The Physics and Chemistry Sub-Corpora ..............................214
5.3.4 The Physics Sub-Corpus: Gulliver’s Travels.............................221
5.3.5 Comparison with other Genres in Biber’s Variation Studies ....223
5.3.6 The Chemistry Sub-Corpus: Salvaging the Tapes from the
Challenger .......................................................................................231
5.4 Mathematics .....................................................................233
5.4.1 Mathematics in the Gulliver’s Travels Text ............................236

Chapter 6 Discussion of Results ............................. 241

6.1 Discussion of the Results .....................................................241
6.2 Coursebooks and Multimedia Encyclopedia Frequency and
Range Results............................................................................243
6.3 Textual Features Compared with Biber’s (1988) Variation
Results ......................................................................................247
6.3.1 Discussion of Dimension 1 ‘Involved versus Informational
Production’ ................................................................................251
6.3.2 Discussion of Dimension 2 ‘Narrative versus Non-Narrative
Concerns’ ..................................................................................257
6.3.3 Discussion of Dimension 3 ‘Explicit versus Situation-Dependent
Reference’ ..................................................................................262
6.3.4 Discussion of Dimension 4 ‘Overt Expression of Persuasion’.266
6.3.5 Discussion of Dimension 5 ‘Abstract versus Non-Abstract
Information’ ...............................................................................269
6.3.6 Discussion of Dimension 6 ‘On-Line Informational Elaboration’..
............................................................................................272
6.4 Academic Prose Sub-Genres.................................................274
6.4.1 Discussion of Dimension 1 ‘Involved versus Informational
Production’ for the Sub-Genres ..................................................276
6.4.2 Discussion of Dimension 2 ‘Narrative versus Non-Narrative
Concerns’ for the Sub-Genres ....................................................277
6.4.3 Discussion of Dimension 3 ‘Explicit versus Situation-Dependent
Reference’ for the Sub-Genres ....................................................278
6.4.4 Discussion of Dimension 4 ‘Overt Expression of Persuasion’ for
the Sub-Genres ........................................................................279
6.4.5 Discussion of Dimension 5 ‘Abstract versus Non-Abstract
Information’ for the Sub-Genres ................................................282
xiv
 6.4.6 Discussion of Dimension 6 ‘On-Line Informational Elaboration’
for the Sub-Genres ...................................................................284
6.5 The English of the Students in the First year of University....284

Chapter 7 The Syllabus........................................... 291

7.1 Study Skills ...................................................................... 292
7.2 Student Needs and the Syllabus........................................ 296
7.3 The Students’ Background knowledge and the Syllabus .... 299
7.4 Data-Driven Learning........................................................ 304
7.5 Methodological Implications .............................................. 319
7.6 Modern Technology and the Syllabus ................................ 324

Chapter 8 Conclusion ............................................. 331

Bibliography ............................................................. 341

Appendices............................................................... 369

xv
Figures
3.1 Pie Graph for the Academic Year 1993/94 showing the Students’
Number of Years of English ........................................................145
3.2 Pie Graph for the Academic Year 1994/95 showing the Students’
Number of Years of English ........................................................146
3.3 Pie Graph for the Academic Year 1995/96 showing the Students’
Number of Years of English ........................................................146
3.4 Pie Graph for the Academic Year 1996/97 showing the Students’
Number of Years of English ........................................................147
3.5 Pie Graph for the Academic Year 1997/98 showing the Students’
Number of Years of English ........................................................147
6.1 Dimension 1 ‘Involved versus Informational Production’ ...........250
6.2 Dimension 2 ‘Narrative versus Non-Narrative Concerns’...........256
6.3 Dimension 3 ‘Explicit versus Situation-Dependent Reference’...261
6.4 Dimension 4 ‘Overt Expression of Persuasion’ ..........................265
6.5 Dimension 5 ‘Abstract versus Non-Abstract Information’ ..........268
6.6 Dimension 6 ‘On-Line Informational Elaboration’ ......................271
6.7 Dimension 1 ‘Involved versus Informational Production’ for the
Academic Prose Sub-Genres............................................................275
6.8 Dimension 2 ‘Narrative versus Non-Narrative Concerns’ for the
Academic Prose Sub-Genres............................................................277
6.9 Dimension 3 ‘Explicit versus Situation-Dependent Reference’ for
the Academic Prose Sub-Genres......................................................278
6.10 Dimension 4 ‘Overt Expression of Persuasion’ for the Academic
Prose Sub-Genres ............................................................................280
6.11 Dimension 5 ‘Abstract versus Non-Abstract Information’ for the
Academic Prose Sub-Genres............................................................281
6.12 Dimension 6 ‘On-Line Informational Elaboration’ for the Academic
Prose Sub-Genres ............................................................................283

xvi
Tables
2.1 Munby’s Communicative Needs Processor ...............................90
2.2 Texts, Categories and Numbers of Words in the LOB Corpus .101
2.3 Texts, Categories and Numbers of Words in the London-Lund
Corpus .......................................................................................104
2.4 Categories and Percentages in the British National Corpus....106
2.5 Conversation in the British National Corpus .........................107
2.6 Number of words and texts for Academic Prose and Fiction in the
Longman/Lancaster Corpus ......................................................108
3.1 Huddlestone’s Level of Science Texts ....................................126
3.2 Darian’s Level of Text and Audience .....................................126
3.3 Students’ Number of Years of Study of English.....................143
4.1 Analysis of 1995/96 Test Results by Item.............................166
4.2 Analysis of 1996/97 Test Results by Item.............................168
4.3 Analysis of 1997/98 Test Results by Item.............................171
5.1 Grolier Frequency and Range List.........................................185
5.2 Frequency and Range Results for Abstract Nouns and Adjectives
............................................................................................188
5.3 Normalised Frequencies from the Main Corpora compared to
Biber’s Academic Prose with Statistical Significance Values (chi-
square χ2) ..................................................................................209
5.4 The Physics Main Corpus: Significantly Higher and Lower
Results ............................................................................................
212
5.5 The Chemistry Main Corpus: Significantly Higher and Lower
Results ......................................................................................212
5.6 Normalised Frequencies from the Sub-Corpora compared to
Biber’s Academic Prose with Statistical Significance Values (χ2)..215
5.7 The Physics Sub-Corpus: Significantly Higher and Lower
Results ............................................................................................
218
5.8 The Chemistry Sub-Corpus: Significantly Higher and Lower
Results ......................................................................................219
6.1 Mean scores of each of the Dimensions compared with Biber’s
Academic Prose corpus results...................................................249
6.2 The main physics and chemistry corpora compared with Biber’s
Academic Prose sub-Genres .......................................................274
6.3 The physics and chemistry sub-corpora compared with Biber’s
Academic Prose Sub-Genres ......................................................275

xvii
 Abbreviations used in this thesis:

CBL Computer Based Learning

CD-ROM Computer Disk – Read Only Memory

DDL Data Driven Learning

EAP English for Academic Purposes

ELT English Language Teaching

EOP English for Occupational Purposes

ESP English for Special Purposes

EST English for Science and Technology

EGAP English for General Academic Purposes

ESAP English for Specific Academic Purposes

ICT Information and Communication Technology

LSP Languages for Special Purposes

MT Mother Tongue

OCR Optical Character Recognition

xviii
Chapter 1 Introduction
 Chapter 1

Introduction

Since the early 1980’s I have been fascinated by the use of

computers in language teaching. From the moment the BBC computers
became available, it was possible to begin to use computers in the
classroom with students as computers had become small, relatively
cheap and above all reliable machines. Nevertheless, first of all teachers
themselves needed to find out how to work with the computers and, as
there was very little software available, to design and write their own
programs to use in class. However, projects were soon started to try to
bring some system and principle into software design for educational
purposes1 and conferences and workshops helped to divulge information
and provoke reflection on the role of computers in the classroom (Higgins
1985, Jones and Fortescue 1987, Evelyn Ng and Olivier 1987). Very soon
the machines were updated and the amount of memory the new
machines made available increased. This meant that the tedious, time
consuming cassettes used to load programs changed to floppy disks
providing relatively greater speed and also greater user-friendliness. The
rate of change continued with the American IBM computers coming to
dominate the education market because they were both more powerful
and cheaper. This led to a state of confusion in many institutions which
now had a mixture of different hardware and software, different sizes of
floppy disks and programs that would not work on some computers

1
For example, John Higgins’ and then later Martin Phillips’ work with the British Council Project to
develop software for language teaching much of which was published in the 1980’s in collaboration with
Cambridge University Press, initially for the BBC computers and then for IBM compatible
microcomputers.
3
because they had incompatible operating systems. This situation was
made even more difficult as many academics had chosen the Apple
Mackintosh computer as the most suitable for academic research.
However, at this stage, education policy was encouraging the use of
computers and the teaching of information technology as the use and
application of computer technology came to be known2. This meant that
towards the end of the 1980’s teacher training courses were beginning to
include CALL training (see Birnbaum 1987:19-20, Heppel 1987:20-21).
Being at the cutting edge of the technological revolution was seen to be of
prime importance not only to teachers (Dunn and Morgan 1987) but also
to governments who believed that their future economic success in the
world depended upon this change in education (see later 1.1 Science and
Technology Education).
The availability of computers and their more widespread use in
education also led to changes in the way that teachers prepared their
work. Initially the opportunities for word processing made a big change in
the preparation of materials and tests. Teachers began to write their own
materials directly through the computer rather than relying on the
support of a secretary or on the traditional cut and paste techniques
which followed widespread use of photocopying. Predictions were made at
this time that we were on the verge of the “paperless” office. It was
believed that the need for paper filing systems would disappear because
computer data was stored on floppy disks. It is now recognised that the
contrary is true, the use of computers has led to far more paper being
used as people who once would not have written anything themselves
began to do so and documents can be revised much more easily, leading
to more and more printouts of documents as changes are made to them
in order to achieve greater accuracy or to bring documents up-to-date. As

2
In Portugal the MINERVA Project to introduce Information Technology across the curriculum in state
schools finished its pilot phase in 1989 (José Moura Carvalho, 1991).

4
in the example of teachers preparing their own materials through word
processing, it became possible to produce much more specific, tailor-
made material for individual classes and so more and more materials
have been produced. Equally well, the mixture of computer operating
systems has led people to be much more careful of how they store their
data. Paper is much more accessible than a floppy disk which is the
wrong size for the current computer system or which cannot be read by
the latest machine. Contrary to popular belief some years ago (Jones and
Fortescue 1987:129) computer hardware has shown that it is prone to all
kinds of mechanical breakdowns and floppy disks often become corrupt
at the most inconvenient moments.
However, computers were also being used to investigate language
itself and huge projects were set up in universities, some on artificial
intelligence (particularly in America with the ELIZA project) and others on
lexicography and dictionary writing (such as the COBUILD Project in
Birmingham). These projects have given way to Natural Language
Processing which is basically a sub-field of computer science directly
related to artificial intelligence, human computer interaction, machine
translation and multimedia and the Bank of English which is now used
for linguistic analysis of general language. They have been joined by such
projects as the European funded ELRA project which is the European
Language Resources Association which aims to include such things as
recorded speech databases, lexicons, grammars, text corpora and
terminological data in collaboration with European countries. There has
also been a burgeoning of corpus work in many European countries
themselves. The Universities of Oporto and Lisbon in Portugal are cases
in point carrying out translation studies and linguistic analyses of
corpora under the supervision of Profs. Belinda Maia and João Malaca
Casterleiro, respectively. It has now been seen that with appropriate
software (Tribble and Jones 1990) that teachers could also carry out

5
linguistic research themselves and because of this teachers carried out
work on error analysis of students’ errors (for example, in Portugal,
Fordham 1997) and the use of concordancing for both teaching and
research developed.
Coupled with this interest in computers and computer-assisted
language learning and research was an interest in special or specific
language teaching. In 1986, I was required to design and teach a special
course for people working for the post office (then the CTT - Correios,
Telégrafos e Telefones, see Howcroft 1986). This particular course was for
personnel working with computers in the head office of the post office in
Coimbra and required reflection and research into the language of
computers together with consideration of an appropriate methodology for
a mixed ability group of foreign language learners dealing with
computers. This was a special situation in which the English language
was to be learnt through specific activities using a computer as part of
the process of learning. The use of computers was also the product or
goal of that language learning situation which made it particularly
stimulating. The fact that English is the main language of the computer
is important in any consideration of teaching students with or through
this technology. Stubbs (1992:203) says that the Cox Report (Department
of Education and Science 1989) on which he worked also argued that
most interactions with computers were language experiences. (This is
taken up again later in 1.4 The Dominance of English in Science and
Technology) However, those that design the programs and operating
systems of IBM computers are not linguists and they, therefore, cannot
be expected to have taken into consideration the fact that their audience,
the user, will often be a foreign language learner in a country far away.
The language of the computer is often idiosyncratic from a linguistic
point of view, but it has had and is increasingly having an effect on
English language usage in the modern world. The number of new terms

6
and concepts that are now employed because of the common use of
computers and electronic communications is legion. The latest
communications through the Internet and electronic mail have only
served to emphasise this state of affairs, as has the huge increase in the
amount of information available to computer users in the 1990’s through
the Internet and CD-ROMs containing the equivalent of whole
bookshelves of knowledge. Crystal (1998) describes the Internet as a
“semiotically sanitized medium” because of restricted turn taking and the
fact that messages are received in order, one by one but he reminds us
that we are not at the end of the technological road and that other
technologies are still to come which may change all this. The language
student of today faces a much greater barrage of applications and specific
language which educational policy makers believe the working population
should be capable of handling efficiently and rationally to maintain
economic advantage in the world but this language will almost certainly
be constantly changing as the technology itself changes. The types of
education policy that are relevant to university undergraduates will be
discussed in 1.1 Science and Technology Education and 1.2 Lifelong
Education.
In order to design a syllabus for a modern day undergraduate
student, whose first language is not English, all of these technological
innovations and changes in language usage will have to be taken into
account. Moreover, the sheer size and detail of the available knowledge
on any subject which the student has to cope with also requires new
learning strategies to be found. Education itself has had to change and
will have to continue changing under the weight of what is now known
and needs to be learnt by students today so that they can keep abreast of
their subject. Political changes that have taken place such as joining the
European Union have also had and are continuing to have an effect on
educational policies within the member states. Harmonisation of policies,

7
specific language training and cultural studies are deemed to be
important for the European citizen (van Ek 1990, Kubanek 1998, Byram
1999). The Committee of Ministers stressed the political importance of
intensifying and diversifying language learning as recently as 1998 when
they reviewed the Council’s earlier initiatives (Byram and Riagain 1999)
The attempt to standardise courses and the subsequent qualifications
obtained from them to allow recognition of academic qualifications
throughout the member states is also seen as an important aspect of
educational harmonisation. The whole issue of what skills and what
knowledge are needed by the people who will make up the workforce in
the future has led to changes in the perceptions of learning as will be
discussed later in 1.2 Lifelong Learning.
Although much research has been carried out in the past, the
changes that have taken place over the past two decades mean that the
present situation is often quite different from the ones those studies refer
to. The English of science and technology and the teaching/learning of
English as a foreign language of undergraduate students studying
languages as part of their courses has changed because of all the issues
discussed above. Whilst some of the findings from previous research will
remain valid, other findings will be called into question. This is especially
so because of the possibility of conducting research on specific situations
using up-to-date computers with their much greater memories, speed
and capabilities. In some cases, as will be argued later in 1.10
Appropriate Text Types, there has been little scientific rigour or little
information that would show the relevance of the work carried out for the
undergraduates on science and technology courses in Portugal. Swales
argues against overgeneralising and applying solutions found to work in
situation x to situations y and z as he (1985:188) suggests “there are
rarely global solutions to local problems”. In other cases, there has been
no published research at all on the specific kind of English that the

8
undergraduates will come into contact with (see later 2.5 The Corpus
Analysis Approach).
The undergraduate of today is also in many ways a different entity
than undergraduates in the past. Many more students will arrive at the
university with expectations about and knowledge of computers and their
applications. This may be because of stimulating projects that were
carried out in their schools in different subject areas and through
computer clubs or because of the availability of computers in their homes
and visits to ‘cyber’ cafés. Not only has this aspect of education changed
the profile of the average undergraduate but also changes in language
studies and other subjects in schools will affect the undergraduates’
profiles. Moreover, the universities themselves are just as vulnerable to
change as the schools and the students are from the impact of computers
and their applications. In fact, the universities lead in this change by
influencing the kind of training teachers are given (see 1.5 The Situation
in Portugal). New possibilities for different learning systems have been
opened up and explored by universities using modern communication
systems. What and how language should be learnt in universities by
undergraduates studying science and technology must therefore be
considered in the light of the above changes. Some form of quantification
of the changes that have taken place in both the students and their
subjects must be carried out to take principled decisions about the
syllabus that would be suitable for these particular undergraduates.
The need for this particular piece of research therefore arose with
the advent in the University of Aveiro of the combined first year for
students of science and technology. These students had English as one of
their core subjects in this first year and so it was felt that a syllabus had
to be designed to meet these students’ needs. As Swales (1985:188) says,

9
 If those of us in ESP have thought long and hard about how best to serve
our students’ interests, it is simply because circumstances have tended to
make us do so. In circumstances of restricted educational opportunity we
have been forced to search out ways of providing maximum educational
value.

The belief that English is of great importance for students of science

and technology as discussed later (in 1.4 The Dominance of English in
Science and Technology) led to the need for research that would benefit
the students and provide them with “maximum educational value”. This
was particularly important as the discipline was to last only one year and
the students’ needs were expected to go far beyond this limited time scale
right through their working lives. The questions that needed to be
answered therefore were:

• What was the level of English of the students taking the course? In
other words, what did the students already know or what had they
already learnt prior to starting their undergraduate studies?

• What English did these students need to know? In other words, what
kind of interaction with the English language could these students be
expected to have and what was the nature of that English?

The difference between the results of these questions should give

the answer as to what it is that the students need to learn. The
information obtained must, in turn, form the heart of the syllabus drawn
up for these students.
The second of these questions can be answered partly by looking at
what the students need English for which will have at least two aspects
to it. The general situation in the country, in Europe and the world will
impinge upon these students just as much as the very specific context in
which they find themselves in the university will affect their needs.

10
 Considering the first of these aspects, the more general situation,
we can see that the world is changing ever more rapidly and general
educational priorities are also changing to prepare people for the modern
world. When, where and how people study are undergoing changes
brought about by changes in the technology of communication. These
changes will have a number of effects on undergraduate students who
are being prepared to take their places in society. The two strands of
education policy and technological change will have an effect on
university courses themselves and new curricula and courses are being
and will be started. Course content and even the structure of courses will
change. The level or length of the courses will change and different
systems, for example a modular system, and different academic
timetables are being and will be experimented with. Schemes to
understand and standardise different countries’ credit systems for
studies in order to allow movement across borders between different
countries’ education systems and recognition of educational
qualifications or part qualifications have started and will become more
usual and widespread. The materials published for use on university
courses will change and modern technology will have a profound effect on
those materials and even on the means by which they are delivered.
Contact with professors can now by means of e-mail and lectures and
notes to accompany coursework be accessed through university
computer networks. Assignments carried out by students can be
produced and printed in sophisticated styles, transmitted electronically
to the professor and comments received through the same channel. The
actual information content of the work done by undergraduates can be
affected by the information that they can obtain ‘on-line’. The skills
undergraduates are expected to have and to develop through their
courses will therefore also undergo profound changes. Stubbs (1992:220)
says that

11
 The evaluation of educational change due to the new technologies involves
the analysis of changed cognitive and social relations in the classroom. We
therefore need simple but powerful concepts to study the pedagogic and
cognitive logic of such situations.

1.1 Science and Technology Education

The attitudes towards science and technology education have first

to be seen in the light of educational policies. Educational policies are not
fundamentally different in the countries of the developed world and, as
was mentioned above, the European Union countries are endeavouring to
harmonise their policies on education.
There is an increasing preoccupation with science and technology
education in the world in general and in the more developed countries of
the world in particular because skill in science and technology in an
educated workforce is seen as a means of maintaining or obtaining a
position with the economic front runners in the modern world. There was
felt to be a ‘crisis’ in science education in America in the 1970’s and
1980’s when fewer students were studying science at university and
there was dissatisfaction with the lack of scientific knowledge in the
general population (Matthews 1994). This preoccupation has been
gaining pace since the 1980s when the small, robust, inexpensive
personal computer was introduced into schools in Europe and America
on a large scale and also because educational policy-makers felt that
changes needed to be made in order to develop a workforce which could
handle such technology. The developed countries’ educational policies
were widely studied to find common ground which would ensure that
people were trained successfully for the challenges the future was
expected to bring. European countries particularly developed many
12
programmes to study and make recommendations on all aspects of
educational policy from specific course content (for example, on language
Trim, Richterich, van Ek and Wilkins 1973/80) to mutually acceptable
accreditation schemes (ECTS – European Credit Transfer System). What
the United States of America, as the leading nation in the world, does to
try to solve the crisis in education has an effect on the rest of the world
and the policies adopted and technology produced and developed in the
USA will affect education in Europe.
In the U.S.A. in 1985 Frank Press, then President of the National
Academy of Sciences, Robert White, then President of the National
Academy of Engineering and Winston Lord, then President of the Council
on Foreign Relations wrote in the Preface to Keatley (ed. 1985:iii)
Technological Frontiers and Foreign Relations, “If one compares the world
of today with that of one or two decades ago, it is clear that modern
science and technology have become profound influences on economies,
societies, and international relations.” They also suggest that “This
process certainly will continue, and may accelerate.” The result of which
will be that it “will deeply affect American foreign policy, whether with
friends or adversaries, competitors or collaborators, rich or poor.” They
may not have foreseen the changes that have taken place in American
relations in the ensuing years but the position they outline is still having
an effect on educational policies throughout the world. In the case of the
developed countries this is so that economic and financial positions can
be maintained in relation to the USA. In Britain Winston Churchill is
reputed to have predicted that “The empires of the future are the empires
of the mind.” Churchill’s comment implies, amongst other things, that
education must be given priority in order to be successful.
Similarly, the European Round Table of Industrialists in its report
Investing in Knowledge: The Integration of Technology in European
Education (February 1997:5) says that “European society is running the

13
risk of an increasing mismatch between the requirements of our new
environment and the capabilities of our people.” This report (ibid:5)
identifies particularly noticeable mismatches between intellectual
aptitudes in the areas of maths/sciences/technology and behavioural
aptitudes which lead to “professionalism, excellence, (and) distinctive
competitive edge.” Furthermore, this emphasis on learning and using
science and technology is seen as a problem. Allen Luke (1992), the
series editor of Critical Perspectives on Literacy and Education, writing in
the Introduction of Halliday and Martin’s Writing Science (1993) says that
the “very dependency on corporate science and technology expansion as a
means for the expansion of state power and legitimacy have translated
the crises of economies and cultures into the crises of sciences.”
Matthews (1994) regards this as the narrow ‘economistic’ view of
education which is designed merely to develop ‘human resources’ so that
countries can overcome their balance of payments deficit, or stay
competitive with other economies. He believes that there is a need for a
much more liberal science education which endeavours to develop
scientific literacy in students which includes the understanding of
concepts and learning about the nature of science both through its
historical and social dimensions.
Nevertheless, because science education is seen by politicians
particularly in this narrower economistic way, different countries are
dedicating time and money to research and development in science and
technology teaching. Mike Robinson (1994) from the University of
Nevada describes why American government funding is being directed
towards science and technology education and the training of teachers.
He says this is because “science, technology, engineering and
mathematics (STEM) are usually singled out as the most pressing
educational areas” as shown by the Carnegie Commission on Science,
Technology and Government Report (1991): In the National Interest: the

14
federal government in the reform of K-12 math and science education. New
York: Carnegie Corporation. This report (1991:7) also mentions the
problem of proficiency in the English language saying that in the year
2000 ‘one child in twelve will lack the English language proficiency
required for learning’. Robinson suggests that the national interest is
best protected by training “hi-tech” workers in order to “maintain the
diminishing US technological advantage in the world economy (Office of
Science and Technology Policy, 1992 Science and Technology.
Washington: Executive Office of the President.)”. The fact that the
economic situation of the USA has changed for the better since 1992
might make the authors of the report change their minds about whether
the US is in fact losing its technological advantage in the world, but the
concern is still to produce the right profile of a technologically competent
workforce. Amongst other things, in order to achieve this goal Robinson
places particular emphasis on the use of E-mail and the Internet in
science teaching. The combination of science and technology education
and the use of modern technology is a common theme amongst
educators (Laurillard 1993).

The American Association for the Advancement of Science, Project

2016 was dissatisfied with the failure of curricular changes made in the
1960s as a response to the Russian launch of the sputnik satellite. The
curricula in this period were designed by scientists rather than teachers
and became over full with facts. The report, 1989:5 states that the
curriculum they recommend should have the following effect:

To ensure the scientific literacy of all students, curricula must be changed

to reduce the sheer amount of material covered; to weaken or eliminate
rigid subject-matter boundaries; to pay more attention to the connections
among science, mathematics, and technology; to present the scientific

15
 endeavor as a social enterprise that strongly influences – and is influenced
by – human thought and action; and to foster scientific ways of thinking.

1.2 Lifelong Education

Coupled with the preoccupation with science and technology

teaching is a desire to improve the education of the workforce both
through school curricula and through further education programmes to
produce an elite workforce which can cope with the demands of this new
technological world. One means that is seen as being capable of doing
this is to develop flexibility in workers so that they can be retrained to do
different work whenever the need arises. Such workers must not expect to
do only one kind of work throughout their lives. The average number of
jobs workers are expected to do in their working lives in the future ranges
from four to fifteen rather than one lifelong career. This means that
people must accept the need for sustained learning in order to face the
different challenges they are expected to meet in their different work
situations. Differences in working arrangements are also expected
because of demographic changes. The workforce is ageing in many of the
developed countries and so the size of the working population in
comparison with the population as a whole is expected to fall which
means that new work patterns and perhaps a longer working life will have
to be accepted in the next few decades if countries are to maintain their
actual economic positions. The process of retraining people throughout
their careers has come to be seen as a process of Lifelong Learning; an
idea which was put forward in the 1970’s with the British HMSO (1973),
OECD (1973), UNESCO (1973) and Council of Europe CCC (1972, 73 and
74) reports on adult education.
President Clinton (1998:59) stated in his State of the Union Address
that the “Information Age is, first and foremost, an education age, in

16
which education must start at birth and continue throughout a lifetime.”
However, the means to achieve this objective often appear elusive. Luke
(1992) suggests that there is a serious ‘time-lag’ between the debate about
educational change and the actual ‘remaking of science education’. He
discusses the work done by Lingard, Porter and Knight, (1992) who say
that the post-war human capital model of education3 in the USA and
Canada, the UK and Australia, has proved resilient and recyclable,
despite there being little evidence that it works. Gerald W. Bracey
(1997:52) suggests that “The biggest threat to the American educational
system may come not from within our schools but from the depth of our
divisions over what exactly they should accomplish and how best to get
them to accomplish it.” He goes on to point out that there has been an
enormous shift in policy which has led to 62 percent of high school
graduates being deemed capable of studying at a higher level and
therefore enrolling in college as opposed to 20 percent after the Second
World War. The massification of education is another factor that will
affect teaching and learning expectations and goals. These changes will
lead to a different, wider range of courses and qualifications becoming the
norm with shorter, less sophisticated and more modular courses being
given as and when required by the individual.
Educational policy emphasising more students enrolling in higher
education is also reflected throughout the European Union and its
member state Portugal. The proliferation of new universities and
polytechnics in Portugal in the last two decades is a clear example of the
increasing need to provide places for the greater number of students
leaving school who could benefit from further years of study and higher
qualifications4. The recent changes in the number of years of obligatory

3
This model of education is one which stresses universal education that is, education for all a nation’s
children without financial or other constraints.
4
The change that took place in Britain in 1995 when all the higher education institutes became known as
universities is also an example of this phenomenon.
17
schooling requiring students to complete the ninth year also reflects these
changes in educational policy. Despite the fact that Portugal is one of the
countries in Europe with the lowest rate of unemployment, further
education is seen as a means of achieving greater prosperity and
increasing the chances of finding a ‘good’ job. Universities themselves are
in competition with each other to provide more up-to-date courses which
give students the preparation necessary for the types of employment in
the Europe of tomorrow. Indeed recently much forward planning has gone
into attracting new faculties or universities to cities in the interior of the
country, which bears witness to the importance which is attached to
higher education in developing the poorer less industrialised ‘interior’
regions. Educational institutes are also seeking to forge links with
industry in order to work towards providing professionally oriented
training for the local workforce.
Countries throughout the world are constantly producing league
tables of the most advanced economies and comparing and examining the
ability of different educational systems to produce the best scores on tests
of mathematics and science subjects and to equate these findings with
spending on education, number of hours devoted to these subjects in
school, amount of homework set, and other parameters to try to discover
the most successful formula to produce the elite workforce deemed
necessary in the future. The world of education is fraught with insecurity
as to what constitutes a technocrat’s training and how to evaluate
‘quality’ in education. ‘Standards’ of education normally translate as the
results of tests. Recently testing carried out for the Third International
Maths and Science Study (TIMSS, 1997) produced a league table of
nations which showed the concern that many countries feel about their
results in such comparisons. Both Britain (England 25th in Maths and
10th in Science, Scotland 29th and 26th respectively) and America (28th

18
in Maths and 17th in Science) feel that they are doing “poorly” and the
Economist (March 29th 1997) reports that
In a television interview in December, the French president, Jacques
Chirac, described as “shameful” a decision by his education ministry to
pull out of an international study of adult literacy which was showing that
the French were doing badly. And in Britain last year Michael Heseltine,
the deputy prime minister, brushed aside objections from officials in the
Department for Education and Employment, and published the unflattering
results of a study he had commissioned comparing British workers with
those in France, America, Singapore and Germany – chosen as key
economic competitors.
The Germans, in turn, were shocked by their pupils’ mediocre
performance in the TIMSS tests. Their pupils did only slightly better than
the English at maths, coming 23rd out of 41 countries. In science, the
English surged ahead (though not the Scots) while the Germans were
beaten by, among others, the Dutch, the Russians – and even the
Americans. A television network ran a special report called “Education
Emergency in Germany”; industrialists accused politicians of ignoring
repeated warnings about declining standards in schools.

It could be argued that if major countries perceived to be

economically successful like Germany and America are dissatisfied with
their placement on these scales, something is fundamentally wrong with
the yardstick being used for measurement. Controversy over the
yardsticks used to evaluate children in different educational settings was
indeed one of the reasons that the British National Foundation for
Educational Research refined the test for the latest study, giving the
teachers concerned precise instructions on how the test was to be carried
out and monitoring schools at random. Nevertheless, the attitudes
expressed highlight the emphasis that governments are giving to

19
education as a means of achieving success in world markets. Success in
education and commercial success are seen as going hand-in-hand.
The American Secretary of Education, Richard W. Riley, trying to
justify the American results, (1997:60) says, “students’ proficiency in
science and math is up about one level compared to what it was a
decade ago. One reason we have been behind countries such as Japan is
because that nation’s public schools always have put extremely heavy
emphasis on science and math. We still have a long way to go.”
President Clinton, not surprisingly, in a speech to the National
Association of Black Journalists (July 17, 1997) put a much more
positive interpretation on the International Math and Science Test
results. He claimed that recent results for 4th and 8th graders showed
an improvement which in turn proved that his policies were working and
that therefore America could achieve “international excellence in
education”. Eight Goals have been identified by the National Educational
Goals Panel which was set up in 1990, these are: Goal 1: Ready to
Learn; Goal 2 : School Completion; Goal 3: Student Achievement and
Citizenship; Goal 4: Teacher Education and Professional Development;
Goal 5: Mathematics and Science; Goal 6: Adult Literacy and Lifelong
Learning; Goal 7: Safe and Disciplined, Alcohol and Drug-Free Schools;
Goal 8: Parental Participation. There is also a similar list of priorities
from the U: S. Department of Education (February 1997):
“All students should be able to:
1. Read independently by the end of the third grade.
2. Master challenging mathematics, including the foundations of algebra and
geometry, by the end of the eighth grade.
3. Be prepared for and be able to afford at least two years of college by age
18, and be able to pursue lifelong learning as adults.
4. Have a talented, dedicated, and well-prepared teacher in their classroom.
5. Have their classroom connected to the Internet by the year 2000 and be
technologically literate.
6. Learn in strong, safe, and drug-free schools.
7. Learn according to challenging and clear standards of achievement and
accountability.

20
 The differences between these two lists is one of specifying in more
detail when and what is necessary in education for the future by the
Department of Education, such as the need to be able to afford further
education and the year 2000 being given as the objective for Internet
connection.
One of the policies introduced into the American education system,
as point 4 on the list of priorities shows, is the testing of both teachers
and pupils. A similar system to that found on the American National
Educational Goals Panel website (http://www.ed.gov/pubs/StratPln/priority.html), which
allows people to find out about different states and their educational
achievements, is one which has been introduced into European
educational systems where testing and grading of results from school to
school to compare those schools that are doing well with those that are
doing badly. The idea is that the better schools can be used to show
what should be done to achieve the desired test results and that
teachers can benefit from visiting those excellent schools to learn about
their methods which they can then apply in their own schools to improve
standards5. A similar system of analysing which education systems
teach science and maths best is forecast to explain what conditions are
necessary to promote effective learning.
The Organisation for Economic Co-operation and Development
(OECD) has collected data on how governments spend their combined $1
trillion annual education budgets and explains that their new studies
(launched December 1996) will compare how schools, colleges and
universities are run in each country and analyse the implications for
policy makers. The fact that some countries with low education budgets
achieve high scores on the TIMSS tests has caused politicians to seek
alternatives to more spending on education. Similarly class sizes vary

5
This is similar to the European LEONARDO project which encourages the movement of professionals
between countries in order to find and emulate excellence in teaching.
21
from country to country but results do not seem to support the
contention that only small classes achieve good results in science and
maths. The methodologies used to teach these subjects appear to be as
important as class sizes are.
As there is no consensus on what causes optimal learning,
research for practical application on undergraduate disciplines is, in the
light of the above concerns, an essential prerequisite to aid success in
science and technology education in a foreign language. One of the
implications of Lifelong Learning is that the emphasis in teaching should
be on the learning process itself and not the product of that learning, so
that people learn how to learn rather than learn a particular finite body
of knowledge.

1.3 The Impact of New Technology

Many other changes have taken place in the last three decades
which will also influence the teaching and learning situations of students
of Science and Technology. One of the principal changes is the advent of
the personal computer with sufficient memory and processing speed to
enable specific situation research work to take place but the implications
of the personal computer go much further than this. Students of English
and those involved in science will find their lives are surrounded by the
specific English of the computer, the associated word processor,
multimedia applications and increasingly the world of electronic mail and
the Internet. It is also the case that the software and computer
communication systems that dominate the world, such as the Internet,
originate from America and are therefore usually originally written and
manipulated through English.
Bucy describes the differences that have taken place in the concept
of what a computer is. He (1985:46) explains how the advent of the

22
silicon chip has made it possible to incorporate the computing power of a
main frame computer into such products as microwave oven controls and
handheld calculators. He also suggests that even by 1985 computers
were thought of as relatively inexpensive machines that can be used in
“any number of activities, such as education, household data storage,
increased job productivity, or entertainment.”
Consequently this greater capacity for data storage has also opened
up the possibility of conducting individual, empirical research into the
language of science and technology in a manner that was unthinkable
only a few decades ago. New software has also been developed to allow
this kind of research to take place as a result of work on corpora (see 2.5
The Corpus Analysis Approach). The problems associated with obtaining
information through the Internet seem to be much more a problem of
obtaining information about where data is available in such a vast
resource and of framing the right sort of question to obtain the desired
result. If the question posed is too general, the enquirer will be inundated
with information which will be hard to sift through in order to locate
appropriate data within a reasonable amount of time. If however the
questions posed are too precise, little or no information may be obtained.
Often the answers to questions surprise because the area the enquirer is
contemplating does not match the results of the search. For example, a
search for data on “bands” would produce results on both the musical
variety, i.e. brass bands, and the electronic forms, i.e. wave bands, either
of which would be inappropriate if the enquirer wished information on
the other. The significance of the data obtained also has to be judged by
the enquirer to ascertain if it is of an appropriate level of sophistication
which in turn requires sufficient background knowledge of the subject by
the person requesting information.
The prominence and utility of modern technology urges therefore
the teaching of a foreign language that is somewhat different from that

23
taught in schools, although many schools are taking part in exciting
projects including distance communication and e-mail. The combination
of modern technology and language use has even created new styles of
language. This difference between the new style of language and other
styles will also have an effect on teaching methods. Teaching techniques
which according to Kelly (1969:120) were “in constant use in the
language classroom right through the history of language teaching” but
which recently have fallen into disuse because they appeared to be
‘contrived’ and inappropriate will have to be reappraised in the light of
the new genre being created. An example of a teaching technique that
has fallen into disuse is written dialogue, which was a common tool for
the presentation and practice of language in the structural model of
language teaching. Modern e-mail seems to be more like this, that is, it is
more like written dialogue than formal letter writing. Written dialogue
was challenged as being inauthentic in the 1970’s and 80’s and,
therefore, not suitable as a model of actual usage but, through e-mail, it
now takes on renewed significance (Leech 1997). There is also to be found
on the Internet written lectures which reflect a little of both worlds, being
written text which is meant to be spoken and which therefore contains
comments and asides that have a specific listener in mind (Stubbs 1996,
McCarthy and Carter 1994). The manner in which students interact with
technology is the object of much research and often the results are
disappointing to those who believed that the technological revolution
would revolutionise teaching6. Most CALL specialists reached the
conclusion that computers are an aid to teaching which, rather like other
modern technologies such as the video, depend upon the ingenuity of the

6
Seymour Papert (1980) Mindstorms: Children, Computers and Powerful Ideas, Harvester Press is an
example of this view of the huge changes (and improvements) that the technological revolution would
bring to education. Robin Goodfellow of the Open University reports (1999) Language Learners’ I.T.
Strategies will they be the Death of CALL? that with university language learners in an open-access IT
environment, CALL is “vulnerable to the growth of IT sophistication in learners” . So he recommends
that teachers need to turn their attention to “the IT choices that learners make when they embark on self-
study” otherwise carefully prepared CALL designs will be sidelined.
24
teacher to make them relevant and useful resources for students
(Kenning and Kenning 1990, Higgins 1988, Phillips 1985, Leech and
Candlin 1986). In 1987 Eastment predicted that computers would not be
found in computer rooms except for computer literacy courses but would
be located in normal classrooms as part of everyday teaching. He also
suggested that at the time of writing (1987:10) concordancing was limited
by the rather unreliable software that was available. This problem has
now been largely overcome but Eastment’s prediction that we would have
“pedgagogical concordances” of varying levels and language types has still
to be realised. Warschauer (1999) argues that in education there is no
BALL (book-assisted language learning), PALL (pen-assisted language
learning) and no LALL (library-assisted language learning) because these
are such powerful technologies. Therefore, he argues, it is only with the
integration of computer technology into teacher education and language
learning that computers could be seen to have taken their place as a
natural and powerful part of the language learning process. The change
towards data-driven learning (DDL) is one of the more exciting new
trends which will be discussed later in relation to the use of corpora for
teaching purposes (Chapter 7).
The use of computers has also had an impact on linguistics and
the description of language through corpus studies and these discoveries
must be exploited and integrated into the curriculum (see 2.5 The Corpus
Analysis Approach). Particularly in the area of collocations, new
information is more easily obtained and is available for use by the
teacher and the learner. Research work on language acquisition has also
suggested that ‘chunks’ of language are used in natural language
acquisition (Hakuta 1974; Huang 1971; Brown 1973; Clark 1974;
Cruttenden 1981; Wong-Fillmore 1976; Newmark 1979, Peters 1983) and
so the use of materials derived from concordancing the target student
texts will provide one more tool to be added to the repertoire of teaching.

25
It is my contention that collocations are appropriate ‘chunks’ of language
that can and should be used use to teach to language learners (Tribble
and Jones 1990). The collocations can be obtained from the corpora
compiled from the textbooks on the bibliographies which the students are
meant to consult. In this way, the language that is being studied becomes
entirely appropriate for the purposes of the students. The specific lexical
semantics of science and technology is being presented rather than
general English. Furthermore, the language being studied could be
brought under the control of the student, thereby customising the
learners materials for study. These aspects will be taken up in more
detail later in 7.4 Data-Driven Learning.
In Portugal, and in the University of Aveiro in particular, there are
now many homepages and interactive websites which the undergraduates
are encouraged to consult and even study from. Most higher education
institutes, like the University of Aveiro, have home pages for each of the
departments. The University has its Informatics Centre through which
students can gain access to the Internet, to say nothing of the facilities
the students have at home or contrive for themselves. There are also
Open University and distance learning courses for undergraduates
making use of the Internet. The first-year students therefore soon
become, if they are not already, quite sophisticated in their knowledge,
use and expectations of modern technology.

1.4 The Dominance of English in Science and Technology

With the economic supremacy of America in the world now and for
most of the 20th century, the English language has also come to
dominate the world of science and technology. Most research work is now
published in English no matter where the research was carried out.

26
Kaplan (1993:156) claims that “something on the order of 85% of all the
scientific and technical information available in the world today is either
originally written in, or abstracted in English.” Furthermore, many of the
books used to teach science and technology are based on American
models.
The significance of this is that in most cases the English that
students encounter and the English that students therefore need will be
predominantly American and it will also be language that is not
specifically prepared for the student of English as a foreign language. It
will, however, be predominantly written language.7 Despite the fact that
all European languages are supposed to be equally important in the
European Union some are seen to be more prevalent than others. Sheer
numbers of speakers have an obvious impact upon this so that the
Portuguese language is not one of the languages that the scientific
community sees as essential for the people who will run the businesses of
tomorrow in Europe. The European Commission 1997 Eurobarometer
reported the results of a survey conducted in 34 countries in Western,
Central and Eastern Europe in which Russian was the principal language
of 35% of the 555M people in these countries, English 28%, German
20%, French 17% and Italian 10% and suggests that the languages at the
upper end of this spectrum, that is Russian and English, appear to be
spreading whilst those at the lower end are declining. Crystal (1997:10)
puts forward the financial argument for using a lingua franca in
international bodies which is that the cost of translation can swallow up
to half the budget for such organisations. The European Union has yet to
come to terms with this problem.

7
Research carried out by Prof. Drª Ana Margarida Barros of the Department of Chemistry, University of
Aveiro, published in her (1998) report on the European Chemistry Thematic Network (ECTN) work on
Communication and Management Skills shows that reading and analysing texts in a foreign language
(usually English but possibly in French) is considered indispensable by 100% of those answering her
questionnaire from Universities in Portugal, and that this activity was classified as indispensable by 88%
of the Industries consulted and as very important by the other 12%.
27
 Whilst it can be argued that Brazil, Mozambique and Angola are
very important markets outside Europe, the problems that these
countries are experiencing means that this potential may not be realised
for some time to come, if at all, and that, therefore, the Portuguese
language will not be seen to be as important at the moment as it might be
in the future. Crystal (1997:7) argues that a language becomes a global
language because it is the language of power, both political and military,
which explains why Portuguese found its way into the Americas, Africa
and the Far East during the period of colonisation. However, with
Mozambique joining the Commonwealth countries there is a suggestion
that it feels drawn more towards countries that had a connection with
Britain and the English language. The proximity of South Africa,
Zimbabwe, Zambia, Malawi and Tanzania where English is either an
official language or retains some influence may also help to explain this.
Crystal (1997:61) suggests that whether English becomes a global
language in the twenty-first century depends upon what happens in
countries with the largest populations, notably China, Japan, Russia,
Indonesia and Brazil. University students who will be the leaders of
tomorrow will need to learn at least one of the dominant languages. From
the numbers of students studying English in Portuguese secondary
schools it can be seen that the language that is often being chosen is
English (Ferreira, Ramos and Braga da Silva 1999).
A significant factor in the dominance of the English language is the
overwhelming expansion in the use of computers in the world. Kubanek
(1998:202) points out that “the lingua franca function of English would
become obvious” to students using the Internet and contacting websites
for information. This technological revolution is having an enormous
effect on education and employment.

28
1.5 The Situation in Portugal

Given the emphasis on the study of science and technology in the

world, there is also a need to study the English of Science and
Technology in order to assist students to study language of science and
technology. However, secondary school English courses continue to teach
the English of the Humanities which ill-prepares the students for tertiary
education in science and technology and the variety of English that this
represents.
The changes that have taken place in educational objectives
because of, and deriving from, the European Union have emphasised
multicultural education. The intention is to prepare students for
European citizenship. This requires both understanding and tolerance of
other cultures together with more language training so that all citizens
will have knowledge of a minimum of three European languages. School
programmes in Portugal reflect this preoccupation with cultural identity
and acceptance of other cultures. The programmes for schools and the
manuals used also reflect this state of affairs. Culture and language are
seen to be mixed together in such a way that it is impossible for them to
be separated (Fligelstone 1998), however, the reader or listener may be
unaware of differences between cultures and therefore their
understanding of what is being transmitted may be faulty (Scollon and
Scollon 1995). Those students who have chosen to follow sciences are
given the same programmes in English as students who wish to follow
humanities courses, rather than special programmes to help them to
cope with the enormous amount of scientific and technical information
which is published in English and which they will certainly meet if they
continue their studies to university level or in their future employment.
This is not to say that the English they have learnt will not stand them in
good stead in many general situations but it will not prepare them

29
adequately for their future studies. Indeed it may even lead to confusion
and misunderstanding in their scientific and technological studies.
Whilst it cannot be denied that this preoccupation with
“multiculturalism” is important in the context of liberal education for
citizenship, it is less useful for the needs of the science and technology
students who require both this ‘liberal’ education and more or further
language support with their specific English needs.
Furthermore, the teachers, who form the bulk of EFL teachers in
Portuguese secondary education and are required to teach language to
students in schools, have followed typical humanities education courses
themselves. Several ESP theorists (for example, Widdowson 1979, Ewer
1975, Strevens 1978, Hutchinson and Waters 1987 and Kennedy 1983)
have pointed out the fact that those who are required to teach the
language of science and technology feel they are themselves ill-prepared,
and are therefore often reluctant to do so. Even in terms of technology,
teachers with a humanities background were seen to have an antipathy
to “machines”. However, most undergraduates these days in this and
other modern universities are positively encouraged to confront the latter
problem through educational technology disciplines on their courses and
by being expected to submit word processed assignments for other
disciplines. The use of computers in schools however for the most part
continues to be considered the province of the maths department (Moura
Carvalho 1991, Stubbs 1992). There is awareness of a need to change
this state of affairs but as White (1988) highlights it can be difficult to
achieve innovation in schools and the process takes a long time.

“The discrimination and adoption of an innovation – in language teaching

as elsewhere – follows and S-shaped curve (...). There is an early stage
during which a very small percentage of innovators decide to introduce the
new idea. This is followed by a second stage during which the early

30
 adopters, who have noted that the innovation produces no harmful effects,
take on the innovation. During the middle stage, the majority adopt
quickly, influenced mainly by the innovators. At a late stage, the laggards
or late adopters finally give in. A minority who never adopt lie outside the
curve.”

In addition to this, for innovation to arise, be taken up and successfully

installed White believes that effective management must exist in the
organisation.
The University of Aveiro is aware of the necessity for English
language proficiency for science and technology students and has
introduced innovations into the structure of the science and engineering
courses in order to address the students needs. The identification of core
disciplines which almost all of the undergraduates in science and
engineering have to study is the main one of these. Other universities,
such as Porto and Coimbra8 continue to teach ESP within the different
faculties rather than as a separate or common core subject for all of the
science and technology students. This system was an earlier model used
for most ESP courses, known as ‘content-based’ syllabi focusing on the
particular requirements of specific academic disciplines. This explains
why published materials are usually entitled English for Electronics,
English for Telecommunications or English for Computer Science and so on.
Very few textbooks attempt to address the English of Science and
Technology today. Previous coursebooks have attempted to isolate some
form of sub-technical language which was thought to underpin the
language of science and technology as a whole.
There are obvious advantages to the type of course which has
common core subjects not least of which is the fact that it allows efficient

8
ESP also exists of course in institutes such as ISCAA here in Aveiro where there is English for
Accountancy.
31
use of staff, providing staff to student ratios which can cope with the
huge entry to university which is taking place in most developed
countries as mentioned earlier. Another advantage of this system lies in
what Laurillard (1993) describes as the need for undergraduates to
develop concepts rather than merely gather facts. The undergraduates
need to learn how to learn autonomously and need to be guided to that
end. The individual scientific concepts that can be found in any specific
science subject nevertheless present some differences from what students
have been taught before. The students are often unaware that this is the
case in university and this in itself can lead to a lack of success and
indeed to considerable frustration if students regarded themselves as
good students and now at university they suddenly begin to get
unexpectedly low marks. In English language studies the students will
also have to undergo this transformation and recognise that what they
have learnt before is only part of the story and what may appear on the
surface to be the same may in fact be quite different in this new context.
The fact that the discipline has to be directed more generally to science
and technology can therefore be an advantage because the students can
become aware that the skills they need to acquire will stand them in good
stead no matter what their subject speciality is. Content knowledge will
also be acquired by the undergraduates along the course so that all of the
students have a similar lack of specific subject knowledge on entering the
university and can benefit from adopting certain strategies when faced
with new material, especially if this new material is in English.
The study which is presented here is focused on University of
Aveiro students and courses as a sample of the language needs and
teaching requirements of undergraduate university students matriculated
in a number of different courses preparing them for the future.

32
1.6 Science and Technology Undergraduates and English

The University of Aveiro introduced a foundation year for all

students taking the various engineering courses of the University in
1993. Five disciplines make up the core subjects taught to most of the
new students (English, Chemistry, Physics, Mathematics and Computer
Science), but there are some exceptions such as the Licenciatura em
Novas Tecnologias da Comunicação (NTC) which only has the English
and Computer Science components, the Licenciatura em Gestão e
Planeamento em Turismo (GPT) which has no chemistry or physics, the
Licenciatura em Planeamento Regional e Urbano (PRU) which has no
physics. This innovation has meant that the English discipline has had to
be revised.
Although English had been taught for many years to many of the
engineering students, the foundation year or Common Year (Ano Comum)
as it is known, has led to a considerable change. On the one hand, there
has been a loss of specificity, that is the specific English genre for the
course the students were studying, for example, English for Electronics
and Telecommunications. On the other hand, there is a need for a wider
but nonetheless specific English genre to be taught to these science and
technology students, the English of Science and Technology. This is in
order to help them to cope both in the first year and in subsequent years
with the English they will need for the many and varied courses and
bibliographies containing books in English that they are going to come
into contact with. The bibliographies for core subjects contain many
books in English or originally written in English and translated (up to
75% in some cases). Although bibliographies are liable to change and
change quite quickly from year to year, some core texts in English remain
for a number of years and are available in the library for consultation by
the first year students. Therefore, the English the students have to cope

33
with needs to be defined in order to be able to produce a syllabus which
makes the optimum use of the limited time9 and resources available to
this annual first year course.
The immediate short-term needs for these students can be
identified through the bibliographies they are asked to consult in their
first year science and technology courses. These include a number of
books in English on the core science subjects taught in the first year of
the University. As scientific literature is seen as becoming more and more
incomprehensible in the latter half of this century for all but a few
specialists (Hayes 1992:739-740), undergraduates will need help in
reading and understanding scientific texts.
The kind of language the students require to be able to read these
books successfully can be identified by detailed study of the textbooks.
The physics and chemistry textbooks have been studied in order to
identify their needs in respect of the syllabus and they will be presented
here. However, as the students come from over 25 different courses, there
is a need to provide a baseline corpus for comparison for the
comprehensive syllabus to be drawn up. In order to recognise what is
normal use in a particular genre a very large corpus has to be consulted,
the baseline corpus, in order to avoid generalising from what may be an
abnormal or exceptional example of language use found in one or a small
number of texts. As the study of these textbooks is based on variation
studies in order to see how far they differ from other genres or text-types,
some form of comparison needs to be made to highlight the differences
and to add scope to the syllabus. As was mentioned above, the science
and technology courses cover a much wider field than the books on the
bibliography alone can represent. What would be most appropriate, given

9
The discipline has one two-hour class per week across the two terms of the first year for most of the
students. Exceptions to this are the Licenciatura em Novas Tecnologias da Comunicação (NTC) which
has a term of four contact hours of English per week in the second year and the Licenciatura em Gestão e
Planeamento em Turismo (GPT) which has 5 hours of English per week for the first term in the second
year.
34
that the students’ physics and chemistry textbooks on their
bibliographies are overwhelmingly American publications, would be an
American general science textbook aimed at undergraduates. One
textbook would nevertheless not fulfil the criteria of a baseline corpus as
it would itself be liable to offer an exceptional or aberrant style for the
genre so a number of American general science textbooks would be
needed to analyse the genre. As such a number of suitable textbooks
could not be identified for this type of tertiary level student, a multimedia
encyclopaedia will be used. The advantages of such material are its wide-
range and huge size to meet the demands of generality in order to identify
linguistic trends and tendencies in the genre. The range would be more
than adequate to cover the basic science of all of the courses included in
the first year foundation course and the size runs to hundreds of millions
of words rather than the tens of thousands of words to be found in one
general science textbook. This will be taken up in more detail later in
1.10 CD-ROM Material.
Implicit in deciding what to include in the syllabus is what English
the students have already learnt or already know. In order to answer this
question the students were tested and their results analysed in order to
identify areas which need to be addressed by the syllabus. Chapter 4 Test
Results for New Students discusses the test used in each of the academic
years from 1993-1998 and the results found for new undergraduates in
those years.

1.7 The Ano Comum

As there are approximately twenty five different courses being

catered for in the Ano Comum, the first-year foundation course, each and
every one of the different specialisations will be addressed nevertheless
through recourse to a multimedia encyclopaedia. As mentioned above, a
35
CD-ROM encyclopaedia covers all of the undergraduates subject
specialities and so it is argued that generalisations about English for
Science and Technology can legitimately be made because of its huge size
and comprehensive nature. The study will also make a detailed analysis
of published texts contained in the students’ bibliographies for the first
year for Chemistry and Physics10. Despite the fact that the language
needs of the students in truth go beyond the boundaries of the first year
of undergraduate studies, other language needs will not be addressed per
se. The language appropriate for conferences and post-graduate work for
example will not be considered, it is too remote and lofty an objective for
an English discipline syllabus designed for the first year of
undergraduate studies with the considerable time constraints imposed
upon it. The use of strategies for coping with texts will certainly prove
useful however in the later stages of the students’ courses. The
methodology used with the students and skills required by same will
therefore also be of importance in enabling them to cope in the remainder
of their courses with the English that they will meet.
It must be acknowledged that many of the students who take the
science and engineering courses in the university will often become
teachers or take up jobs in business and management and only a rare
few will become pure or research scientists11. For this reason, a concern
for undergraduate work and success within university courses in science
and technology would appear to be not only a more realistic aim of this
research but would be addressing a common need of the students in
university science and engineering courses. Moreover, the content should

10
It is admitted that English for Information Science is as relevant as the English contained in textbooks on
the bibliographies for Mathematics. However, the language in Information Science is English even if the
explanations for use (and pronunciation of the terminology) is given in Portuguese. There are also
appropriate glossaries that students can consult for this discipline. The language of Mathematics is
subsumed by the mathematics contained in the physics textbook analysed and has been shown to be a very
restricted genre (Biber 1988).
11
See Arroteia, Jorge Carvalho; Martins, António Maria (1997) Inserção Profissional do Diplomados pela
Universidade de Aveiro: Trajectórias Academicas e Profissionais, Aveiro: Universidade de Aveiro.

36
be made appropriate to the learning purpose of the students. Wilson
(1997:130) suggests that databases designed for use with language
students should contain texts that relate to students’ tasks and interests
in other disciplines in order to make the “students’ goals in the language
learning programme … coincide as far as possible with the students’
wider goals.” Despite this, she identifies the fact that her computer-based
materials were too general as a “disappointment” as they “had none of
the quality control for style and linguistic coverage that good CALL
demands”. In other words, Wilson reminds us that the materials that are
used with undergraduates need to be carefully selected so that they are
sophisticated enough and they must be tried out and improved upon or
abandoned if necessary should they prove to be unsuitable.

1.8 Appropriate Text-Types

Similarly, the materials used for analysis need to be carefully

selected. Despite the fact that many studies have been carried out on
specific English in the past, they often include a wide range of text-types,
including text-types12 that would be appropriate for post-graduate
specialists. One early example of this is Barber 1962. Barber took three
texts that ‘straddled’ disciplines, two came from university textbooks on
engineering applications of electronics and astronomy and one from a
journal on biochemistry. The latter text-type would be suitable only for
those (few) students who enter university to study science and technology
and who continue with their studies to a very high level. Similarly,
Tarone et al (1981) studied two papers from one Astrophysics journal. In
this case it is too small a study to allow generalisations to be made.
Indeed Swales (1985:192) comments that this work by Tarone et al is an

12
text here should be understood to include written and spoken language.
37
“inadequate sample”. Even Tarone et al (1981:191) indict themselves they
say

While extensive use of the passive is shown by frequency counts of verb

tense and aspect which are performed on corpora combining texts from a
variety of scientific and technical fields, significantly different results may
be obtained when one compares the frequency of the passive and active
voices within a single scientific or technical field.

The only study that they report they have found that was on only one
field was Wingard’s (1981) work on medical texts. There has been an
increase in recent years in the numbers of students going on with post-
graduate studies, and it may well be that at a later stage those advanced
students’ specific language needs must be studied to see if further
language training is necessary.
This pattern of combining texts-types often including journalese or
popular science texts has continued in many cases even in some corpora
that are regularly used by researchers. Therefore, there is no suitable
study of undergraduate textbooks for science and technology students
that could usefully be used to identify the target language of these
undergraduates and hence the necessity to start from the beginning to
analyse specific texts appropriate for these students.

1.9 The Corpora

Added to the problem of defining exactly which text-type might be

appropriate is the difficulty of obtaining a sufficiently large corpus in
order to produce significant results. Some of the studies done in the past
were based on very small corpora or one very small corpus which now
would be considered dubious as a basis from which to draw valid

38
conclusions (for example, as mentioned earlier in 1.8 Appropriate Text-
Types, Tarone et al 1981 only used two texts). However, more detailed
studies of a smaller corpus may show features that would be lost in a
very large corpus (Robinson 1991). This will be taken up in more detail
later in 2.2.3 Scientific Specificity. For this reason two sub-corpora are
included in the analysis. Five corpora will be used: a large physics
corpus, a small physics sub-corpus, a large chemistry corpus, a small
chemistry sub-corpus and a corpus from a multimedia encyclopaedia for
strictly comparative purposes. Biber, Conrad and Reppen (1998:136) go
even further and suggest that studies based on very small corpora are
likely to be inaccurate and a ‘baseline’ is needed for comparison to
identify significant variation. Halliday (1993) suggests that the
development of the modern corpus is that “we can now for the first time
undertake serious quantitative work in the field of grammar” but he
points out that in order to be able to do this “Quantitative studies require
very large populations to work with.” The multimedia encyclopaedia will
provide that baseline for comparison as will the large physics and
chemistry corpora when used in comparison with the sub-corpora in
these same subject areas.

1.10 CD-ROM Material

CD-ROM material in the form of a CD-ROM encyclopaedia

provides an extensive baseline corpus from which to draw generalisations
about linguistic phenomena in the study undertaken. It also provides a
wide enough range in terms of subject matter to address all of the
varieties of courses included in the foundation year for undergraduates.
Although it will be argued that a multimedia encyclopaedia on CD-ROM
is of a less sophisticated nature than a general science textbook, this is
denied by a number of researchers (Huddlestone 1971, Swales 1985,

39
Halliday and Martin 1993). These linguists argue that encyclopaedia texts
are intended to be instructional and so are textbooks. Furthermore,
textbooks and encyclopaedia are seen to be of a similar level or standard.
They are also aimed at a similar reader, that is one who has knowledge of
the subject but is not a specialist and is in the process of learning more.
Although encyclopaedias may of course be used for more general
purposes, more in-depth information can be obtained if the user so
desires. The multimedia encyclopaedia chosen provides reading lists for
further study on any topic.
The CD-ROM encyclopaedia also shares a number of features that
are particularly relevant for our tertiary level students. First of all almost
all of the widely published CD-ROM multimedia encyclopaedias are in
American English. This is partly as a result of Microsoft’s dominance in
the computer market as mentioned earlier and their marketing strategy
of linking other products to the sale of their personal computers.

Secondly the length of the texts reflects new technology where

‘screenfuls’ of information have to be coped with. Search facilities are
provided. Educational policy in Europe suggests that being able to handle
searches and obtain information from new technology are skills students
need to have in order to work successfully with such important
applications as the Internet.
Third, an added dimension of “range”, that is, how many texts the
items are used in, can be analysed from the CD-ROM so that the usual
context of lexical items can be examined. A wide range means that the
item is more useful as it can be used in more situations. This is
consistent with Michael West’s (1953) idea of a high ‘surrender value’ for
student learning or Swales’ (1985) “maximum educational value”. The
aspect of context is also relevant for syllabus design where an
appropriate context for lexical items is essential in order to make the
teaching materials designed for use on the syllabus reflect authentic or
40
natural contexts of use in the genres of science and technology as
opposed to in general English genres. Robinson (1991:20), discussing the
difficulty in delimiting different forms of English as every situation
overlaps with another, prefers the use of the term ‘technolect’ after
Lauren and Nordman (1986) because it suggests a form of language
rather than an independent language as opposed to the term ‘special
language’ to describe the difference between general English and this
form of English.
Multimedia technology applications are seen as a possible future
for study for students at this level in particular. All the developed
countries are keen to foster the use of new technology in education and
the University of Aveiro is committed to this objective too. The European
Union is involved in many projects in this area, including going as far as
the development of virtual universities like that of the Universitat Oberta
de Catalunya, Spain13. Furthermore, advances in language teaching
methodologies also suggest incorporating technology in learning
strategies. Computer driven searches of appropriate corpora which may
be from CD-ROMs are increasingly being used to present to learners
precise, authentic language use in specific genres where learning is seen
to be data and learner driven. (cf. Wichmann, A., Fligelstone, S.,
McEnery, T., Knowles, G. (eds. 1997) Teaching and Language Corpora,
London: Longman.)
Much documentation supports the spread of these technologies.
The European Round Table of Industrialists Report (1997:6) says that
“the number of CD-ROM drives rose from 2.7 million to 9 million between
1994 and 1995, and is expected to reach 35 million units by 1998.” This
report goes on to suggest that “Not integrating ICT (Information and
Communication Technology) in the education process would further

13
This is an example of an Open University for distance learning programmes which makes use of computer
technology only.
41
widen the gap between real life and education. Youngsters are growing up
in an informatics and media world: education should respond to their
cultural expectation pattern, use their language.” The CD-ROM
encyclopaedia fits this role, but maintains an educational rather than an
entertainment perspective. Young people all too often use games on CD-
ROM as their “language” and although motivation is extremely important
in teaching and learning, this thesis argues that education at tertiary
level should be both stimulating and demanding.
The number of CD-ROM encyclopaedias has increased in recent
years but in the early 1990s there were only a few widely available ones
such as the Grolier, Compton’s and Encarta by Microsoft. Compton’s was
not very user friendly while the Encarta tended to take on a more
entertainment type of format including quizzes and games. For these
reasons the Grolier was chosen as it combines a suitably academic style
with user-friendliness. The report written by Jeremy Fox, Anne Matthews,
Clive Matthews and Arthur Rope for the British Government Employment
Department Group Training Agency Learning Technology Unit by the
University of East Anglia and the Bell Educational Trust, March 1990
Educational Technology in Modern Language Learning in the secondary,
tertiary and vocational sectors, describes the GROLIER ELECTRONIC
ENCYCLOPEDIA which will be used here (ibid.1990:26) as “an excellent
example” of a CD-ROM encyclopaedia which “holds the equivalent of 20
bookshelf volumes plus an index of all the occurrences of every word in
the encyclopaedia.” and the report grades this encyclopaedia for
“secondary, tertiary and vocational” levels, with emphasis on the latter
two which they indicate by means of the italics used. Furthermore, the
report claims that the encyclopaedia is applicable to the areas of reading,
writing and vocabulary and can be used “in a hypertext-like way down a
track of cross-references”.

42
 The appropriateness of hypertext in teaching Portuguese students
of English has been explored by Prof. Doctor António Moreira in his
doctoral thesis Desenvolvimento da flexibilidade cognitiva dos alunos-
futuros-professores: uma experiência em Didáctica do Inglês (1996
University of Aveiro). He finds cognitive learning of this type to be
successful with students in Portugal. He found that (1996:x) “hypertext
systems based on an approach that uses cases which are structured in
such a way that they offer multiple representations of knowledge which
in turn emphasise critical interconnections between different structural
and surface knowledge components can be superior in their effectiveness
for the preparation of students in their use of knowledge in new and in
novel situations”. This form of transfer is extremely important for the
students under study here who are attempting to use English in a ‘new
and novel’ situation - that of tertiary level study in science and
technology.

1.11 The Syllabus

This study will attempt to examine and define just what the
appropriate specific English for these undergraduate science and
technology students is. This will be found through a linguistic analysis of
computer corpora, from the textbooks for physics and chemistry found
on the students’ bibliographies, contrasted with an analysis of the
students’ language needs as obtained from the results of the tests
described later in Chapter 4. The areas that must be addressed in
undergraduate English language studies will then be identified. The
results of this research are to be applied to the development of a syllabus
and teaching materials for the discipline appropriate for the entry
standard of English of the students taking the discipline and for their

43
overall course needs in terms of bibliography in English for science and
technology.
Many other considerations will have to be taken into account as
well such as the amount of contact hours available, the size of classes
and the heterogeneity of the students in those classes. All of these
features of the discipline will influence the syllabus that can be used with
these undergraduates. The fact that these are undergraduates just
starting their courses in university will also have to be taken into
consideration as mentioned earlier as they are going to have to adapt to
many new aspects of life as well as new aspects of learning in an entirely
different environment from the one the have been used to up to this
point. Simply adapting to the size and complexity of university life is a
major difficulty for many of the new students who may also be coping
with being away from home and family for the first time as well. Students
cannot be seen divorced from these different aspects of their lives which
will colour their learning and attitude to learning and which the teacher
and syllabus designer have to take into consideration in their work. The
syllabus then needs to address the state the learner is in at the beginning
of the course not only from the point of view of their level of knowledge,
which will vary from student to student, but also from their personal
situation with regard to university life. There will be a need to draw
together a number of strands to blend the classes into some form of co-
operative body where the differences of level in background knowledge,
both of their subject specialities and language level, together with other
more mundane problems of their new lifestyle will be addressed. Simply
getting the students into contact with each other and making friends is
important for the well-being of undergraduates and their success on their
courses (Tavares, Santiago, Lencestre, Soares 1996).

44
1.12 The Research

The principles of orientation underlying this research in order to

produce valid materials for Science and Technology students in their
undergraduate years at university must therefore take into account the
following criteria:
1. The English must be the American English found in those academic
textbooks in the students’ bibliographies, which were written to be used
to teach native speaker students at undergraduate level. Biber (1988:201)
finds there are systematic differences in British and American written
texts in that “American written genres are consistently more colloquial
and involved than British written genres, while at the same time
American written genres are consistently more nominal and jargony than
British genres.” This is a general observation of differences between
British and American English. However, the specific differences between
genres with respect to the undergraduate science textbooks has to be
studied.
2. The fact that there has been a technological revolution, which has
affected both teaching and learning, must also be considered. The latter
is reflected in this study through the use of source material taken from
multimedia encyclopaedia which will serve as comparison (variation
study) with the textbook corpora and provides the necessary scope of
information for science and technology students studying on a wide
variety of science and engineering courses.
3. The corpora must be compared with other studies to identify how far
they vary from the results obtained for other science and technology
texts. Biber’s (1988) algorithms on academic prose will be used for this
purpose.
4. Finally, the results obtained must be compared with the language
the students have already acquired or learnt, which will be established

45
through testing, in order to identify mismatches with the English needed
by those students coming into the University of Aveiro to take up places
on Science and Technology courses.

1.13 Methodology

The research undertaken follows this pattern: firstly frequency and

range studies of the multimedia encyclopaedia are made and then these
are compared and contrasted with scientific text-types taken from the
actual coursebooks used in Aveiro University for science and technology
students in the first year. Frequency lists are used as one of the bases of
data on variation from which descriptions of language and therefore
decisions about appropriate language can be drawn for application in the
materials taught in this discipline.
Comparison of textual features are also made following Biber’s
(1988) variation studies methodology to provide a detailed scientific
comparison for the research. As some linguists have argued (cf. Roberts,
1983) the definition of an area is complicated as there is often an overlap
between disciplines and between text types, which makes it difficult to
describe the features which uniquely pertain to that discipline. Variation
studies overcome this complication by looking at how far the texts under
study differ from other texts rather than at an absolute contrast between
them.
The level of the English the students coming in to the first year in
the university have acquired is examined through tests carried out over
five years of student intakes, together with the English needs identified
by the different university departments for their university students. The
latest research on language acquisition and corpus linguistics will be
applied in order to determine what is relevant for the teaching or learning

46
of such students. The role of modern technology in education is also
addressed in the learning strategies proposed for these students.

47
Chapter 2 Historical and Theoretical
Background to ESP
 Chapter 2

Historical and Theoretical Background to ESP

The difference between general English and specific varieties of

English is a very problematic area as distinctions can be drawn either
between the differences in what language is used for or the differences in
the language used. Despite this, many kinds of analyses have been
carried out over the last few decades. The argument has come full circle
and often now revolves around what general English consists of as
opposed to what specific language is in either literature or science.
Recently there have been new attempts to define this general English (see
later 2.5.7 The British National Corpus). The reason for the attempt to
define specific varieties of English was often in order to apply the findings
to syllabi for teaching purposes to provide the students of those courses
with tailor-made material and information about the language they
needed in their studies of a particular subject The research that has been
carried out has had a much more widespread effect than this however.
Mainstream English as a Foreign Language (EFL) teaching has benefited
greatly from the results of linguistic thought and analysis carried out
ostensibly for ESP. Before it is possible to move forward in research for
ESP it is necessary to review what has gone before to find what can
usefully be done to try to add to the store of knowledge that has been
built up. It is also important to examine how the work done in the past
has affected syllabi and the teaching of English to science and technology
students. In other words a synthesis of the research and ideas that have

49
gone before can be of enormous help in defining what should be included
in a syllabus for university students of science and technology.
A methodology for the teaching of language can be traced right
back to Quintilian (Marcus Fabius Quintilianus. 35 - 95 A.D.) with his
Instituto Oratoria. He outlines the teaching of rhetoric or bene dicendi
scientia as being made up of the study of grammar which is sub-divided
into correct expression or recte loquendi scientia and interpretation of the
poets or poetarum enarratio which, in turn, requires the study of writing
and reading or scribendi legendique facultas. Quintilian was aiming to
produce the perfect orator through his system of linguistic studies and
states that the first requirement of an orator is that “he should be a good
man”. Quintilian’s methodology was that a second (or foreign) language
should be taught to children through total immersion in the target
language, although he also advocated adapting materials to suit different
types of learners and of motivating students to learn. This idea of
different types of learner requiring different types of materials is
fundamental to the modern study of languages for special purposes.
Similar to Quintilian’s ideas on motivation is the modern idea that
motivation is a necessary prerequisite to facilitate learning which is
advocated by those involved in special language training today (cf.
Hutchinson and Waters 1987).
From the seventeenth to the eighteenth and nineteenth centuries,
from Locke to Horne Tooke and Humbolt theories about language led to
etymological studies and then descriptions of languages being made. The
emergence of a method of ‘scientific’ study of language based on empirical
research continued into the twentieth century with work such as that of
Bloomfield on indigenous American Indian languages. All of these strands
of theoretical linguistics have had and still are having effects on syllabus
design, materials and the teaching and learning of special languages like
that of science and technology under study here.

50
 2.1 English for Special Purposes

The historical background to English for Special (or Specific as it

was known in the 1960s) Purposes (ESP) and in particular, to English for
Science and Technology (EST)1 has neither the Quintilian requirement of
‘goodness’ nor the possibility of using his methodological approach. This
field was developed for working adults, not children, and often for those
in tertiary education who were studying science and engineering and for
whom English was a foreign language. Locke had already, in the
seventeenth century, shown concern about language as a(n) (imperfect)
vehicle for the acquisition and spread of knowledge. However, he
proposed remedies for these imperfections of language so that language
could be used safely for the purposes of science and philosophy. Locke’s
remedies consisted of the definition of complex ideas through the use of
simple ones so that communication between people was possible and a
bridge between minds could be built. This process Locke believed to be
similar to demonstrating a mathematical conclusion and that thereby
this use of language would become adequate for scientific and
philosophical discourse. As a result of Locke’s ideas, there was great
interest in etymological studies and many dictionaries were subsequently
written in the eighteenth century. This process cannot be seen to have
come to an end and research into language use for the writing of
dictionaries (see 2.5 the Corpus Analysis Approach) continues today but
modern technology is now used to improve the accuracy of the
information included. Swift also argued for an Academy to ‘fix’ the
language for scientific purposes. In other words, historically there has
been a tendency to invent a language of science which is removed from

1
The use of the term English for Science and Technology (EST) is usually attributed to Ewer (1971).,
although Trimble (1985) attributes it to Selinker.

51
general language. This was particularly the case when Latin stopped
being the lingua franca of scientific thought. The idea of language as fixed
is contrary to fact whether for scientific purposes or any others (White
1998 and see 5.1.5 Plurals from Latin and Greek) which suggests that
there will always be a need to analyse its use both diachronically and
synchronically.
Theories about language have not stopped being put forward either
and language studies for application in teaching have added more to the
understanding of specific varieties of English. For example Swales
(1985:x) sees EST as underpinning the development of ESP. He says that
“With one or two exceptions …English for Science and Technology has
always set and continues to set the trend in theoretical discussion, in
ways of analysing language, and in the variety of actual teaching
materials.”

2.1.1 Phrasebooks

Special purpose language has a very long history with the

‘traveller’s language course’ for those who intended to visit a foreign
country and who wished to obtain a smattering of the language of the
phrase book type. Course books aimed at this type of learner have been
in existence since the sixteenth century according to Archibald Lyall
(1932) in his Guide to the Languages of Europe: a practical phrase book
where he quotes from ‘Colloques ou dialogues avec un dictionnaire en six
langues of my earliest predecessor, Henry Heyndricx, Antwerp, 1576’
(Strevens 1978). Strevens (1978:190) claims that phrase-books for foreign
tourists have been in existence for four hundred years and Opitz (1983)
that mariners have been using specialised bilingual maritime dictionaries
for more than two hundred years. This type of phrasebook language is an
example of the approach which emphasises the objective or outcome of

52
language use or ‘specific purpose’ as opposed to ‘special language’
(Turner 1981) and is based upon the idea that there is an equivalent in
one language for an item found in another language. This approach takes
the view that one language corresponds directly to another language
although it is in code and ignores the idea that there are cultural
differences between languages which need to be coped with.
The implications for language teaching for foreign travel and
phrasebook language has been significant in that different emphasis was
placed in teaching on speaking and listening, although traditional
teaching methods in the past would have favoured pronunciation and
reading aloud of phrases. The results of theories about the reasons for
teaching English has changed the methodologies and materials used for
that teaching. The final example given above for mariners would be
entitled EOP, English for Occupational Purposes, today and might well
restrict itself to very elementary goals. Similarly, just as new dictionaries
continue to be produced, the study of English for Special Purposes
continues to this day and there have been four, often overlapping and
interconnecting major schools of thought for the teaching of ESP and
science and technology this century. These can be described as: the
register analysis approach, the discourse analysis and variation studies
approach, the needs analysis approach and, most recently, the corpus
analysis approach2 which I shall go on to describe in order to show how
they influence the study of science and technology for university students
today.
Some of these approaches derived from the need to respond to a
practical crisis like the need during the Second World War for a means of
teaching/learning foreign languages quickly. Others have had the benefit
of taking up theoretical work done by linguists which has then been

2
The latter also often being developed and used for the gathering and analysis of data for writing new
dictionaries.

53
applied to teaching/learning situations. The teaching of foreign languages
has been benefited by studies in special language teaching/learning and
to a lesser extent special language teaching/learning has benefited from
general language teaching/learning (Robinson 1991, Swales 1985).
Hutchinson and Waters (1987, Ch. 2) identify four stages of
development of special language analysis, with a fifth emerging. To a
greater or lesser extent, all are germane to my project. They are:
1. The concept of special language: register analysis,
2. Beyond the sentence: rhetorical or discourse analysis,
3. Target situation analysis,
4. Skills and strategies
5. A learning-centred approach.
In this analysis the last two of these stages are considered to be
methodological approaches and not language research as such. Indeed,
Hutchinson and Waters themselves identify this division as one of “new
ideas about language and new ideas about learning” (ditto, 1987:14). The
methodological implications of learning English for science and
technology will be addressed later (Chapter 7 The Syllabus) after
presentation and discussion of the results of the research undertaken.
I will briefly discuss each approach before showing how they feed
in to my thesis.

2.2 The Register Analysis Approach

The very first research material published for teaching was not
meant for special or specific purposes but rather for the general learner.
The idea that students could be helped to learn languages more easily
and quickly by having a select list of words came into vogue between the
wars. In terms of word lists or frequency counts, the earliest which was

54
used to provide a scientific foundation for teaching was developed in
America by Thorndike. Thorndike produced a list of 5,000 words for
teachers, culled from a corpus of four and a half million words, which he
published in 1921 as the Teacher’s Word Book. Subsequently, Horn
published 10,000 words taken from business and personal letters in
1926. The number of words published then multiplied to 20,000 in 1931
with Thorndike’s The Teacher’s Word Book of 20,000 words and in 1944
to 30,000 words with Thorndike and Lorge’s The Teacher’s Word Book of
30,000 words. These counts were used to decide on appropriate reading
materials for school children and are still popular means to decide on
appropriate materials for different school levels. In Britain, Michael West
published his A General Service List of English Words in 1953 which
contained 2,000 words and a supplementary list of scientific and
technical vocabulary. Similarly, the graded readers in English such as
those published by Penguin are meant to correspond to levels of reading
competence in EFL learners learning British English. West’s work like
that of Palmer was specifically addressed to EFL learners whereas the
American work by Thornton and Lorge was for general reading in
mainstream education systems. However, all of these word lists were
based on written material but in France work by Gougenheim was taking
place on a spoken corpus which was to provide the basis of Français
Fondamental, first published in 1954. The first revision of this was
published in 1959 and was composed of 1475 entries, 1222 of which
were lexical items and 253 grammatical words. As will be shown later
(2.5 The Corpus Analysis Approach) the whole concept of distinctions
between grammatical words and lexical items is called into question
through corpus research. Nevertheless lists can still be used to help
define a specific type of language or special language and to underpin
syllabi.

55
2.2.1 European Languages for Special Purposes

Just as the Second World War had served as an impetus for

teaching foreign languages expeditiously and for research into teaching
foreign languages in Britain and America, the European Union has been
and still is a driving force behind research on foreign language teaching
and languages for special purposes (LSP) in the new Europe. There have
been studies published from work commissioned by the Council of
Europe, namely Van Ek’s the Threshold Level (1976) and Waystage
English (1977) designed to specify exactly what children should be taught
in their secondary school courses in English as a Foreign Language.
These last two continue to be published and similar books are now
produced for all the member countries’ languages. The Threshold Level
has in fact been republished as recently as 1994 and there is now a
Threshold 1990 published together with Trim which purports to set out
“how a learner should be able to use English in order to function
independently in everyday communication.” The focus given in this kind
of work is on secondary education particularly and is designed to
encourage all European citizens to speak other European languages. The
intention that is usually explained is that all citizens should know a
minimum of three European languages. However, as was mentioned in
Chapter 1, 1.4 The Dominance of English in Science and Technology,
certain European languages like English are becoming much more widely
used and is learnt at school by an increasing number of European
citizens.

2.2.2 Methodologies

These different lists of what children should be taught on language

courses in schools were produced based on different methodologies. The

56
Threshold Level reflects the functional/notional approach popular in the
1970’s because of Wilkins’ work (among others), Wilkins (1976) Notional
Syllabuses, Oxford: Oxford University Press. The lists produced based on
the functional/notional approach were not produced empirically and
many other possibilities of what to include at what stage in learning
exist. The variety of coursebooks produced to teach functions/notions in
the late 1970s and early 1980s reflect this and authors such as Abbs at
the TESOL Conference in Lisbon in 1979 admitted that things were
beginning to get out of hand when students were being taught how to
react angrily to situations! So as a basis for syllabus design this
approach left something to be desired.
If an empirical approach to register analysis is to be adopted for
syllabus design, then there is more to the production of frequency counts
that needs to be taken into account. For example there is the question of
the definition of the words listed, reflected in the description of the
Français Fondamental list mentioned above. What is lexical and what
grammatical? The whole question of what a word is must also be
addressed. Moreover, the corpus from which the word lists were
produced must be clearly categorised so that the results can be seen to
be pertinent to the learners’ needs. These aspects must be made very
clear if any comparison between research undertaken can be made and
scientific results verified. Therefore, further consideration will be given to
these aspects later in examining the results obtained from the corpora.

2.2.3 Scientific Specificity

Work by linguists, such as Halliday & Martin, Swales and Abbott,

has argued for scientific specificity in order to make research comparable
and, therefore, to add more to our understanding of language and
language contexts. Abbott (1980:121), although writing on error analysis,

57
argues that any published analysis should provide enough information
for another researcher to reproduce the results, as with any other piece of
scientific research. Halliday (1993:103-4) wants the research to make use
of the same theories and methods of analysis (understandably advocating
his own systemic functional grammar analysis here, above all others), so
that comparison with other studies is possible. Robinson (1991:31) wants
specific information on the materials used for the research to be
provided. She (ibid.) says:

“First, the fact that research exists on the same topic or subject matter that
the students are interested in is not sufficient to make that research useful.
We need to know the source of the material that has been researched: its
date and geographical origin. In addition, we need to know the level of the
material: does it represent specialist to specialist communication, or
specialist to non-specialist? What was the mode of the material? Was it
originally spoken or written, prepared or unprepared? All these alternatives
will have an effect on the language forms selected.
Second, we need to know the size of the corpus that has been researched
(.....) Larger-scale studies may be able to arrive at reliable generalisations.
Smaller-scale studies, however, may be able to go into more explanatory
detail.”

Robinson identifies one of the major difficulties of making use of

much of the early research work carried out. Many studies exist but the
exact situations to which they refer and/or the material on which the
studies were based were often not clearly identified. This meant that it
was difficult to build on those studies and many researchers were in
danger of repeating work that had gone before. An example of this is the
number of studies carried out (and the range of results obtained) on the
active-passive in science texts. Even distinctions between what is
analysed as active or passive has been discussed and various categories
58
like the stative (Lackstrom, Selinker and Trimble) or equative be as main
verb (Wingard) have been isolated and investigated as sub-categories of
this dichotomy. This situation of reproducing work on similar topics
probably arose because researchers were trying to cope with a situation
they were presented with, usually abroad, and which needed solving in a
short space of time. This was often the case for many of the university
level courses in English in the 1970s designed for undergraduate
students on many different courses, from economics to engineering.
Porter (1976) laments the fact that much of the research that had
been carried out in other places was never published which meant that
valuable data may well have been lost forever. Porter (1976:77) is
particularly concerned that those linguistic analyses of ‘the language of
science’ which have been lost would have been of “inestimable value” for
others engaged in constructing teaching materials. Swales (1984:1)
echoes Porter’s sentiments when he says that although research was
taking place in England in the sixties, it was usually being carried out by
students on postgraduate courses and the “resulting projects and
dissertations have long been buried in departmental and university
libraries and are now almost completely forgotten”. Even more recently
linguists like Stubbs (1996:152) assert that “it has often been difficult to
find studies which build on previous work”.
Robinson (1980:18) argues that any useful research carried out
“tends to confine itself to very limited, clearly defined, areas” such as, for
example, prepositions in chemical abstracts. There is considerable
difficulty in identifying only one “English of Science”, because there is
such a confusion of sub-genres or modes.3 This state of affairs has made
it difficult for textbook writers to base their material on register analysis.

3
See McCarthy and Carter 1994 Pp 4-16 for a discussion of modes and their features.

59
 Stern (1983:131-2) argues that the study of lexis or vocabulary
has received little attention from English-speaking linguists4 because it
does not lend itself easily to structural and systematic treatment in the
way that syntax and phonology have done but that this is an area of
research which is very important for language teaching. More recent
studies (Sinclair 1991) would suggest that the learning of lexical items in
isolation does not reflect actual English usage where words and their
meanings are associated with particular structures and contexts.
Sinclair therefore suggests that words must be studied in context in
order to show their specific meanings and associated structural
restrictions.

2.2.4 Syllabus Implications

Halliday put forward the idea that language shows variety in terms
of its use and not its user. For example, there is language specific to food
or cooking: tomato, apple, bread, butter or sport: referee, goalkeeper but
the user can only show different dialects which may be regional or social
and so on. This sets the course for discourse analysis which will be
described later but which Halliday (1993) still refers to as Register
Analysis. Investigation of varieties of English can then show what the
learner needs to cope with in a specific area.
Register analysis is based on the idea that nouns make distinctions
because they are used for concepts or principles (which is somewhat
similar to the ideas of Condillac in the eighteenth century, Philosophical
writings of Etienne Bonnot, Abbé de Condillac, Hillsdale, N.J.:Lawrence
Erlbaum, 1982). Consequently, if the register of say biology, as distinct
from other registers, can be identified, then a specific syllabus can be

4
Stern points out that research has been carried out by French and German linguists and reference is made
later (2.2.7 The Impact of Modern Technology on Register Analysis) to Hoffman who describes the
research done (on English) by German linguists in the GDR for teaching purposes.

60
drawn up which would be more limited in range and therefore, would not
diffuse students’ learning energies. West describes this as the ‘surrender
value’ of the course of study. A high ‘surrender value’ would mean greater
efficiency in terms of meeting the student’s language requirements or as
Swales (1985) puts it, as getting ‘maximum educational value’ out of the
course.
White (1975) reached the conclusion that it was a “unique
constellation of features rather than any single characteristic” that made
one register distinctive from another. These features however have to be
identified from the kinds of materials that are likely to be used by the
students so that a pedagogical selection can be made for course design.
Mindt (1997:42) describes this process of designing a grammar for foreign
language learners as:

•the compilation of a corpus of language data

•the construction of a didactic grammar from this corpus
•the derivation of pedagogical grammars from the didactic grammar

Sinclair and Renouf (1992) suggest that in general language courses “the
main focus of study should be on:

a. the commonest forms of the language

b. their central patterns of usage
c. the combinations which they typically form”

These principles should hold for the results of register analysis of a

particular variety of language usage such as that of science and
technology, where the context in which language occurs would be of
particular importance.

61
2.2.5 Publications and Coursebooks based on Register Analysis

Following this principle Barber (1962) published “Some Measurable

Characteristics of Scientific Prose” in Contributions to English Syntax and
Phonology, Stokholm: Almquist & Wiksell, in which he identifies three
subject areas which are common to all fields: sentence structure, verb-
forms and vocabulary. He analyses 350 sentences and finds 2
commands, 3 statements and commands and 345 statements which he
compares with unpublished data from a colleague doing postgraduate
work at the University of Leeds, W. Rumszewicz, who had carried out
similar work on both English textbooks on agricultural studies and
passages of prose drama. He concludes that in his colleague’s scientific
texts all sentences are statements but in the dramatic texts only two-
thirds are statements, the remainder being questions or requests. There
were also a number of articles and books published following the register
analysis approach: Herbert (1965) published The Structure of Technical
English, London: Longman, Swales (1971) based on his experience in the
Middle East Writing Scientific English, London: Nelson, Ewer and Hughes-
Davies (1971-72) based on their work in the University of Chile “Further
notes in developing an English language programme for students of
science and technology”, English Language Teaching, 26, 1& 3, Strevens
(1973) Technical, technological and scientific English, Sudhikam (1975)
“Lexis” in Guidelines Sample Materials for the Teaching of English to First
Year Tertiary Level Scientific and Technical Students in Universities in
Thailand, SEAMEO/RELC (Mimeo), Friel (1978) “A verb frequency count
in legal English”, ESPEMA Bulletin 10, Wingard (1981) “Some verb forms
and functions in six medical texts” in English for Academic and Technical
Purposes (edited by Selinker, Tarone and Hanzeli), Newbury House, and
Palmer (1981) Register research design. Rowley, MA: Newbury House.
Coursebooks such as Ewer and Latorre’s (1969) A Course in Basic

62
Scientific English and Herbert’s (1965) The Structure of Technical English
demonstrate this approach. Swales (1985:18) comments that Herbert’s
book was still in print and still being used when he published his book
twenty years later. He (ibid.) attributes this to the fact that it “shows a
highly professional concern with the language of EST” however the
methodology used was rather dull and the combination and connection
between the diagrams used and the accompanying text was often
obscure.

2.2.6 Criticism of Register Analysis

One of the first criticisms levelled at register analysis was the

fundamental one that there was no such register as a “general” register
which could be used for comparison with a “special” register (see above:
Syllabus Implications). Halliday takes this position but argues that it is
useful to recognise a category of “special-purpose” language or language
varieties.
A second kind of criticism of the methodology of register analysis
began to appear grounded in the idea that no specific information on the
corpus that had been used to derive the results was presented and what
information there was showed that no spoken language had been
included, which would imply that the analysis was not sufficiently
representative to be able to provide generalisations about language.5
Swales (1984:10) claims that frequency analyses do provide
evidence for generalisations about a “variety, type or style of the
language” and that the people who criticise frequency work most are
“those most given to making claims that such-and-such a feature is
important, frequent or interesting without any evidence to support these
claims”. Porter (1976:77) also argues that attempts at linguistic

63
characterization have been “surprisingly careless” and that even though
Bloomfield wrote forty pages in 1938 on “Linguistic aspects of science”
only one explicit example is given throughout those forty pages. This
problem continues to the present day for example, although Stubbs
(1996:152) argues for clarity on texts analysed, he is himself open to
criticism (Hoey 1993) for not giving sufficient information on the school
textbooks in his own analysis.6
The criticism was based on the fact that researchers appeared to be
claiming that features were unique to one type of text or that one feature
uniquely characterises a text. Once it was understood that the distinction
was much more of degree the objections were largely overcome. As White
(1975) said:

Firstly, it became clear that ... it is not possible to take the occurrence of
any specific feature as being criterial of one and only one particular
register. Secondly, it was obvious that what made one register distinctive in
comparison with another was a unique constellation of features rather than
any single characteristic.

However, the difficulty of finding sufficiently large scale studies to work

from remained.

2.2.7 The Impact of Modern Technology on Register Analysis

Before the advent of the computer it was generally believed that it

was impossible to distinguish unambiguously between texts and that,
therefore, such analyses were not worth carrying out, owing to the
difficulty in coping with large corpora and because much of the language

5
Although Robinson’s comment on small scale specific studies, see above P 55 Scientific Specificity, could
still hold true.
6
First presented with Andrea Gerbig (1993) as “Human and Inhuman Geography: On the Computer-
Assisted Analysis of Long Texts.” In Hoey (ed.1993) Data, Description, Discourse. London: HarperCollins.
64
that is found in corpora is not distinctive7. This led to the idea that there
is a core of language that is common or sub-technical (cf. Robinson
1991). Trimble (1985:129) equates sub-technical vocabulary with “those
words that have the same meaning in several scientific or technical
disciplines” together with “those “common” words that occur with special
meanings in specific scientific and technical fields”. This approach goes
hand-in-hand with the methodological principle that only students of an
intermediate level of English competence could or should be exposed to
an English for Special Purposes course, as these students will already
have sufficient basic knowledge of the language to be able to appreciate
the difference between these common forms and the language that is
scientifically specific or purely technical. Trimble (1985:7) suggested that
students at the tertiary level are “assumed to be fairly advanced in
English” but, nevertheless, recognised that not all of the students could
be assumed to be equally accomplished in all of the language skills.
In the same line of argument, Hoffman (1981:114) claims that:

“It is the significantly frequent occurrence of certain speech elements, forms

or structures that characterizes scientific writing and spoken discourse. As a
consequence statistical methods play an important role in selecting an
inventory for teaching purposes... It is the word and the phrase levels that
yield the best results, i.e. lists of typical lexical items which may serve as a
highly effective teaching/learning minimum.”

This statistical approach to language teaching, ‘lexicostatistics’ was

particularly prevalent in the universities and technical colleges in the
Eastern bloc countries such as the German Democratic Republic where
frequency counts and terminological dictionaries were produced. These

7
Recent large corpus studies by Sinclair (1991) demonstrate that there is a body of very frequently
occurring lexis and that if those items that only occur once are removed from the frequency list it shrinks
to half its size.

65
days strict adherence to this sort of approach would be deemed too
limiting for syllabus design, although it is generally accepted that it still
has a part to play in it.
Swales (1984:1) reported that although frequency analyses found
little favour in British and American ESP work, a revival of this form of
study is taking place because of the fact that “frequency analysis is
ideally suited to computerization”. He also predicted (1984:214) that ESP
would only come of age when computers and video recorders were used
and the processes of technical and sub-technical vocabulary acquisition
were properly investigated and not merely imagined. Computers, through
concordances, can already provide learners with much better
investigative tools and give access to real language use instead of
inventions by the course writer. Tribble and Jones (1990:15) find that the
results obtained from a concordance “will only be as interesting as the
raw material on which you put it to work” so that appropriate corpora
must be used to generate instances of language usage which match the
students’ needs.
Biber, Conrad and Reppen (1998:136) say that teachers must
understand the processes by which register is understood so that they
can facilitate its acquisition. However, they go on to suggest that the
ability to describe and understand the differences between registers has
proved to be very difficult without the use of corpus-based studies.
Furthermore, the features that distinguish one register from another are
rarely features unique to that register. Registers usually share many
linguistic features; it is the relative use of these features that usually
distinguishes one from another. Therefore what is needed is a
comparative quantifying approach in order to know whether one feature
in a register is rare or common.

66
2.2.8 Variation Studies

Ewer and Hughes-Davies’ (1971) syllabus suggests that particular

forms such as the present simple tense, the passive voice, conditionals,
compound nouns and anomalous finites are more prevalent in scientific
writing, and so many of these areas were researched in the 1970’s. Some
of that research led to further divisions being made, for example between
passive and stative (Lackstrom, Selinker and Trimble 1972). Strevens
(1973) put forward the suggestion that different kinds of language could
be distinguished “within ‘scientific discourse’. These were: the scientific,
the technological and the technical, each of which, according to him,
contains its own characteristic range of style.
Halliday (1993:37) points to nominalization as a characteristic of
scientific prose. Halliday takes this further and argues that “at the level
of social context, ideology is realized by genre, which is in turn realized
by register”, in other words he sees connections of an ideological nature
within text from its register to its genre. Halliday (1998:186-7) says that
grammar, “in its ideational guise” transforms “human experience into
meaning” but it also “impose(s) a categorisation” on “our perceptions of
phenomena”. This preoccupation with ideology is now dealt with in
studies of social semiotics in discourse analysis.
More recently variation studies by such people as Biber (1988) and
Finnegan have postulated criteria for the identification of genres and text-
types and have reached the conclusion that several combinations of
factors (both present and absent in specific genres) help to pin down the
linguistic features that can be found in those genres. Their research also
shows that there are no absolute black and white differences between
genres and that this is particularly the case in the various sub-genres of
academic prose. Biber (1988) postulates a continuum as a better
representation of differences between genres with speech at one extreme

67
and formal writing at the other. Nevertheless, academic prose was found
by Biber to be one of the most widely differing sub-genres in the analyses
he carried out. Biber’s work will be discussed in more detail later in 3.3
Biber’s Methodology of Variation Studies and Corpora Analyses when
making a classification of the textbooks on the undergraduate students’
bibliographies.

2.2.9 Recent Studies

Work on register analysis is still very dynamic at the moment

particularly with Halliday and the ‘Australian school’. Halliday (1993)
dates scientific discourse as having arisen following Newton, in the early
eighteenth century and describes some changes that have evolved in
scientific writing. Halliday (1993:58) asserts that Newton showed no
coyness about using I in his writing and that the ‘suppressed person’
passive favoured by teachers and scientific editors only came into fashion
towards the end of the nineteenth century. The point being that scientific
language has not only been invented for the purpose but that it is also
historically mutating.
Dissatisfied with register analysis as a means of identifying
differences between genres, some researchers put forward the idea that
what made one genre different from another was at a ‘higher’ level than
the sentence (or supra sentential, see Discourse Analysis below) which
was initially termed discourse or rhetoric analysis. The preoccupation of
linguists such as Strevens (1977), White (1975) and Candlin et al. (1978)
was the purpose of the language use. Nevertheless there is a perpetuation
of register analysis within discourse analysis so that these two terms are
often used indiscriminately (or together) in modern methodologies.
McCarthy and Carter (1994:36) represent the division of register and
genre for ‘reporting genres’ in the following way:

68
 Discourse world
↕
Core generic function
↕
Genres
↕
Generic blends
↕
Registers

Where the Discourse world is divided into spoken and written; the Core
Generic Function is Reporting; the Genres are for example, the
Information Report, the Progress Report or the Weather Report; the
Generic Blends are for example, Reporting and Predicting, Reporting and
Recommending or Reporting and Evaluating; and the Registers are for
example, a Weather Forecast which can be even further sub-divided into
TV/Radio and Newspaper. These can then be linked to different
“Prototypical linguistic features”, such as “Past tense, passives, relational
processes”. McCarthy and Carter (1994:33) survey some of the
uncertainties that are found about definitions of genre and suggest that
the results are that “the notion of genre becomes as slippery as the
notion of register.” They ask if this distinction is necessary at all but they
note that it has had important implications for discourse analysis.
Biber (1988) uses genre to refer to “categorizations assigned on the
basis of external criteria” and text-type to refer to “groupings of texts that
are similar with respect to their linguistic form.” The term register is still
used most extensively by Halliday and the Australian school.

69
 2.3 The Discourse Analysis Approach

2.3.1 Definition

Coulthard (1978) defined the concern of discourse analysis as ‘the

identification and description of supra-sentential linguistic structure in
written and spoken texts.’ Widdowson (1971) exemplified the discourse
analysis approach to teaching as asking what the students need to use
the language for, its uses and functions. Students must be able to
describe, explain cause and effect, make comparisons and classify. His
argument is that these functions of language do not change whether
students are studying geography or chemistry but that they cross the
subject boundaries. Furthermore, it was suggested that since students
have been taught these functions in their first language (L1), teachers
can appeal to this framework and with some adjustments only need
teach the English realisations of these characteristics.
Widdowson (1977) advocates a distinction of terms between ‘text’
and ‘discourse’. He suggests that text is viewing a stretch of language as
an exemplification of the structure of the language, especially of devices
to indicate structuring above the level of the sentence. Discourse he
describes as viewing a stretch of language as a unique piece of
communication.

2.3.2 The American School

A similar and parallel school of thought developed in America with

Selinker, Trimble, Lackstrom and Todd-Trimble, who were concerned
about the functioning of language above the sentence level and with
identifying the organisational patterns in texts and the linguistic means
whereby these are realised, as compared with register analysis which had
concentrated only on sentence grammar. Discourse analysis does not rule
70
out the analysis of individual sentences and so subsumes register
analysis to some extent (see Widdowson 1977 above). However, in order to
discover how given grammatical structures come to have given meanings
in given contexts, or how larger textual or topical constraints affect the
choice of individual lexical, exophoric or anaphoric items within a given
clause, discourse analysts believe it is necessary to study whole texts8.

2.3.3 The Prague School

In the decade preceding World War II a Czechoslovakian branch of

‘functional’ linguistics came into being, which is now referred to as the
Prague School. These linguists were led by the work of Vilém Mathesius,
the Russian Nikolay Trubetskoy and the Czech born American Roman
Jakobson. The dominant characteristic of the Prague School approach
was its combination of structuralism with functionalism and how it
stressed the importance of the cognitive, the expressive and the conative
(or instrumental) functions fulfilled by language. In their approach the
cognitive function of language refers to the transmission of facts, the
expressive to the indication of the mood or attitude of the speaker (or
writer); and the conative its use for influencing the person addressed or
bringing about some practical effect.

A number of scholars of the Prague School have suggested that

these functions correlate in many languages with the grammatical
categories of mood and person. The cognitive function is fulfilled by 3rd
person non modal utterances; the expressive by 1st person utterances in
the subjunctive or optative mood; the conative by 2nd person in the
imperative. In their work on criticism of literature and their stylistics

8
Once again text here should be understood to mean both written and spoken language.

71
studies, their key principle is that language is being used poetically or
aesthetically when the expressive aspect is predominant.

The Prague School also conducted a lot of research into phonology

and ‘markedness’. The latter was then extended into morphology and
syntax and it was suggested that morphologically unmarked forms have a
much wider range of occurrences and a much less definite meaning than
a morphologically marked form. For example, morphologically jumped is
‘marked’ for past tense by means of the ed ending or suffix, whereas jump
is ‘unmarked’. Jump is not ‘marked’ for present tense and therefore
occurs more widely. Similarly, in vocabulary dog is ‘unmarked’ and bitch
‘marked’. Dog is a more general category and can therefore be used in
many more contexts than can bitch which is a specific (female) type of
dog.

More recently their work has been on the distinction between

‘theme’ and ‘rheme’ and the notion of ‘communicative dynamism’ or
‘functional sentence perspective’. By the theme of a sentence is meant
that part that refers to what is already known (also called ‘topic’,
‘subject’, ‘old’, ‘given’ or ‘shared information’ by other linguists) and by
rheme the part that conveys new (or unknown) information. Halliday,
although a member of the British School of linguistics (see below 2.3.4),
uses both ‘theme’ divided into ‘theme’ and ‘rheme’ and ‘given’ and ‘new’
information in his discourse analyses. It can be seen from this how the
different schools in fact interact with one another and often deviate only
through their primary focus.

2.3.4 The British School

The British school of ‘functionalism’ was led by linguists like J.R.

Firth, Michael Halliday and John Sinclair. These linguists rejected the
72
study of language ‘by itself’ and resolved to study what people actually
say in context. Firth considered three features of “typical repetitive events
in the social process”, which were:
1. The participants: persons, personalities and relevant features of these.
(a) The verbal action of participants .
(b) The non-verbal action of the participants.
2. The relevant objects and non-verbal and non-personal events.
3. The effect of the verbal action.
Sinclair’s group pioneered discourse analysis through fieldwork on
classroom discourse. They rejected the traditional linguistic terminology
and referred to discourse ‘moves’ such as ‘initiation’, ‘nomination’ and
‘follow-up’ by the teacher and ‘bid’ and ‘response’ by the learner (Sinclair
and Coulthard, 1975; Sinclair and Brazil, 1982). The hypothesis that has
developed today (Sinclair 1992) from this work is that a three-part
structure (Initiation, Response, Follow-up) is always the option in spoken
interaction and is virtually obligatory in many types of discourse. All of
the neo-Firthian linguists have taken the view that the analyses they have
made and theories they have developed could have a practical application.
Van Dijk (1997:29) identifies this preoccupation with actual
occurrences of spoken language as the basis of most of the current
research in discourse analysis. However, Halliday and Martin (1993) are
more interested in how the translation of one type of discourse changes
its status by being translated into another discourse. In their case it is
the loss of status involved in mediating of the language science in schools
through speech and testing through short answer writing as opposed to
through writing reports like those the students will need to interact with
in order to become proficient in the language of science. Halliday
(1993:202) sums up the argument thus:

73
 To rehabilitate literacy in science teachers and students will have to work
towards a much clearer grasp of the function of language as technology in
building up a scientific picture of the world. Technical language has evolved
in order to classify, decompose and explain. The major scientific genres –
report, explanation and experiment – have evolved to structure texts which
document a scientist’s world view. The functionality of these genres and the
technicality they contain cannot be avoided; it has to be dealt with. To deal
with it teachers need an understanding of the structure of the genres and the
grammar of technicality. With this knowledge they can begin to tackle the
problem of science literacy ... Without it they will continue to focus on
content without taking language into account, probably with an increasing
emphasis on science activities rather than science texts. The linguistic
technology is the key -–not just to science literacy but to understanding and
practising science itself. Ways must be devised to provide access to this
technology. And the answer must not involve watering the technology down.

2.3.5 Systemic Functional Grammar

Halliday built on Firth’s schema and developed his systemic

functional grammar to analyse genres. His systemic functional model
divides the communicative function of a text into three:

• Field of discourse (social action): the processes, purposes and subject

matter with which the participants are engaged.
• Tenor of discourse (role structure): the linguistic and extra-linguistic
role relationships of the participants. The linguistic roles are speaker
and hearer, initiator and responder and so on. The extra-linguistic
roles are the social identities and relationships of participants.
• Mode of discourse (symbolic organisation): the part that language plays
in situations and how it does so. This includes such notions as

74
 medium (e.g. written, spoken, typed) and immediacy (e.g. face-to-face
or distant).
Halliday further defined these aspects of register (organisation of
content) through their metafunctions (organisation of language). The
choices for meaning are organised into the following metafunctions: Field
is associated with Ideational meaning (resources for building content);
the metafunction associated with Tenor is Interpersonal meaning
(resources for interacting) and the metafunction associated with Mode is
Textual meaning (resources for organizing texts).
Halliday argues that scientific texts are derived historically from
the need to condense information about previous scientific discoveries.
They are therefore characterised by dense nominalization as this is the
best means of conveying dense information. Here, once again, is the
notion that solid background (or underlying) scientific knowledge of the
subject being studied is necessary. Furthermore, this background
knowledge is assumed by authors of textbooks to exist in their readers.
This is in conformity with Labov’s (1972) suggestion that only by using
the concept of ‘shared knowledge’ can discourse be interpreted correctly.
However, Trimble (1985:114) says that his research “showed and
continues to show that the majority of non-native students lack the
cultural background that enables them to bring more than a very limited
amount of the presupposed information to their reading of EST
discourse”. Here he is referring to the information presupposed by the
writers of scientific and technical discourse to be ‘possessed’ by the
reader.
Horne Tooke (1778, 1786) argued back in the eighteenth century
that language contained ‘abbreviations’. Like Halliday’s ‘condensation’ of
knowledge he argued that abbreviations were necessary so that thoughts
could be expressed in real time and that abbreviations had been
developed over time so that an (empirical etymological) analysis of

75
language would show that abbreviations such as prepositions could be
traced to their historical (nominal) roots. In the Diversions of Purley: 9-15
he says:

“Abbreviations are employed in language in three ways:

1. In terms
2. In sorts of words.
3. In construction.
Mr. Locke’s Essay is the best guide to the first; and numberless are the
authors who have given particular explanations of the last. The second only
I take for my province at present; because I believe it has hitherto escaped
the proper notice of all.”

Halliday (1993:148) argues that scientists have appropriated

resources that already existed in English for their own purposes “to
create a discourse that moves forward by logical and coherent steps, each
building on what has gone before” and that this had emerged as the most
highly valued model by the end of the eighteenth century. Nevertheless,
he suggests that modern school syllabuses reflect an increasing
emphasis on doing as opposed to learning science which means that
“children are not taught to access the genres science has evolved to store
information which leads to tremendous inefficiency in the science
curriculum.” Children, in his view, should be encouraged to study the
text organisation of scientific texts to become literate in science.

2.3.6 Rhetorical Moves

Swales, in a seminal article, (1981) examined the introductions to

48 scientific journal articles using a “top-down” approach to the analysis
of science texts. This means that he began with a consideration of overall
text organisation and any statements regarding choice of structure at the
76
sentence level or below are related to that higher level organisation. He
concluded that there are certain rhetorical moves or ‘macropurposes’
within discourse. In the case of the introductions, he claims there are
generally four such moves;

• Move 1 Establish the field,

• Move 2 Summarize previous research,
• Move 3 Prepare for present research,
• Move 4 Introduce present research.

Swales argued that these moves were more important than standard
English grammar. However, in his later work (ibid. 1984:213) he says
that genre-analysis has a “price to pay” in that by revealing something of
the “internal logic and external language of a conventionally-constrained
communicative event” it may “have little to say about other, apparently
quite similar, communicative events”. He gives the example of his own
work when he says that there is “no such thing as an Introduction in
academic writing” and explains that introductions “would appear to be
quite differently organized in different genres such as scholarly papers,
theses, projects and essays”. These findings suggest that analyses of text
organisation must be carried out on specific genres which are relevant for
students in a particular setting in order to develop specific teaching
materials that the students could use to develop their understanding and
scientific literacy in the Hallidayan sense given above in 2.3.5. McCarthy
and Carter (1994) give some suggestions about how this kind of text
organisation analysis can be carried out in the classroom using
frameworks.
There is however another practical application of this type of
analysis in specific fields such as business negotiations. Johns (1991)
reports that interest in Uljin and Gorter’s (1990) work on

77
discourse/rhetorical moves in business negotiations has increased with
the enlargement of the European Union and the consequent enlargement
of different language contexts/interfaces that must be dealt with.

2.3.7 Organisational Features

Some of the early claims made about the organisational features of

texts particularly that of paragraph structure (see Ewer 1975) have been
called into question by other researchers. Trimble (1985:14-18) finds that
there is a ‘conceptual’ paragraph as opposed to a ‘physical’ paragraph in
EST. He argues that the ‘standard’ definition of a paragraph is that it is
made up of a group of sentences that express a complete thought and
which are set off on a page of text by indentation or spacing. However, his
analysis of written EST discourse showed that the ‘complete thought’
idea, given in the first part of this definition, might in fact be realised by
two or three physical paragraphs which deal with lower-level
generalisations or details supporting the main generalisation, which all
together would make up the ‘conceptual paragraph’. Most linguists would
now accept that the organisation and cohesion of a text is not restricted
to discrete paragraphs but goes over and above that level and may be
found even at the level of chapters in coursebooks. (Phillips’ unpublished
Ph.D. thesis Birmingham 1985 cited in Hoey 1993). Hunstan (1994:204)
argues for “unit boundaries” in discourse which, in her study of a report
on experimental work, sometimes coincided with paragraph boundaries
and sometimes occurred within paragraphs. She (1994:206) concludes
that “Changes in status coincide with transitions from one unit to the
next, while value and relevance serve to bind together sections that may
cover several status categories.”
Heslot (1981) argues that studying a full text section by section can
show important features of those sections that would otherwise be

78
masked by a study of the whole text. She suggests that her research is,
nevertheless, in line with Selinker and Trimble’s (1976) recommendation
to work at a higher than sentence level.
The ‘whole text’ approach has gained more and more adherents in
recent years. Hoey (1991) argues for it as does Stubbs (1996). Much
published work however still concentrates on detailed analyses of small
fragments of texts. Stubbs (1996) recognises this fact but goes on to
argue for complementing the analysis of text fragments by the analysis of
long texts. McCarthy and Carter (1994:112) claim that “Matters
traditionally thought of as the domain of semantics and syntax can be
placed squarely at the heart of discourse analysis.” They also suggest
that (1994:106) a top-down approach can “assist the job of relating
higher order categories in the syllabus (such as text-type) to the micro-
syllabus elements of grammar and lexis”.

2.3.8 Discourse Rules

Cooper (1981:426) argued that “the unpractised reader of academic

discourse in a language other than his own may not know the rules of
discourse for that language; or if he knows the rules (because they
underlie discourse in his own language), he may not know the ways in
which they are realised in that language.” He suggested that it may only
be necessary to encourage the student to use his existing discourse
knowledge to ‘decode’ the second language. Trimble (1985:120) takes
exception with this view which he believes leads to confusion; he cites the
example of ‘should’ in its everyday use in contrast with its specific
(different) meaning in scientific texts, he says:

“In our experience non-native students tend to transfer their reading

techniques developed for ‘general English’ to reading EST discourse without

79
 realizing that adjustments are often necessary. As a result they read ‘should’
with the meaning found most commonly in ESL/EFL grammars and so
assume that a choice is possible.”

The distance between Portuguese and English is not very wide

compared with other pairings of languages but Scollon and Scollon
(1995), discussing professional communication, warn that there may be
problems of interpretation because the communication is between
members of different discourse systems and it must be assumed
therefore that, “knowledge, assumptions, values and forms of discourse”
will not be shared and the problems that are likely to arise must be
anticipated in order to achieve effective communication. Some of the
dominant features of Portuguese academic prose are elaborate involved
subordinate clause structures, the use of the first person plural ‘we’ verb
form to describe work carried out in preference to the passive and
paragraphs that may be composed of only one sentence. However, the
influence of different schools of thought also underpin academic writing;
the English often follow empiricist lines such as that of Firth mentioned
above in the British School of linguistics whereas the French often follow
more rationalist lines. It would be possible to find both of these
influences in Portuguese academic prose together with the more modern
integrative or ‘ecological’ approaches of the 1990s. Swales (1985:72)
argues that “in cross-cultural situations we have again to take up the
matter of ‘teaching’ in some way the rationale of scientific communication
in English” and the idea of using a methodology based on the teaching of
science in the first language put forward by Widdowson (1974) must be
abandoned because that is now recognised to be a local phenomenon.

80
2.3.9 Coursebooks based on Discourse Analysis

A number of coursebooks began to come onto the market which

reflected the principles of discourse analysis notably the Allen and
Widdowson (eds.) English in Focus series which covered nine areas:
Mechanical Engineering (1973), Physical Science (1974), Workshop Practice
(1975), Basic Medical Science (1975), Education (1977), Agriculture (1977),
Social Science (1978), Biological Science (1978), and Electrical Engineering
and Electronics (1980).
Professor Widdowson assumes that students already have some
knowledge of science and some knowledge of English which need to be
put together. He sees the student as looking for the means to express in
English what he already knows in his first language. However, despite the
theoretical underpinning brought to the series by the editors, these books
were not a great success which Swales (1985:72) attributes to a certain
rigidity in the format and exercises which fail to take sufficient account of
teachers’ experience.

2.3.10 Criticism of Discourse Analysis

Despite this series, criticism was levelled at the utility of discourse

analysis for teaching materials. Some felt that in the 1980s discourse
analysis had only shown the relevance of coherence in text based on
Halliday and Hasan’s (1976) work in this area and that this was really
rather too little as an entire methodology for generating teaching
materials.
Furthermore, whilst it is undoubtedly true that certain
characteristics like describing and comparing are essential for science
and technology students, it does not follow that students have been
taught these in their first language. Nor does it follow that the

81
realisations of these are the same in the target language as in the mother
tongue. Langkilde (1981:517) found, for example, that for undergraduate
students of economics in Copenhagen long adverbials (in French)
interfere with the comprehension of sentences and “disturbs the well-
established patterns that the students are used to finding”. This
phenomenon may well operate in the opposite direction between English
and Portuguese academic prose as Portuguese corresponds more closely
to French in its use of long adverbials9. Trimble (1985:131) believes that
one of the features of scientific or technical discourse that is a “special
problem for the majority of non-native students” is the use of noun
compounds or strings which are Germanic in origin and so not natural in
many languages which is certainly the case with Portuguese native
speakers.
It is also a fallacy that there is a universal ‘scientific’ way of looking
at things and that everyone with adequate intellectual gifts thinks
‘scientifically’. One of the characteristics of science is that it needs to be
taught to people; it does not exist naturally. Moreover, many scientific
discoveries have been shown to be rather haphazard (or in a hypothetico-
deductive form rather than an inductive one10) and order and method
have only been imposed when the scientists concerned have written up
their work as a paper for other scientists to read.
Beaugrande’s (1997:44) offers an ideological critique of education
and scientific training. He argues that “In theory, all citizens have the
same basic human rights to freedom of speech, public education,
scientific training,” and yet he claims that “in practice, the great majority
are systematically excluded.” This he argues is because they cannot

9
However, Quirk (1995:127) finds a higher proportion of adverbials in speech than expected which he says
“runs counter to the widespread belief that written English is more complex syntactically than impromptu
speech and that the incidence of ‘adverbial clauses’ is a significant marker of relative syntactic
complexity.”
10
An example of this is the book by Watson and Crick (1968) The Double Helix which describes their
discoveries. Karl Popper (1972) in The Logic of Scientific Discovery argues that scientific method is
hypothetico-deductive and not, as many believe, inductive.
82
understand the discourse of science and therefore are not science
‘literate’ as Halliday and Martin (1993) term this phenomenon. When
looking into the future Beaugrande (1997:59), claims that less attention
has been focused upon the ‘twin knowledge crisis and communication
crisis’. However, he believes that there is ‘an exploding body of knowledge
that is locked up in discourse accessible to only a few people
concentrated in centres of wealth and power’ which, he argues, needs to
be made available to everyone through the results of the analysis of
discourse being applied to teaching. So, for Beaugrande, discourse
analysis is to be seen as the key to unlock the door of scientific language.
This position seems reminiscent of the plain English group whose aim is
to make bureaucratic jargon much more transparent for the average
person so that they are not considered “functionally illiterate” as the
Americans describe the inability to cope with filling in forms and other
such language manipulation activities which the average person can be
expected to meet in their daily lives. The language of science and
technology as it has developed over the last two centuries would seem to
be a far cry from bureaucratic jargon because it demands sufficient
background knowledge of the concepts concerned to understand rather
than a certain legalistic hedging of the terms used as in bureaucratic
jargon.

McCarthy and Carter (1994) discuss perspectives of discourse

analysis for use in language teaching and give a number of examples of
how this can be done. They (ibid:122) find that students can be taught
what they call ‘natural interactiveness’ when actual, naturally occurring
language is analysed and the features that are used to achieve
‘successful communication’ are isolated. However, much of this analysis
and description has yet to be carried out to provide a comprehensive
guide for the teacher. It is possible that such work will become
increasingly available for use by teachers in the classroom but the task is
83
huge and it may be that access to data on CD-ROM will be one of the
ways in which students themselves can gain insights into how the
language operates naturally. Training will have to be given for this to be a
feasible prospect with the teacher acting as guide so that neither
overgeneralizations nor simplifications are made. With more work on
discourse analysis and the analysis and use of corpora being included on
teacher training courses this situation should become more viable as
time goes by.

2.3.11 Educational Structures

Moreover, Martin (1993) suggests that students in schools (in

Australia) are being misled about the nature of scientific register and are
not being prepared to understand and write scientific English. He
ascribes the lack of success in science studies to this deficiency. In other
words Martin sees a mismatch between the teaching of science and
acquiring a useful scientific style, leading to a lack of success in science
at school which leads on to fewer students taking up science in higher
education.
Adams, Heaton and Howarth (1991:2) also point out that there is a
“false expectation that educational structures and systems do not differ
internationally” and that students believe that foreign universities
operate very similarly to their own. American textbooks written for
undergraduates would, from this perspective, be working on assumptions
made about the American educational system which is different from the
academic situation of the Portuguese students under study in this
research. The differences between these two systems would reside in the
teaching that goes before, whether science subjects were taught in
secondary school as combined subjects, optional subjects or ‘pure’
science subjects, how these subjects are linked at different stages in the
84
curriculum, the amount of and amount of time dedicated to practical
experimental work conducted by the students themselves or as
demonstrations by the lecturer, the contact hours per week, staff to
student ratios, funding and resources available and so on.
As with Halliday and Martin (1993), Swales and Najjar (1987) had
noted a discrepancy between advice given to teachers and actual
scientific texts they would meet later in their careers. Martin (1993)
describes teachers being instructed to teach secondary school children
science through description and personal involvement which he finds to
be damaging to their later scientific writing and understanding of
scientific texts, which often contain neither of these features. Moreover,
Swales and Najjar (ibid.) found significant variability in the styles of
writing in different scientific disciplines.

2.3.12 Student Competence

Hutchinson and Waters (1987) argue that it is the “underlying

competence” which a student brings to the learning process that must be
examined in order to decide what and how to teach. Widdowson (1974)
argues that students being taught how to read scientific English should
be encouraged to translate in order to discover the functional equivalence
of what they have already learnt about science in their L1, to learn how
‘certain acts of communication which are central to scientific enquiry’ are
realised. The level of the students’ English on entering the university
could thus be seen as only one side of the coin, with previous scientific
learning being the other. Nevertheless, the university level that needs to
be achieved must be sufficient to carry the students on to levels of
understanding they do not yet possess, such as that we would expect
from a university-level textbook.

85
 One aspect of teaching that Hutchinson and Waters advocate
strongly is motivating ESP/EAP students to learn. In universities in
Portugal language studies are often seen as a necessary evil by both
students and staff in science and technology departments. It is therefore
often relegated to a minor position on the curriculum which cannot fail to
reinforce the idea in some students that it is of little importance to their
overall studies when in fact the ability to function effectively in several
languages will often become increasingly important as far as both their
courses and later careers are concerned.

2.3.13 Register and Genre Theory or Variation Studies

Strevens (1978:193-4) answers the question of what is different

about scientific discourse in the following way:

It is not the basic components of his (the author’s) language that differ, it is
the statistical properties of the mixture in which they occur, and the
intention, the purpose, behind their selection and use.

He claims that, among other features, long sentences, long nominal

groups and frequent passives are characteristic of scientific discourse.
However, Mary Todd Trimble and Louis Trimble (1981:199) declare
that contrary to their initial ideas, surface syntax was not a matter of
personal choice on the part of the author. They found

that for native writers of scientific and technical discourse these

grammatical choices were not arbitrary: in fact, we found them sufficiently
patterned that we were able to make generalizations concerning the
relationship between specific rhetorical functions and the grammar chosen
to express those functions.

86
Thus there is no absolute division to be found between register (word and
sentence) and discourse (above the sentence level) analysis and that both
of these continue to be studied by linguists in order to define text types or
genres.
Eggins and Martin (1997:230) describe Register and Genre Theory
as “linguistic approaches to discourse which seek to theorise how
discourses, or texts, are like and unlike each other, and why.” They go on
to define the steps that need to be taken when applying such a theory.
The first step, they maintain, is to describe the linguistic patterns or
“words and structures” in the texts being analysed. The second step is to
try to explain the linguistic differences found between the texts being
studied. In short, Register and Genre Theory is a theory of functional
variation or how texts coincide or differ one from another for a particular
purpose.
Eggins and Martin (1997:251) define the terms register and genre
as ‘context of situation’ and ‘context of culture’, respectively and they say
these “identify the two main dimensions of variation between texts.”
Register is seen as lower level (bottom-up) realisations of variation and is
constituted by lexical, grammatical and semantic choices. This theory
brings together work from both Register and Discourse Analysis as
described above and will be called Variation Studies (after Biber 1988) in
this thesis as it is applied to the textbooks under study.
Eggins and Martin (1997) explain that genre can be seen in many
different ways. There is the conventional literary model of “types of
literary productions” including short stories, poems and novels. Then
there is the linguistic definition Bakhtin (1986) gives which broadens
genre to include everyday speech and writing with the literary genres.
Genre in linguistics is also defined functionally in terms of its social
purpose. Eggins and Martin (1997:236) summarise this saying “Thus,
different genres are different ways of using language to achieve different

87
culturally established tasks, and texts of different genres are texts which
are achieving different purposes in the cultures”, or what may more
simply be described as text and talk in context.
Similarly, the needs of students and course needs have to be
studied alongside these analyses and continue to be important for the
study of English for Special Purposes and syllabus development. Stubbs
(1996:19) criticises the work carried out on scientific research articles by
Swales (1990) because he failed to relate the linguistic features he found
to a theory of variation in English. Stubbs (ibid.) suggests that any study
of genres “must be located in a description of variation in the language
overall” and that Biber’s work is a good example of how wide a range of
variation there is within academic prose.

2.3.14 Discourse Analysis and Computers

Sedelow and Sedelow (1994:160) acknowledge that the work that

has been carried out using computers for discourse analysis to date are
“purely piecemeal approaches” to “highly restricted discourse domains”.
They advocate the development of a conceptual thesaurus that is based
on associative semantics in order to transcend narrow domains, which
they believe is essential to deal with semantic space. They examine the
lexical cohesive ties described in Halliday and Hasan’s Cohesion in
English (1976) and find (1994:167) that “cohesion does form a major
component of our perception and analysis of discourse”. They also
bemoan the fact that many computer scientists are “rediscovering,
laboriously, many of the relationships already worked out” by Halliday
and Hasan.

The difficulty of applying modern technology to discourse analysis

is quite simply the difficulty of dealing with whole texts. The computer
can handle precise searches and programmes are available for
88
grammatical tagging but the overall organisation of a text is much more
difficult to capture on computer. Biber, Conrad and Reppen (1998:106-
131) put forward a scheme of corpus-based research into discourse
features such as given and new information and the use of discourse
maps of verb tense and voice in research articles in experimental science.
They conclude (1998:131) that although discourse features cannot be
investigated completely automatically, interactive computer programmes
and innovative output formats will be exploited in the future in order to
show patterns of discourse across texts and registers.

2.4 The Needs Analysis Approach

2.4.1 Needs and ESP

Richard West (1994:1) claims that the term ‘analysis of needs’

dates from the 1920’s in India with Michael West and his consideration of
the ‘surrender value’ of learning for secondary level learners. The
connotation that Michael West uses here is that of language
requirements of the students studying English. After this the concept of
‘need’ does not appear again for about fifty years until ESP research
started in 1960. Michael West is more closely associated with register
analysis and his (1953) General Service List of English Words but here he
represents a number of researchers working abroad in universities where
they were trying to find solutions to the problem of what and how to
teach students who required English for their studies, either as a second
or a foreign language. Halliday, McIntosh and Strevens (1964:189) refer
to ‘English for Special Needs’ although their use of the term ‘need’ is also
in terms of special language or register as discussed above. The term
‘need’ only began to take on its more modern connotation of why and for
what purpose students learn a foreign language in the 1970s.

89
2.4.2 The Development of Needs Analysis

West (1993) gives four stages in the development of needs analysis:

Stage 1 in the early 1970’s which focused on English for Occupational

Purposes (EOP) was concerned with target situation analysis and is
exemplified by Richterich (1971/1980) on English for adults, English
Language Teaching Document Unit (1970) on business English, and
Stuart and Lee (1972/85) on English for industry and commerce;

Stage 2 later in the 1970’s which focused on English for Academic

Purposes (EAP) and was also concerned with target situation analysis
and could be exemplified in the work of Jordan and Mackay (1973) and
Mackay (1978);

Stage 3 in the 1980’s ESP and general language teaching which covered
a range of analyses, target situation analysis, deficiency analysis,
strategy analysis, means analysis and language audits as exemplified by
Tarone and Yule (1989), Allwright (1982), Holliday and Cooke (1982),
Allwright and Allwright (1977), and Pilbeam (1979);

Stage 4 in the early 1990’s with integrated/computer-based analyses

and materials selection exemplified by Jones (1991) and Nelson (1993).

West (1993), naturally enough, sees this latter stage with computer-
based analyses as the future of needs analysis. The use of technology in
both analysing and selecting materials is purported to make the syllabus
more appropriate for learners needs.

90
2.4.3 Needs and Syllabus Design

In 1978 Munby published his book Communicative Syllabus Design

which discusses the questions that have to be asked (and answered)
before designing a course. The size and scope of his achievement in this
book have meant that needs analysis has now come to be seen as crucial
in any ESP course design. Munby’s theoretical bases were contemporary
views on the nature of communicative competence, derived principally
from Hymes (1971). In his Communicative Needs Processor the following
parameters are identified as being pertinent to syllabus design:

0.0 Participant
0.1 Identity (Age/Sex/Nationality/Residence)
0.2 Language ((L1/L2/Present level of L2/Other L2s known)
1.0 Purposive Domain
1.1 ESP classification (English for Occupational Purposes (EOP) or English for
Academic Purposes (EAP), if EOP, pre- or post-experience, if EAP, discipline
based or school subject)
1.2 Occupational purpose (specific job or post/central duty/other duties)
1.3 Educational purpose (specific discipline/central area of study/academic
design classification)
2.0 Setting
2.1 Physical setting: spatial (location/country/town/place of work/place of study)
2.2 Physical setting: temporal (point of time/duration/frequency)
2.3 Psychosocial setting (noisy, demanding, culturally different, aesthetic -
unfamiliar)
3.0 Interaction (with others)
3.1 Position (role relationships - dependent on purposive domain e.g. student)
3.2 Role-set (other interlocutors etc.)
3.3 Role-set identity (number/age/sex/nationality of interlocutors thus affecting
role relationship)
3.4 Social relationships (or role relationships e.g. superior-subordinate, peer-
peer, official-member of public, doctor-patient, teacher-learner)
4.0 Instrumentality
4.1 Medium (spoken or written)
4.2 Mode (monologue/dialogue)
4.3 Channel (e.g. face-to-face, text for silent reading, phone)
5.0 Dialect
5.1 Regional (and British English/American English, etc.)
5.2 Social class
5.3 Temporal
91
6.0 Target Level
6.1 Dimensions (size and complexity of utterance/material (text), range and
delicacy of forms and functions, speed and flexibility of communication)
6.2 Conditions (degree of tolerance of: 1. error, 2. repetition, 3. hesitation,
4. stylistic error, 5. reference)
7.0 Communicative Event (i.e. what the learner has to do, either/and productive and
receptive)
7.1 Main (macro activities e.g. waiter serving customer in restaurant, student
in university seminar)
7.2 Other (micro activities e.g. taking down customer’s order or student
introducing a new point)
8.0 Communicative Key (i.e. how the learner does the activities above determined by
1,2,3 - attitude factor).
Table 2.1 Munby’s Communicative Needs Processor

In other words, the communicative needs profile that results from

applying these parameters is a very detailed description of
communicative needs but with no specification of the language items
which will realise these needs. That is to say, communicative needs
profiling is at the pre-language stage and is designed for curriculum
development. This approach was later broadened to include practicalities
and constraints, teaching methods and learning strategies and materials
selection, which were areas which were not considered by Munby. Nor
can Munby’s needs analysis be considered learner-centred, as all the
information to be collected is about the learners and not from the learners
themselves.
Although Hutchinson and Waters (1987:12) recognise that Munby’s
needs analysis was an attempt to put the learner’s needs “at the centre of
the course design process”, they attack Munby’s concept of needs as
being “far too simple” and propose an improved version of their own. They
argue that ‘it is necessary to examine the underlying competence which
the learner must bring to ... the study of any specialised subject’
(Hutchinson and Waters, 1980:178). They propose a classification of
needs which includes:

92
(a) Necessities which are ‘the type of need determined by the demands of
the target situation, that is, what the learner has to know in order to
function effectively in the target situation’ (Hutchinson and Waters,
1987:55). Identifying these necessities is often referred to as target-
situation analysis (see Chambers, 1980).

(b) Lacks. Analysis of what the learner already knows leads to recognition
of the gap which exists between this and the target situation in other
words the ‘learner’s lacks’ (Hutchinson and Waters, 1987:55-56).

(c) Wants. These wants are the learners’ perceived needs or subjective
needs. (Hutchinson and Waters, 1987:57). The learners’ needs
(subjective needs) may be in conflict with the needs analysis that has
been carried out and therefore may be in conflict with the aims of the
course, as determined by those responsible for the course. However, in
such a situation it may be possible to incorporate some of the generally
perceived (subjective) needs of the learners into the course. An example
of this would be to incorporate speaking tasks into courses which are
predominantly designed to aid reading.

(d) Learning Strategies. Hutchinson and Waters (1987:60-2). The

identification of the learning strategies that the learner prefers to use
in order to deal with the target situation may also come into conflict
with the teacher’s identification of suitable strategies.

(e) Constraints. Hutchinson and Waters consider the external factors

which may condition the learning situation such as the resources
available, which would include the length of time the course will run as
well as the materials, aids and methods available. These constraints
are also referred to as means analysis (see Holliday & Cooke, 1982;
Holliday, 1984).

93
 Hutchinson and Waters criteria are highly relevant for the research
carried out here on undergraduate students with (a) the necessities, what
the students need to know, being found from the corpora analyses, and
(b) the lacks, what the students already know, being identified from the
results of the language tests carried out on the undergraduates joining
the first year science and technology courses. Incorporating students
wants would be a more difficult task given the numbers involved and
certain constraints such as the requirements to test all of the students in
the same way at the same time.

Some linguists felt that there was a need to examine language

much less intuitively in order to produce a more objective description. In
many cases the corpora that were built up were designed primarily for
other reasons but soon it became obvious how useful corpora could be in
a number of areas of research into actual language use such as
differences between different text-types and genres, which as I said is
now called variation studies. This led to corpus analysis research as a
means of describing language through examination of actual examples of
use instead of invented examples. This requirement of only employing
actual language is also considered crucial in modern discourse analysis.

2.5 The Corpus Analysis Approach

Modern computers have removed many of the barriers to

comprehensive corpus research and have enabled significant and far-
reaching specific studies to take place. These studies are not restricted to
English, although many of the early corpora developed were based on
both American and British English (see later the Brown and LOB
corpora). There is for example the French Trésor de la Langue Française
which is largely historical and lexicographic in conception.

94
 Sinclair (1991:1) suggests that traditional linguistics had been
limited by the amount that one person could experience and remember
and he equates the situation with that of the physical sciences 250 years
before. Halliday (1993:7) sees the start of corpus-based linguistics as
laying the foundations for a quantitative and qualitative breakthrough in
understanding linguistic systems and of this having started in the 1960’s
with Randolph Quirk in Britain and Freeman Twaddell in the United
States.

2.5.1 English Corpora Development

The study of computer corpus data in English was started in order

to find out ‘the true facts of English grammar’ because it is more accurate
than old manual methods and also more likely to take into account all
the unpredictable features which occur naturally in speech and writing.
Sinclair and Francis (1994:191) suggest that:

“Corpus data provides us with incontrovertible evidence about how people

use language. ... lists are both a continuation of a tradition, and an
innovation. ... the tradition was to observe what people did and record it in
a reference book. In recent years, the language teaching trade has lost the
ability to appreciate the prime importance of this tradition, and has relied
rather too heavily on intuition.”

Originally, the idea of using computer corpora in this way was

derided as being a waste of time and money because it was thought that
traditional methods could provide suitable examples of English grammar
drawn from native speakers’ insights into how the language operates.
Biber, Conrad and Rippon (1994:169) point out, however, that “...corpus-
based analyses frequently show that earlier conclusions based on
intuitions are inadequate or incorrect.”. It should also be pointed out that
95
intuitive decisions are even more questionable in scientific and
technological language, as no-one is a native speaker of ‘scientific
language’. Meijs (1992:146) says that while the great traditional
grammarians such as Poutsma, Kruisinga and Jesperson were ardent
data-gatherers, their work could be criticised on the grounds that they
had a bias for the unusual and irregular which made them overlook the
larger generalities in their search for “interesting” phenomena. Stubbs
(1996) compares Chomskyan principles with neo-Firthian principles. He
(1996:24) argues that the Chomskyan view that “linguistics is a branch of
cognitive psychology”, that “it can be based on intuitive data and isolated
sentences, that corpus data are unrevealing”, and that “the study of
language in use is essentially uninteresting” is in direct contrast with
neo-Firthian views as represented by Sinclair (1991:4) who argues that
corpus data provide “a quality of evidence that has not been available
before” and only considers actual, authentic, attested data to be of
interest. Meijs (ibid.) claims that computer corpus studies can develop “a
more balanced view of the ‘spread’ of linguistic phenomena - lexical as
well as syntactic ones”.11
The use of computer-based text corpora has become increasingly
important for research into natural language processing, lexicography
and descriptive linguistics, issues relating to corpus design have also
assumed central importance. Biber (1994:179) says that there are two
main considerations “1) the size of the corpus (including length and
number of text samples), and 2) the range of categories (or registers) that
samples are selected from.” Biber explains that corpus designs can differ
in whether they are bounded or unbounded and therefore static versus
dynamic; whether they are richly encoded or minimally encoded (that is,
whether there is grammatical tagging, phonological or prosodic encoding

11
Meij’s ‘spread’ could be equated with the computer corpus concept of ‘range’ of an item in frequency
studies.

96
or some kind of social characteristic tagging of the participants and
situational tagging); whether it contains complete texts or samples from
texts; and the selection of texts may be either made by convenience
versus purposeful versus random within strata versus proportional
random.

2.5.2 Corpora Use

Much of the work that has been undertaken is on general language

use and has attempted to find generalisations about normal usage rather
than the specific language associated with English for Special Purposes.
Leech (1993:10) identifies the use of computer corpora for the following, a
list which he claims is “by no means complete”:

• linguistic theory: improving models of language generally

• computational linguistics: natural language processing by computer; machine
translation; etc.
• grammar: syntax, morphology, automatic parsing
• dictionaries: lexicography, lexicology, word-formation
• the study of meaning: semantics and pragmatics
• discourse analysis and conversational analysis
• language variation: spoken and written language; language and gender; general studies
of style; dialect
• speech technology: automatic speech synthesis and speech recognition
• speech science: phonology, phonetics, stress, intonation, etc.
• historical studies of language
• child language acquisition, psycholinguistics
• applied linguistics: language learning, language testing, etc.
• orthography: punctuation, spelling

97
Leech does not, for example, consider translation studies which is an
enormous area of its own with very specific views on language although
this could be subsumed under computational linguistics in general and
is perhaps hinted at in machine translation. Nor does he make special
mention of collocations, an area of language learning and teaching that
has become extremely important in recent years (Sinclair 1991, Tribble
and Jones 1990). Collocations are now seen to be the building blocks of
language and can be used for vocabulary management, to disambiguate
similar terms and formulate or check hypotheses about language use, to
help learners to understand texts, for self-access outside the classroom
and to provide teachers with suitable teaching materials. However, this
list does show some of the areas within which computer corpora can be
applied to the study of the English of science and technology in applied
linguistic research and language teaching.
Biber (1994:180) suggests that recent debate centres around
whether to use large corpora as opposed to what is known as “balanced”
corpora (that is, made up of a number of similar sized texts possibly from
a wide range of registers) for the design of general purpose corpora. He
argues (1994:180) that “it is important to address the question of
whether the varieties represented match the intended uses of a corpus”
and he claims that studies of a single sub-language are “legitimately
based on corpora representing only that variety”. This is the view that
this study takes about corpora; they must represent the variety of
English that the students are expected to come into contact with and
need to understand for their studies in science and technology.
Biber (1994:11) lists the following areas of study in linguistics that
corpora can help with: “individual words, grammatical features, men’s
and women’s language, children’s acquisition of language, author style,
register patterns” and goes on to suggest that dialect and register
patterns could be investigated for sociolinguistic fields when looking at

98
the “complex co-occurrence patterns among features in different
registers” which would be difficult to do without recourse to computers
on a large scale. He also (1994:12) mentions the study of styles across
historical periods which could provide the opportunity of investigating the
development of registers over time and emphasises the role of corpora in
educational linguistics. With respect to the latter he says that “large-scale
studies of use are helpful in designing effective materials and activities
for classroom and work-place training, allowing us to help students with
the language that is actually used in different target settings.” He also
recommends corpora use in language testing, that is, “making tests
which conform to the actual language that students will be using on a
regular basis”. These conclusions form the basis of the working
presuppositions of this study.
The preliminary tests that were designed for the undergraduates,
described in Chapter 4 went some way towards this goal of conforming to
what was seen as the target language that the students would be coming
into contact with in science and technology. The tested items were from
the materials that were to be taught in the discipline. They were not
however, derived from corpora developed for the purpose, which, in the
light of modern computer corpus methods is a weakness of the testing.
The reason for this was that the testing had already got underway before
the corpora used in this study had been developed but now that they are
available there is no reason why they should not be used for this purpose
in the future.

2.5.3 The Birmingham Corpus COBUILD

The corpus work of Sinclair, started in 1960 in Birmingham, was

primarily concerned with lexicography and the production of the
COBUILD Dictionary. This dictionary was published in 1987 and was

99
based on what was then considered a huge corpus of 7.3 million words of
written and a smaller corpus of about 1 million words of spoken
language. The ‘main’ corpus was started in 1960 and subsequently
smaller ‘side’ corpora were developed (notably the Bank of English and a
corpus especially prepared for Teaching English as a Foreign Language
(TEFL) textbook writing (see Willis, 1989). Sinclair and Jones (1974)
report that “The first corpus, in 1961, was a mere 135,000 words”. This
reflects the changes that have taken place with regard to the gathering of
data. Initially every text had to be transcribed onto computer manually
and the original computer programs for handling the texts had to be
developed. Later text which had already been transcribed on computer
through word processing became available and later still the use of
optical scanners (usually known as Optical Character Recognition or
OCRs) simplified the transcription of text and speeded up its conversion
into electronic data.
Sinclair (1987:2) describes the criteria on which the ‘main’ 7.3
million word Cobuild corpus was developed to be relevant “for the needs
of an international user” and which the team defined as the following:
- written and spoken modes
- broadly general, rather than technical, language
- current usage, from 1960, and preferably very recent
- “naturally occurring” text, not drama
- prose, including fiction and excluding poetry
- adult language, 16 years or over
- ‘standard’ English, no regional dialects
- predominantly British English, with some American and other varieties.

Sinclair (1987:3) also describes the balance that was given to

the components but gives no clear definition of why this should be so,
he merely suggests that these were chosen for “different reasons”:

100
 book authorship - 75% male: 25% female
English language variety - 70% British: 20% American: 5% Other
language mode - 75% writing: 25% speech

The development of criteria for the Birmingham corpus was

grounded on clear principles as those involved describe (cf. Hoey 1996)
but some areas still need explaining like those given above. It is
questionable whether this balance would be the only suitable choice for
“the needs of an international user” as described by Sinclair.
The spoken corpus was initially a problem because of the difficulty
of obtaining permission to use tape recorded conversations where the
participants did not know that they were being recorded. The university
prohibited such activities which led to structured dialogues and BBC
programme material being used. The latter material can be criticised as it
should be considered scripted or ‘prepared’ and, therefore, the spoken
corpus could be seen as not reflecting ‘natural’ speech or conversation,
thereby destroying the main premise of collecting data of actual (real-
time) occurrences of language and making it prone to be considered
(more) examples of written English read aloud.
These early problems have largely been overcome and the current
corpus, known as the Bank of English, now runs to hundreds of millions
of words but is constantly growing. The working corpus recently (July,
1998) had 329 million words of modern English text. The written texts
include fiction and non-fiction books, newspapers, guides, magazines,
brochures, letters and leaflets. The 20 million words of transcribed
natural speech in the corpus include everyday casual conversation, radio
broadcasts, meetings, interviews and discussions. Most of the texts originate from
after 1990 and are designed, according to the Collins COBUILD website (1998:1),

101
“to provide objective evidence about the English that most people read, write,
speak and hear every day of their lives”.

2.5.4 The Lancaster-Oslo-Bergen Corpus

The LOB (Lancaster-Oslo-Bergen) Corpus of British English has an

even sample size of around 2,000 words all taken from printed sources
published in 1961 and totalling a million words (see Johansson et al.
1978, Johansson 1982 and Hofland and Johansson 1986), compiled in
the 1970’s. There are 500 text samples taken from fifteen genres: press
reportage, editorials, press reviews, religion, popular lore, skills and
hobbies, biographies and essays, official documents, learned writings,
fiction (including general, mystery, adventure, science, and romance),
and humour. The total corpus size is approximately one million words of
running text. The divisions of the corpus into genres is conducted largely
on intuitive criteria. Sinclair (1991:19) criticises this because “a corpus
which does not reflect the size and shape of the document from which it
is drawn is in danger of being seen as a collection of fragments where
only small-scale patterns are accessible.” The breakdown into the
respective categories is as follows:

Category Number of Texts Approx. number of words

Press reportage 44 88,000
Editorials 27 54,000
Press reviews 17 34,000
Religion 17 34,000
Skills and Hobbies 36 72,000
Popular lore 48 96,000
Biographies and essays 75 150,000
Official documents 30 60,000

102
Academic prose 80 160,000
General fiction 29 50,000
Mystery fiction 24 48,000
Science fiction 6 12,000
Adventure fiction 29 58,000
Romantic fiction 29 58,000
Humor 9 18,000
TOTAL 500 1,000,000
Table 2.2 Texts, Categories and Numbers of Words in the LOB Corpus

The LOB Corpus is tagged and part of the LOB known as the
Lancaster Parsed Corpus contains 133,000 words that have been
syntactically analysed.
There is also now the Freiburg corpus with approximately 1 million
words of British English, parallel to the LOB corpus, but compiled from
material published in 1991. The fact that corpora are seen to be
becoming dated means that their authority to describe modern English
usage is also diminished and so many more of this type of up-to-date
corpora are being prepared to keep abreast of changes that are
constantly taking place in language usage. These more modern corpora,
when produced using similar criteria, can be used for diachronic and
other comparative studies. The other reason that more up-to-date
corpora are being produced is that the techniques now available and the
research carried out on machine readable or electronic texts has brought
some of the original criteria into question. The insights gained from such
research now implies that more modern corpora can be obtained in many
more different states of tagging depending on the purpose to which they
are to be put. Biber’s research which forms the basis of this study drew
on some of the earlier LOB texts.

103
2.5.5 The Brown Corpus

The Brown University corpus of written American English (see

Francis and Kucera 1982) is one of the oldest of the large scale corpora.
It was started in 1961. It consists of short extracts of many genera, for
research purposes. The LOB corpus is a replica of the Brown
(1964/1979) corpus so that parallel text samples can be compared
between British and American English. This corpus is tagged. There is
also now a fully tagged subset of the Brown corpus known as the
SUSANNE Corpus, which contains approximately 128,000 fully tagged
words in 64 texts each about 2,000 words long from four genres of the
Brown Corpus; A: press reportage, G: belle lettres, biography, memoirs,
J: learned (mainly scientific and technical writing), N: adventure and
Western fiction.
Leitner (1993) criticises the LOB and the Brown corpora because
textbooks are unrepresented in them and he contends (1993:81) that
“textbooks are a major medium for communication”. Goethals, Engels
and Leenders (1990:237) find that what they describe as “the journalese
style that dominates the Brown and the LOB corpora” causes distortion
in their work on the Leuven English Teaching vocabulary list. This bias
can be confirmed by an examination of the texts included in some of the
sections other than those specifically designated as press reportage,
editorial or reviews. For example, the skills and hobbies section contains
many texts taken from magazines such as ‘High Fidelity’, ‘Dog World’ and
‘Hot Rod’. Similarly, the Popular Lore section contains articles from many
magazines including ‘Vogue’, ‘Family Circle Magazine’ and ‘National
Geographic’. Even Belles Lettres contains texts from ‘The Saturday
Evening Post’, ‘The New York Times Magazine’ and many different,
regional ‘Quarterlies’ and ‘Reviews’. The source of texts has obvious
implications for style and discourse features as Goethals, Engels and

104
Leenders have found. For the purposes of the research described here
these failings make the use of such corpora inappropriate. Neither the
absence of textbooks nor the presence of an overwhelming amount of
journalese is suitable for analysis of the language that undergraduates of
science and technology need to confront and is therefore not suitable for
the purposes of this study.
Furthermore, Minugh (1997:68), despite recognising that the
Brown and LOB corpora were “a revolution in their time”, describes the
difficulty of using such corpora for searches for neologisms because of
their date of development. Minugh (1997) recommends the use of British
and American Newspaper CD-ROMs for this sort of search. In other
words, these corpora are also becoming dated and are therefore not
suitable for finding representatives of colloquial or modern language
terminology or coinings. This limitation is particularly relevant for those
conducting research into speech and current news because of the ability
to change quickly and reflect fads and fashions. Some of those changes
will become part of the language but others will disappear almost as
quickly as they came. This is the heart of the problem that dictionaries
such as the Oxford have every time a new edition is published. Terms
which are regarded as fashionable or corruptions are often decried by
readers and reviewers as having no place in such an established
reference work on the English language.

2.5.6 The London-Lund Corpus

The London-Lund Corpus of Spoken English (Svartvik and Quirk

(1980), Johansson (1982) is a collection of 87 spoken British English
texts12 of about 5,000 words each. The total corpus contains
approximately 500,000 words of different genera. It is divided into half

12
Text here means a communicative event

105
spoken and half written material. Six major speech situations are
represented: private conversations, public conversations (including
interviews and panel discussions), telephone conversations, radio
broadcasts, spontaneous speeches, and prepared speeches divided up in
the following way:

Category Number of Texts Approx. number of words

Face-to-face conversations or 65 235,000
discussions
Telephone Conversation 110 60,000
Public conversations, discussions, 20 85,000
interviews
Spontaneous commentary (radio 20 55,000
broadcasts)
Spontaneous oration 12 30,000
Prepared oration 12 35,000
TOTAL 239 500,000
Table 2.3 Texts, Categories and Numbers of Words in the London-Lund Corpus

This corpus started as Randolph Quirk’s (1980) Survey of English

Usage in a traditional paper filing system but was converted into
electronic form by Jan Svartvik (1980). The corpus is tagged, that is to
say, the texts are annotated grammatically. The practice of tagging is
criticised by Sinclair (1991) as, in his opinion, this makes a corpus less
useful. However, Biber, Conrad and Reppen (1998:31) argue that tagging
makes a corpus more useful for particular kinds of searches. They give
the example of automatic frequency counts for each separate
grammatical word of “deal” in a corpus, that is “deal” as a singular noun,
proper name, verb, plural noun, present participle, past tense verb and
past participle which they explain would help to show which of these is
found in which registers and, therefore, the ways in which words are
used (and the different meanings they show) in different registers.
However, they do accept (1998:59) the need for hand-editing in order to
106
ensure that some searches have not included items incorrectly (in this
case words ending in -ion or -ity which are not in fact nominalizations,
like nation and city). They also (1998:67) discuss the difficulties of
comparing analyses when different criteria may have been used to
describe nouns and verbs. The different studies may, for example, have
included pronouns in the nouns category or not and auxiliary verbs may
or may not have been included in the verb count. Similarly with
automatic grammatical tagging, Biber et al (1998:73) recommend using
interactive techniques and balancing the results found in order to correct
errors made by the tagger even though this process is extremely time-
consuming. The problems discussed above are particularly prevalent with
automatic tagging of speech corpora because of the interjections that
occur.
There is obviously a need here to have some kind of international
agreement on tagging procedures or at the very least to insist that
research carried out is accompanied by very clear descriptions of what
was included in each of the categories that were used. The fact that
English uses a number of different schemes to describe grammar,
depending on the purpose or audience the work is for, means that there
is ample space for conflicting criteria to be used. Sinclair’s idea of
keeping corpora as simple and ‘pure’ as possible until these areas of
grammatical tagging have been clarified is probably the safest and is the
position that will be adopted with the corpora used in this study.

2.5.7 The British National Corpus

The British National Corpus (BNC) was started in 1991 at Oxford

University and is a 90 million word collection of samples of written and
10 million words of spoken language from a wide range of sources,
designed to represent a wide cross-section of current British English both

107
spoken and written. Leech (1993:13) gives the following information on
the composition of this corpus:

Written texts (90 million words)

Selection features
- Informative/imaginative writing
- Subject field
- Date
- Genre
- Level

Spoken texts (10 million words)

- demographic sampling (50%)
- sampling by discourse type (50%)

Written Component: Informative

Primary subject fields Level
Natural and pure science (5%) Specialist (30%)
Social and community (15%) Lay (50%)
Commerce and Finance (10%) Popular (20%)
Belief and thought (5%)
Applied science (5%) Date
World Affairs (15%) 1975-present
Arts (10%)
Leisure (10%)

Genre
Books (55-65%)
Periodicals (20-30%)
Miscellaneous (published) (5-10%)
Miscellaneous (unpublished) (5-10%)
To be spoken (2-7%)

Written Component: Imaginative

(20-30%)
Level Date
Literary (33%) 1960-1974 (25%)
Middle (33%) 1975-present (75%)
Popular (33%)

Table 2.4 Composition of the British National Corpus

108
The spoken, face-to-face conversation corpus is as follows:

Category Number of Texts Approx. number of words

Face-to-face conversation 160 4,000,000
Table 2.5 Conversation in the British National Corpus

There is information for research purposes on the gender,

occupational group, social background and age of the speakers
(informants) taking part in the collection and it has been used as the
basis for the Longman Dictionary of Contemporary English. These corpora
can be consulted in order to obtain other data such as the differences in
spoken language use between gender and age groups, which has entailed
a much more rigorous description of the material held in the corpora to
allow this level of specificity. Leech (1993:10) points out that corpora of
the size and nature of the BNC ‘are often too expensive in time and effort
to be built without commercial or industrial help’ and describes how this
corpus was funded by the British government (hence it is British
English), three major British publishers (Oxford University Press,
Longman and W&R Chambers) together with the British Library and the
Universities of Oxford and Lancaster.

2.5.8 The Longman/Lancaster Corpus

There is also the Longman/Lancaster Corpus containing American,

British and other varieties of written English. It consists of 30 million
words which is described as ‘representative’, by which is meant that ‘a
full range’ of variation of the language is included. What a “full range” of
language is provokes some debate. This question has been answered by
the British National Corpus developers as “the English which most people
read, write, speak and hear every day of their lives”. A new category has

109
emerged because of this, that of “ephemera” which includes any material
that people come into contact with unintentionally, such as unsolicited
mail and advertising.
The number of texts and number of words contained in the categories
Academic Prose and Fiction are as follows:

Category Number of Texts Approx. number of words

Academic Prose 98 2,700,000
Fiction 144 3,000,000
TOTAL 242 5,700,000
Table 2.6 Number of words and texts for Academic Prose and Fiction in the
Longman/Lancaster Corpus

The samples are taken from many registers from the early 1900s to
the 1980s. It can be seen that this corpus also has limitations for a
description of either general English or for analysis of specific varieties of
English usage. It suffers from a lack of balance to provide what is
described as ‘general English’ by the Bank of English criteria mentioned
above (section 2.5.3). It also suffers from having ‘text fragments’ which
Sinclair (1991) regards as a failing of many corpora. It also covers too
wide a period of time for much research on either modern usage or for
diachronic study purposes.

2.5.9 Other Corpora

Many projects on specific issues that researchers feel are not or are
underrepresented in the established large scale projects described in
more detail above are taking place in universities around the world. A
short summary of some of the main areas that these cover is given below
to demonstrate the trends in recent corpora studies.

110
 An Australian corpus (ACE) produced at Macquarie University, New
South Wales and an International Corpus of English (ICE) produced at
University College London (http://www.ucl.ac.uk/~ucleseu/design.html)
have also recently been developed to address other types of Englishes in
the world. ACE contains one million words of Australian English
compiled along the same lines as the Brown Corpus for purposes of
comparison. ICE contains one million words from the English of
Australia, Canada, East Africa, Hong Kong, India, New Zealand, Jamaica,
Nigeria, Singapore and the Philippines. The Melbourne-Surrey Corpus
has 100,000 words from Australian newspapers.
The Kohlapur corpus contains 1 million words of written Indian
English from 1987. It uses the same categories as the Brown Corpus and
LOB Corpus.
A corpus of spoken American English (CSAE) is being constructed
at the University of California which eventually hopes to contain one
million words.
The Northern Ireland Transcribed Corpus has about 400,000 words
of spoken material from 42 locations and over three age groups.
The CHILDES Project (http://poppy.psy.cmu.edu/childes/database.html) is
developing a corpus of children’s spoken and written language. There is
also the Polytechnic of Wales (POW) corpus of 61,000 words of children’s
spoken language which has been parsed using Hallidayian Systemic-
Functional Grammar.
The increase in the number of corpora and such corpora as those
on language development will surely have an influence on teaching and
learning as they show what actually takes place rather than what some
small scale studies have suggested is the case in both language
acquisition and language diversity.

111
2.5.10 EFL Student Corpora

In addition to these, there are even specific corpora for English

language learners such as the International Corpus of Learner English
(ICLE) and the Longman Learners’ Corpus drawn from EFL learners
around the world and associated with the International Corpus of English
which was directed by the late Sidney Greenbaum at University College
London. This ICLE corpus is expected to reactivate certain research areas
such as error analysis and could very probably lead to analyses of the
relative frequency of certain features in learner English as compared with
those of native speakers.
The Hong Kong University of Science and Technology also has a
Learner Corpus with approximately 6 million words of written
undergraduate assignments and “A” level Use of English scripts from the
Hong Kong Examination Authority. This corpus is still growing.
These corpora are beginning to increase our understanding of
learner English so that insight can be gained into learning stages and
strategies, perhaps with comparative studies on language acquisition.
Biber, Conrad and Reppen (1998:172-202) describe some studies that
have been carried out but recognise the fact that computer corpus-based
studies have not hitherto normally been used in these areas.

2.5.11 Specialised Corpora

Although clearly of less interest to this study, more specialised

corpora such as, the Helsinki Diachronic Corpus of English (Kytö 1991)
and the ARCHER corpus (Biber et al. 1994) have been designed to cover
the development of English registers over several historical periods. The
Helsinki Corpus covers the periods from Old English to Early Modern
English, while the ARCHER corpus includes texts from 1650 to the

112
present. These variation studies across time as opposed to across genres
are developments which seem to be harking back to some of the other
traditional (now computer assisted) studies of variation in old
manuscripts.

2.5.12 Concordances

Many programmes have now become available for small personal

computers that allow the teacher and students to conduct research into
their own corpora either to produce accurate materials or for reference
purposes. Concordances, programmes that display the immediate
environment of particular lexical or grammatical items, are available that
will accurately reflect real English use and which can be used by both
students and teachers. Moreover, these programmes can be customised
to reflect the different genres to be studied by the simple means of
choosing specific text types as the basis for the concordance. Teachers
and students can then consult such programmes to decide what is the
correct or natural choice of, for example, a preposition after a particular
word, such as different from, by examining all the instances of different. It
is in this specific area that recent work has been carried out by Jones
(1991) and Nelson (1993) as mentioned earlier (see section 2.4.2).
Published CD-ROM’s on collocations, such as by Collins, are also
beginning to appear on the market. These products, together with specific
material on aspects of English grammar like phrasal verbs and reported
speech, have been developed from research on corpus material. They
contain examples taken from the corpus and not invented ones. Many of
these come from the Birmingham corpus, which can be consulted by
both teachers and students. As with any information, the material
extracted is only as good as the corpus on which it is based. Some of the
examples, whilst claiming to represent general English usage, can be

113
seen to be from novels from earlier periods like those of Jane Austen in
the early nineteenth century. Such examples can be regarded as dated
and often ‘unusual’ rather than reflections of modern-day English usage.
The biggest criticism of many of the first corpora produced is precisely
this, that they have already become dated and cannot be seen to be
representative of modern English usage any longer. They are already
caught in the trap of ‘historical’ rather than ‘current’ usage.

2.5.13 Undergraduate Textbook Corpora

Corpora built up from undergraduate textbooks in chemistry and

physics produced using an OCR are still a prerequisite for specific study
of how these texts differ from or conform to other discourse types, despite
the wealth of material described above, for the simple reason that these
textbooks represent a much more specific genre than any of those
available either for research purposes or commercially, especially as
these textbooks for undergraduates are often in American English. As
was mentioned earlier (2.5.6) the American Brown corpus was criticised
by Leitner (1993) for not containing any textbook material. In this case,
commercially available CD-ROM texts can serve a useful purpose for
comparison and contrast with more general instructional texts in order to
highlight the differences between the general and the specific nature of
the texts of a similar level under study here. Stubbs (1996:5) argues that
there is a “need to analyse not only short text fragments, but also whole
long texts; and the need for the stylistic analysis of individual texts to be
based on comparisons with other texts and with corpus data which
represent (however imperfectly) the language.”. Thus, I have incorporated
it into my approach.

114
 This dissertation will take up three major lines of research from the
register analysis, discourse analysis and corpus analysis mentioned
above which are essential to syllabus development.
First a register analysis will be carried out on the physics and
chemistry books from the students bibliographies, as register analysis
can be applied to syllabus design following Jones’ orientation (1991),
using frequency counts to identify what is lacking in any syllabus or
materials for specific learners. Consideration of cognates will be made in
order to fine tune these lists even further and to anticipate areas of
difficulty for Portuguese native speaker students.
This will be compared with a CD-ROM multimedia encyclopaedia in
order to bring out similarities and differences between texts that are of
the same academic level, according to Huddlestone (1971) and Swales
(1985), and which will serve to reflect the moves in education towards the
use of this kind of technological resource for both student and teacher-
directed learning (see Guillot and Kenning 1995:365). Work with
interactive and multimedia resources such as those available on CD-
ROM and through the Internet are seen as being increasingly important
in education as discussed in the introduction to this study. This
comparison will also provide information on the range of the items, so
that the relevant context (and, therefore, specific use and meaning) of the
lexis can also be determined.
The corpora from the physics and chemistry textbooks will be
explored using Biber’s (1988) methodology for variation studies, in order
to highlight the ways in which these conform to and differ from both
academic prose and general language use he classifies it. Biber’s
methodology is explicitly defined so that it is possible to build this study
on his work in an accurate and principled manner. This is deemed to be
a prerequisite of any research in corpora studies so that a precise

115
description of the criteria used to produce the data is available which can
thus be evaluated in the light of the purpose to which it is to be put.
The students’ language needs will also be ascertained by using the
results obtained from tests on entering the university to determine the
strengths and weaknesses which will need to be addressed by any
syllabus designed for these students. The results of the tests are
classified into grammatical categories that correspond to those used on
the corpora as far as possible and test items are exemplified to provide
clearer description especially for those tests that took place before the
advent of this study. By comparing and contrasting these categories it is
possible to reach some conclusions about the areas that need specific
attention in the syllabus designed for these students.
Finally the results of what can be seen as a data-driven description
will be brought together to suggest what should to be included in any
syllabus adopted for these students. At this point the research findings
from other corpus-based studies have to be taken into account in both
grading of material through core or key patterns of usage. The
commonest forms of language use and the combinations these typically
form or core patterns in these textbooks must be matched with the
abilities shown by the undergraduates with them on the tests they have
taken. Nevertheless, syllabus design calls into question many other
methodological aspects which must be addressed. What can feasibly be
achieved, despite the inherent constraints, is one of the major
considerations here. It is essential that innovation of the kind mentioned
above in terms of modern technology is incorporated into the syllabus
and more of the same kind of teaching/learning is not carried out for
these very specific students with specific goals and requirements. The
wider educational implications of innovation in the syllabus proposed will
be addressed.

116
Chapter 3 Research Methodology
 Chapter 3

Research Methodology

As described in Chapter Two, much work has been carried out this
century and in particular in the last thirty years to try to define exactly
what makes different styles of English different. Lexis, syntax, pragmatics
and discourse features have all been studied in order to discover
differences and many claims have been made, some on rather slim
evidence. For example Tarone et al’s study (1981), although very
professional, was based on only two Astrophysics articles which were eight
pages and seven pages long respectively (see section 2.2.3 for further
discussion of this point). Swales (1985) attributes this state of affairs as
existing because ESP ‘practioners’ that is, teachers who also produce
materials , design courses and conduct research, are usually working in
isolation and do not often look back to the work that has gone before, nor
do they learn from work that is being conducted in parallel to their own
and which might usefully contribute to their work. One thing has become
increasingly apparent and that is that each learning context needs to be
studied in order to provide an accurate picture if the results of such
research work are to have practical applications. Nevertheless, this
specific work must be related to other work in the field.
Halliday (1993:124) says:
“There are practical reasons for analyzing scientific texts. The most obvious is
educational: Students of all ages may find them hard to read, and we know
from various research reports that, in English at least, the difficulty is largely

116
 a linguistic one. So if we want to do something about it we need to understand
how the language of these texts is organized.”

It is especially important that the materials from which the research

data is taken reflect the target material for the students who are to be
taught or else inappropriate conclusions will be drawn. Many studies are
weakened by the fact that the materials used for the research do not
reflect only and uniquely that which is the target material of the learner.
This is very often because of the difficulty in obtaining sufficient material
of the right type. The dangers are that material can easily slip either into a
much more general English category or into much more specific
specialised English, like that of the post-graduate specialist, when the
learning context is for undergraduate students. In this study the first year
undergraduates are the focus of attention. These students have much
more of an end of secondary education profile, which is a difference of five-
years from the post graduate in both science subject specificity and
maturity.
At issue here are the undergraduate textbooks contained in the
students’ bibliographies which will be analysed in order to help the
students to be more successful in their immediate studies and to prepare
them for their futures. Rosenthal (1996:31) suggests that the English
language that causes students (in American higher education) problems
with their science studies is not the vocabulary of science but rather in
comprehending typical college science textbooks whose readability levels
are beyond those attained by English as a Second Language students.
Many of the textbooks recommended for further reading for
undergraduates in universities here in Portugal are either in English or in
translation. The translated books are all too often in Brazilian Portuguese
because Portugal is considered too small a market for educational

117
materials of this level, which suggests unfortunately that this situation is
likely to continue for the foreseeable future.
Researchers often recognise that different styles of English are more
prevalent at different academic levels. The difference between an
undergraduate and a post-graduate science student, for example, would
suggest widely different text types (from textbooks teaching the subject
matter of the course to journals reporting the latest research in very
specific branches of science) and, therefore, styles of English. Similarly,
some English for Science and Technology (EST) practitioners believe in
adopting much more popular and accessible texts which would also bring
with them a considerable difference in style and content than the average
science textbook. One example of this difference in style is shown by
research carried out by Darian (1981) into the manner in which
definitions are handled in such magazines as Popular Science or Time
magazine and the Journal of Astrophysics. Darian finds, not surprisingly,
that definitions are handled differently in popular magazines from those
used in specialist journals, and that these are different again from those
used in textbooks used to teach the subject. Although it can be argued
that the students may be more motivated by certain types of (more
popular) materials, these are not considered to be a suitable basis for an
analysis of syllabus design for tertiary level students. The assumption that
will be made in this analysis is that if students are taught to cope with the
kinds of scientific texts that appear in their undergraduate bibliographies,
they will be better able to cope later on, whether it be with the literature of
their specialisation (where incidentally lexical density has not been found
to be a barrier to the specialist) or in other professional outcomes of their
courses1.

1
such as teaching cf. Arroteia, Jorge Carvalho; Martins, António Maria (1997) Inserção Profissional do
Diplomados pela Universidade de Aveiro: Trajectórias Academicas e Profissionais, Aveiro: Universidade de
Aveiro.
118
3.1 Frequency and Range List

In order to develop a syllabus for students it is necessary to select

from the vast number of words and structures that exist (and which no-
one is capable of learning in their entirety) those which will be most
suitable for each particular student or class of students. This is as true at
the more advanced level as at the elementary stages of learning. One
method of doing this is by means of frequency counts, as described before
in register analysis (Pp. 50). Although it is claimed that the most frequent
words are few in number and the 1,000 most frequently used words make
up about 95 per cent of the total number of words in any randomly chosen
corpus of language, frequency counts can show which word or structure
to choose over another and can throw up evidence of omission of items
that should be included in teaching materials. Frequency counts do not,
however, mean that certain words or structures can automatically be
excluded from teaching materials. It would be ludicrous to argue that
certain language that is appropriate for teaching purposes should be
excluded on the grounds that it did not appear in a frequency count. For
example it may be found that certain days of the week or months of the
year are not found or are found infrequently in the corpora. This does not
imply that only certain days of the week and months of the year should be
taught to students. This is often referred to as opportunism, by which is
meant that some things are available in the immediate situation or are felt
by the teacher to be useful to the students. On the other hand, without
empirical studies it may not be immediately obvious which structures and
items in textbooks designed for undergraduate students of science and
technology are to be preferred.
The frequency counts described below go further than just providing
evidence for the language to be included in teaching materials, they also
show how widely used a word is across texts (its range), increasing its

119
usefulness for teaching purposes and further specifying appropriate use.
Contrasting the three corpora will also give information about coverage,
that is, the number of things that can be expressed by any given item.
Coverage and range together will provide clearer evidence for which items
to include in the syllabus. Furthermore, examination of the frequency lists
allows prediction of areas of difficulty for students whose first language is
Portuguese.

3.1.1 Contrastive Analysis

Lado (1957) puts forward contrastive analysis as a form of language

description across two languages which is particularly applicable to
syllabus design and to the evaluation of the items which would lead to
difficulties for students of particular languages. It was not meant to be a
new method of teaching but to aid curriculum development and diagnose
learning problems so that suitable materials could be prepared for
teaching. Lado was influential because he set out procedures that could
be applied for the comparison between languages. The detailed work
remained to be done however.
Contrastive analysis was not found to be the universal panacea that
it was hoped to be in language teaching, as linguistic theories of
transformational generative grammar came into being in the late 1960s
and ousted structuralism as the dominant theory underpinning teaching
practices. Error analysis has also shown that the whole subject is
considerably more complex than was first supposed. However, contrastive
analysis has been reappraised (initially by Di Pietro 1968, 1971 and with
Danesi in 1990 who brought contrastive analysis in line with
transformational generative grammar but even more recently through
reappraisal of interlanguage by Selinker 1991) and contrastive techniques
can be applied to lexical corpora to highlight and predict areas that,

120
because of similar word formation or shared Latin roots, could be easier
(cognates) or, because of different roots, more difficult for Portuguese
undergraduates, or examples of false friends. For example the word
“abnormal” and its Portuguese equivalent anormal are sufficiently close to
suggest that positive transfer could take place and the students’ ‘guess’ or
semantic prediction would probably be accurate. However, words like
“able” and its Portuguese equivalent capaz are considered to be difficult,
although an alternative hábil could be used in some circumstances and
would be closer to the English form. The latter would be more accessible
provided that the students recognised the similar pronunciation of the two
words rather than their orthographic form. This contrastive analysis aims
to predict the learnability (Mackey 1965) of the language found in the
corpora and is incorporated into this study through examination of the
frequency lists from the corpora described in more detail in Chapter 5 in
order to identify cognates and thereby identify the areas of difficulty for
students.
More recently internationalisms have appeared where the same (or
very similar word) is being used in many countries. An example of this
might be computer jargon like software/hardware which are used in many
languages and have caused the European Union to fund projects to
produce terminology banks in various areas including that of information
technology.

3.1.2 Context

The next step in the development of teaching materials, once the

identifying and contrasting stages have been completed, is to provide a
suitable and typical context for the items regarded as important (and often
posing a certain level of difficulty for students). Bright and McGregor
(1970:16) suggest that it is context that is the determinant in identifying

121
the meaning of a word but they warn “There are even more dangerous
traps when the overseas context that appears to correspond to the native
speaker’s context in fact differs.” They suggest that students should be
encouraged to pay particular attention to collocations. Sinclair (1994-
98:18-19) suggests that the text itself contains everything that the reader
needs but warns that there are restrictions which, with the help of the
computer, can be explored to provide “models which help the text to reveal
itself to us”. Johns (1994-98:103) sees that the text that should be used
by students should reflect the target material the student needs to get to
grips with but should not be treated in a manner that would lead students
to develop ‘bad’ reading strategies and that any simplified text will only be
“used as a stepping stone to the real thing”.

3.1.3 Collocations

The precise collocations to be included will be determined from

concordancing of the corpus material to determine the most frequent
collocations, where collocation is used to mean (after Grenouff 1991) the
usual word or words found in the vicinity of the word being concordanced.
Wilks, Slator and Guthrie (1996:67) regard concordances as “special
scholarly tools” because they do not give explanations of meaning but only
index words against their occurrence in a corpus, leaving out all
information except the text citations. The teaching and learning load of
collocations can also be reduced by a contrastive approach to the concept
of lexical collocation (Bahns 1993). The fact that the undergraduates have
already studied English at school does not negate this need to see
differences in meaning from those that were learnt in different contexts.
The idea that a common word can take on a specialised meaning in
technical writing is discussed by several linguists (cf. Bright and McGregor
1970, Darian 1981, Weber 1981, Hoffman 1981). Weber (1981) gives the

122
example of the word digital and says “it is impossible to decide whether the
term denotes a special technical quality or is just an element of general
language use. The denotative meaning of the word is determined by its
textual venue i.e. whether we encounter it in a technical statement on
computer operations or in a sales talk in a watch shop, where the word
might be in a familiar juxtaposition to the word watch.” Darian (1981)
suggests that “ultimately the fullest meaning of a word lies at the
discourse level, which allows for an extended definition and deeper
exploration.” Martin (1992:172) claims that “technical language both
compacts and changes the nature of everyday words.” Students need to
connect words learnt at school with new contextual meanings in more
specialised contexts to avoid a particular kind of “false friends” where
words change their meanings in these different contexts (Hoffman 1981).
Moon (1994-98:122-124) lists ‘fixed expressions’ from her analysis of a
newspaper editorial. She (1994-98:126-7) finds the most common
expressions in the lexicon as a whole to be ‘functional’ or ‘grammatical’ as
opposed to ‘lexical’ and that (ibid.:134) examining the fixed expressions in
text provides information on the message and the speaker/writer’s
presentation and how this relates to objective statement or subjective
interpretation.
Whilst recognising that the analysis of specialist corpora will not
always reveal what the researcher expected, Tribble and Jones (1990:35-
36) make the following comment in relation to the utility of concordancing
for teaching purposes:
Two generalizations can be made about applications of concordance output,
in spite of their diversity. Firstly, most of them favour discovery learning.
That is, they present language in a way that enables learners to discover new
knowledge for themselves, rather than being spoon-fed. Secondly, they do
this by providing examples of authentic language. The fact that the source
material for exercises is drawn from real life rather than concocted by

123
 teachers increases motivation, as it gives learners immediate contact with
the target language in use.

The objective with collocations would be their application directly in the

teaching materials, ideally being controlled by the students themselves.

3.1.4 The Baseline Corpus

The first stage of this research involves the analysis of

approximately 30,000 science and technology texts taken from a
multimedia encyclopaedia. The advantage in this is that these
encyclopaedia are totally up-to-date in terms of student access to modern
technology, that is, they are to be found on CD-ROMs (see Integrating
Communication Technology 1996). Guillot and Kenning (1995:365),
discussing staff induction at the University of East Anglia, suggest that
“CD-ROM reference and textual databases are likely to become a major
resource in language education at tertiary level in particular: the sheer
magnitude of the information they make available, together with the
information processing and interfacing options they offer, open a vast array of
pedagogic possibilities for self- and teacher-directed learning.”

Although Guillot and Kenning see CD-ROM as particularly

important in tertiary level education, the fact that technology is being
introduced very much sooner in the educational curriculum means that
the students will increasingly be conversant with this type of application.
Students will expect to find it available in the university.
The most common, popular and favourably reviewed of these
encyclopedia are in American English; this reflects the fact that America is
the microcomputer powerhouse producing most of the popular CD-ROM
encyclopedia. The fact that most CD-ROM encyclopedia are in American

124
English fits in neatly with the English requirements of undergraduate
students in the University of Aveiro. One extremely useful addition is that,
not only can a word frequency study be carried out, but a further
dimension can be added to the research and that is the context in which
each high frequency word is to be found. So, not only can useful
information about lexis be obtained, but also a clearly defined use of those
items in specific scientific texts and also the link between the word and its
discourse setting. The number of texts that high frequency words are
found in can also add to the information about which words students are
most likely to encounter and, therefore, need to learn.
In addition, Rosenthal (1996:114) reports that introductory science
textbooks for further education in the United States have been getting
longer, broader and deeper in their coverage and reading complexity
making many of them become encyclopaedic.

3.1.5 The Level of the Material in the Corpora

Huddlestone (1971) gives the following analysis of science texts:

(1) “High-brow”, e.g. scholarly journal articles;

(2) “Mid-brow”, e.g. undergraduate textbooks;

(3) “Low-brow”, e.g. popular science for the general reader.

Table 3.1 Huddlestone’s Level of Science Texts

Swales (1985) argues that this “‘level of brow’ is not as important as the
expected relationship between the author and reader”. He describes these
‘mid-brow’ texts as “essentially instructional”. Similarly, Darian (1981:29-
30) describes the relationship between Material and Type of Audience. His
division is as follows:
1. Popular magazines, newspapers Uneducated layman

125
2. Scientific American and popular books A reader conversant in the
general area (e.g. business, social
science)
3. High school text Layman - limited general
knowledge and technical
background information
4. Introductory college text Layman - educated to college
level of general knowledge
5. Scholarly journal, specialized book-length Specialist and advanced
study (e.g. a volume on optics) graduate student
Table 3.2 Darian’s Level of Text and Audience

Darian claims that for each of his categories the writer assumes a different
level of “presupposition or background knowledge” on the part of the
reader. Glaser (1982:76-77) describes the difference in style between what
she calls “the academic scientific and technological style” addressed to
“‘insiders’ of a particular field of knowledge” and “the popular-scientific
style” used for “a general audience composed of non-specialists”. Glaser
describes the specific features of each of these being governed by the fact
that in the former “knowledge of the subject and the appropriate
terminology, the code of formulas and symbols and the various functions
of the syntactic patterns” is presupposed whereas the latter “show
entertaining deviations from the specialist’s topic for the purpose of
motivating the reader”. Furthermore, she distinguishes both of these
styles from a “didactic” style which attempts to make “a job-specific
problem (a scientific or technological subject) understandable to the
learner” found in textbooks, handbooks and other teaching material used
at schools and universities which are “subject to the didactic principle of
intelligibility of the text.” Similarly, Myers (1994-98: 189) finds that
different styles of research articles and popularizations construct different

126
views of science and that scientists “see their work as much more
tentative and mediated than does the public.” Myers (ibid.) found
differences in syntax, vocabulary and organisation between these two
types of ‘scientific text’ and he believes that teachers and students must
take these differences into account to “follow the entry of students into a
research community.”
This thesis contends that the appropriate material for
undergraduate students is the textbook, corresponding to Darian’s fourth
category above and Huddlestone’s second category of ‘mid-brow’. These
students are at a stage where their bibliographies reflect “instructional”
texts and therefore the encyclopaedia is an appropriate research tool as it
is also ‘essentially instructional’. This is because the students are in
transition form secondary to tertiary education and have yet to develop
greater knowledge of the subjects in their core disciplines on the Ano
Comum. Both of these text-types also fall into the category of educational
texts which will be reflected in their style.

3.1.6 Previous Studies and Text-Types

Swales (1985) takes Barber to task on the latter’s (1962) paper

“Some Measurable Characteristics of Modern Scientific Prose” for having
lumped together two different sorts of texts for his analysis (informational
and instructional from three texts with a combined total of only 22,400
words). Swales quite rightly criticises him for muddling his data.
It is essential if the right kinds of conclusions are to be drawn from
research for application in syllabus design for the corpus chosen to be
taken from the same kind and level of material, and for that corpus to be
of a significant size for generalisations to be made. Too many researchers
in the past have put different types of texts together for study, thereby
confusing rather than elucidating the issue. Myers (1998:179-190)

127
describes the differences between the focus of scientific texts and
“popularisations” which are prepared for a more general audience. It is
only appropriate in variation studies for a variety of text-types to be
examined together, but what those texts are needs to be defined clearly.
Biber (1988:208-210) describes precisely which texts he included in his
variation studies on speech and writing. Many of the texts were taken
from the LOB and London-Lund corpora mentioned earlier in Chapter
Two. This level of specificity makes Biber’s analysis an appropriate tool for
this study.

3.1.7 What are words?

The information obtained for “words” from this encyclopaedia needs

to be discussed. First the concept of word itself has to be addressed. The
Grolier encyclopaedia corpus takes the view that a word is anything which
is delimited by two spaces, rather in the way that a computer itself
‘recognises’ words2. This gives rise to some rather strange occurrences
such as ‘avant’, which on closer inspection is seen to be part of the
expression ‘avant garde’ (this could also be written with a hyphen but this
changes nothing for hyphenated words are also treated as two separate
words by this multimedia encyclopaedia).

Bright and McGregor (1970) discuss the difficulties involved in the

decision-taking stage of presenting lists. They give ten ‘problems’ that
need to be taken into consideration and offer some useful orientation for
my analysis.

Their ‘Problem 1’ refers to regular plurals, they suggest that ‘The pupil
who has mastered regular plurals will recognise monuments instantly if he
knows monument and vice versa. The difference is lexical not grammatical.’

2
Sinclair (1991) uses the term ‘word-form’ for this concept.
128
‘Problem 2’ is whether the word is a noun or a verb. Take, for example,
the word ‘play’. Is this to be regarded as two different words, once as a
verb and again as a noun? The Grolier encyclopaedia corpus does not
make any such distinction and so it is only through a more searching
analysis using concordancing that such distinctions can be resolved. This
is a very important issue however as Biber (1998:34-5) points out with his
finding that “deal/deals functioning as a verb is almost twice as common
as the noun use” in academic prose (from the Longman-Lancaster corpus)
whereas fiction (from the same corpus) shows the opposite with “the noun
use being considerably more common than the verb use”. This kind of
information is extremely important for ESP language learners and should
be brought out in the materials designed for their use.

‘Problem 3’ is whether regular forms of verbs should be considered

different. On this occasion they consider the difference to be grammatical
and not lexical. Each grammatical form is given a separate entry in the
Grolier encyclopaedia which will help to highlight some of the more
obvious grammatical forms to be found in this type of text. Indeed, this
could be regarded as one of the strengths of using this multimedia
encyclopaedia. Sinclair (1991) says

“It is now possible to compare the usage patterns of, for example, all the forms
of a verb, and from this to conclude that they are often very different one from
another. There is a good case for arguing that each distinct form is potentially a
unique lexical unit, and that forms should only be conflated into lemmas when
their environments show a certain amount and type of similarity.”

Sinclair’s concern is with dictionaries but his point would appear to

be even more important for application in teaching materials where certain
forms should be emphasised over others, where this is the form normally
associated with the particular genre of concern to the students, in this

129
case, in EST texts. Halliday (1993:71) also argues that it is impossible to
separate the grammar from the vocabulary and that it is the ‘the total
effect of the wording -words and structures-’ that the reader responds to.

‘Problem 4’ is when one word is in fact two lexical items as in Weber’s

example for “digital” given earlier in Chapter Two. These words are known
as ‘homographs’ and are defined as words that have the same spelling but
are different either in meaning, derivation or pronunciation. In this case
access to the context is necessary to define which lexical item is prevalent.
Once again this can be achieved through the use of a concordance,
examining the collocations the word is associated with and so clearing up
the ambiguity.

‘Problem 5’ refers to suffixes which Bright and McGregor regard as one

lexical item such as ‘young - younger - youngest’ or ‘play - player’
although they recognise that the latter is prone to ‘irritating but irrelevant
spelling problem(s)’ which they claim do not present a large ‘learning load’.
The high incidence of comparatives and superlatives in scientific texts
suggests this is an important structure and that teaching materials would
be wise to address this area of difficulty for students.

‘Problem 6’ refers to extending the previous argument of grammatical

suffixes to cover the relationship between such items as ‘permit’ and
‘permission’. Bright and McGregor say ‘We cannot, however, usually
assume a knowledge of Latin in our pupils’ but in this particular instance
we can assume certain similarities of words because of their common
Latin roots whilst at the same time taking particular care to deal with the
‘false friends’ which occur because of the different developments or
evolutions of meanings of words and their use in a specific scientific
context.

130
‘Problem 7’ is that of prefixes. Bright and McGregor claim that ‘any pupil
will be able to jump to the meaning of such items as ‘action - reaction’.
Whether or not this claim (and other similar ‘leaps’ in understanding by
students) is true would appear to depend to a certain extent on the
contact that students have had with English in their schools and their
understanding of discourse and shared scientific background knowledge
as discussed earlier. This will be taken up again later in the evaluation of
our students’ test results on entering the University (see Chapter 4).

‘Problem 8’ discusses how compound words and hyphenated words are to

be treated. Once again the Grolier encyclopaedia treats these as separate
items (as mentioned earlier). This suggests that further analysis will be
necessary through concordancing to analyse compound words.

‘Problem 9’ is concerned with what they term ‘form words’ such as ‘a, the,
and’. The Grolier encyclopaedia expressly excludes a number of such
words on the grounds that they are too common. A list of these very
“common” items is included at the end of each of the alphabetical lists as
they occur both in this chapter and in the Appendices.

‘Problem 10’ is phrasal verbs which are treated as separate items by the
Grolier encyclopaedia, that is to say, the verb and its particle appear
separately. This is a difficulty that can only be cleared up by examination
of the context of use of the main verbs found to be phrasal. Sinclair lists
the phrasal verbs that account for nearly 30% of all phrasal verbs in the
COBUILD corpus as “bring”, “come”, “get”, “go”, “put” and “take”. Separate
study of these would need to be made if this proved to be an area that the
students had particular difficulty with on their test results. The results
(see Chapter 4) produce mixed results in fact depending on the phrasal
verb being tested.

131
3.1.8 Other Features of the Text and Corpus

In addition to the ten problem areas discussed above other

considerations need to be taken into account such as American words and
spellings, abbreviations, pronunciation conventions, Latin and Greek
influence, word preferences. Each of these will be discussed below.

3.1.8.1American Words and Spellings.

Whilst many terms and spelling differences are relatively minor,

some could lead to confusion especially for those students who had
studied the British English model. Although many of the differences
encountered might be easily deduced by students such as colour – color
spelling differences, vocabulary differences like autumn - fall where the
latter is often taken as a verb in English would need particular attention.
Research carried out by Barber (1962:5) found that may was much more
prevalent in American scientific texts and can was twice as common as
may in British texts. Differences between some common terms for parts of
a car (automobile ) with such noun equivalences as boot - trunk and
bonnet - hood could lead to confusion in some engineering texts, although
these common terms in themselves are often more difficult than other
more specialised or technical terms for students. Portuguese students are
likely to find the scientific and technical terms are much closer to their
own language as these often have Latin roots. One other feature that can
prove particularly confusing for students is the difference in the use of
prepositions, especially such items as through - from...to. Students already
have particular difficulty with prepositions in English and to add to this
confusion by having two systems might be especially daunting.

132
3.1.8.2 Abbreviations.

There are usually differences in the use of abbreviations between

English and Portuguese (and indeed between British English and
American English) which could lead to a breakdown in comprehension.
English uses e.g. to indicate an example and Ex. an exercise; Portuguese
students have a tendency to use ex. to indicate examples. Similarly, nº
indicates number, but may be rendered nr. (near) by Portuguese students.
Students are often unaware of these differences and are confused when
faced with the appropriate form in English. Some of the common
abbreviation differences between British English and American English
are in measurements such as ml. for mile in British English which is
rendered, probably more appropriately to avoid confusion with millilitre
(milliliter in American English spelling), as mi. in American English. Added
to this difficulty is the increase in the use of abbreviations in specific
genres. Some uses of abbreviations are particularly idiosyncratic, like the
pronunciation system found in the CD-ROM material described where
sounds are rendered by groups of letters like ‘ahl’, ‘ahn’ and ‘ahr’ (see
Chapter 5, Table 5.1).

3.1.8.3 Pronunciation conventions.

Encyclopedia texts written for native English speakers often contain

indications of pronunciation which do not follow the dictionary (usually
international) phonetic notation which some students may well have come
across through reference work in their studies. These appear in corpora
and would almost certainly lead to considerable confusion for non-native
language learners. The sound groups would be totally inappropriate for
Portuguese speakers who approach pronunciation using their own
language’s conventions. An example of this problem can be observed

133
through exclamations like Ah and Ha which may be pronounced in very
much the same way for a Portuguese speaker but quite differently by a
native English speaker.

3.1.8.4 Latin and Greek Influence.

There has been extensive use of Latin in scientific writing and

historically Latin was the lingua franca of educated people and most
scientific studies were written in Latin into the Renaissance and thus
could be understood by scientists in other countries. In modern scientific
writing these Latin and Greek roots do still exist, but the use and
knowledge of Latin and Greek is no longer a pre-requisite for most
branches of science. Stubbs (1996:70-1) sees “Graeco-Latin loan words”
as having been used to build up the vocabularies of institutions which in
turn leads to “differential access to subjects on the school curriculum”.
Stubbs is concerned with the “authoritative knowledge” that is expressed
by such features of texts as examples of power relations and the way that
writing is always aligned. Examination of just how Latin and Greek terms
are employed in modern texts can be an important tool in deciding what
should be included in the syllabus for today’s undergraduates. White
(1998:290) claims that the “strangeness” produced in scientific texts
through the use of Latin and Greek derived terms is a deliberate measure
taken to ensure that the reader recognises that these terms are not to be
taken as the normal view of reality construed through the use of the
vernacular. Thus the need to construct different views of reality, which are
not related to “common sense”, is essential for an understanding of (the
language of) science to take place. Laurillard (1993) gives a number of
examples of how the concepts used in science do not match those found
through “common sense” and how lack of success in building adequate
conceptual frameworks by students can occur in many different ways. She

134
claims that higher education has not yet, found a means of coping with
this as yet other than through the tutorial question and answer system to
draw out where and when the misconceptions occur.
Strevens (1978:193) maintains that Latin and Greek roots and
affixes combine to form an extremely large number of words which are
‘science-specific’. He cites the roots aqua-, cyto-, hydro-, plasma-, pyro-,
and the prefixes ante-, anti-, poly-, post-, pre-, sub-, and suffixes -fer, -ite, -
logy, -valent. Strevens (ibid.) maintains that this scientific vocabulary
makes up a ‘normal part of the training of all scientists’. Portuguese
students are fortunate in that they have a Latinate language which may go
some way towards providing them with knowledge of and insight into the
scientific applications of Latin roots.

3.1.8.5 Word Preferences.

Quirk (1995) has shown that some words are preferred in certain
texts or registers even though there may well be a very similar synonym.
“Ancient” and “old” for example may exist in almost equal numbers of
texts (range) and frequencies in the corpora whereas “attempt” as opposed
to “try”, and “change” as opposed to “alter”, may exist in different
frequencies showing preference for one form over the other. Quirk (ibid.)
argues that these kinds of choices, although apparently arbitrary, can
indicate formality in texts and may therefore be representative of the
particular genre they are found in. Lemke (1998:92) suggests that choices
of lexis contribute to the “attitudinal stance of a text to its audience, to its
content, and to other text-embodied viewpoints”. McCarthy and Carter
(1994:104-5) suggest that vocabulary choice is just as discourse sensitive
as grammatical choices and that if language is to be considered as
discourse “vocabulary must be a concern as much as any other aspect of
language form”.

135
 A similar position is adopted by Biber, Conrad and Reppen
(1998:43-54) who demonstrate that “big”, “large” and “great”, which are
often presented to students as synonyms, are usually used in quite
distinct patterns and with specific meanings. They find (1998:51) that
fiction and academic prose have different preferences for these words with
“big” being more common in fiction and “large” in academic prose. While
both registers use “great” with “deal” as a collocate, fiction uses many
more senses of “great” than does academic prose. They account for these
findings by suggesting that “fiction texts contain frequent physical
descriptions” and “more varied descriptions” whereas academic prose texts
“deals with size” and “specific measurements”. They go on to examine the
collocates associated with these words in the two registers. Similarly
(1998:98-99) they examine the preferential use of “begin” and “start” in
fiction and academic prose and discover that the intransitive use of start
is the most common in both registers but is more prevalent in academic
prose. In contrast “begin” is usually used intransitively but in fiction it is
used mainly with a to - clause. In other words, they argue that the
patterns of language use are not synonymous across registers. Thus the
vocabulary preferences found in the corpora are significant both as
representations of the style of the texts and as a means of demonstrating a
model of authentic usage to students.

3.1.9 Optical Character Recognition

Burnard (1992) points out that the optical scanner or optical

character recognition machine (OCR), which in this study was used to
compile the physics and chemistry corpora, can only recognise what is
visibly present on the page and that it cannot undertake any kind of
editing nor can it distinguish structurally different components of the
printed page even if these are visually distinct, such as footnotes and

136
headings. Any corpora searched will not provide these distinguishing
features so that an analysis of the texts themselves as published can often
reveal other interesting and important features of those texts (see later
5.2).
Laurillard (1993-97:27) describes academic knowledge at university
level as a process of ‘mediating learning’ because the students have to
learn what others have given insights into rather than what they can have
direct experience of. She suggests that because academic knowledge “has
this second-order character, it relies heavily on symbolic representation as
the medium through which it is known” and although the medium may be
language it may also be “mathematics symbols, diagrams, musical
notation, phonetics, or any symbol system that can represent a
description of the world.” Therefore, students in university have two
problems to overcome, the first that of handling the representation system
and the second the ideas which they represent. Some features that must
be taken into consideration therefore are the use of typographics, titles,
subtitles, summaries and conclusions, drawings and diagrams, and
formulae, numbers, equations and tables. These features should add to
the student’s understanding of the text, provided that the student is aware
of the conventions used and has been trained in recognising the
multimodality of texts. Lemke (1998:95) suggests that scientific text is not
meant to be read in a linear manner and for him it represents a “primitive
form of hypertext” where “footnotes represent an optional branch for
readers, so do figures and their captions, and the parenthetic or main-text
expressions such as ‘(Table 3)’ or ‘as seen in the first table’ which point to
them.” In contrast speech is linear in this respect. The number of
dimensions that are then available to the reader is much wider and access
to them is much more open, the reader can choose what information to
access then from the different textual and visual information present in
scientific texts. Nevertheless, students have to have background

137
knowledge of the canonical forms used in science in order to be able to
understand and interpret the information available.

3.1.9.1 Typographics.

The punctuation and use of italics are also vehicles of information in

a text. Darian (1981) refers to the use of typographics as vehicles of
definition. The equals sign, the colon, pairs of commas, parentheses, the
dash, quotation marks and italics can all be used to give or signal
definitions in scientific and technical writing and are constantly
interacting with the text itself. In this way it is possible for the definitions
in a text to be either “overt” or “covert” (Darian 1981:36). Often the word
which is to be defined in a covert way is flagged, that is it can be located in
bold, a convention which Halliday and Martin (1993) have explored in
Writing Science and White (1998) discusses in Reading Science, but which
nevertheless harks right back to Locke in the seventeenth century and his
concern with definitions. Lemke (1998:95) suggests that typography
serves to orient as well as organise and that “the use of italic and boldface
types signals emphasis or importance, as does the relative point size of
type in titles, headings, abstracts, footnotes, captions, labels, etc.” while
“paragraphing and sectioning of text, and geometric relations of figure
space to caption space indicate to us which elements are to be
preferentially read in relation to which other elements; what goes with
what.” I would go further than this. This easily accessed presentational
mode affects our ability to bring out latent meaning in texts. It is
interesting to note how the use of computers themselves are even
influencing this kind of discussion. In the past, before the widespread use
of word processors, neither teachers nor students would have been so
ready to discuss “point size” or “font”.

138
 The use of a number of symbols through computers has taken this
further in modern materials and, if over used, these may serve to irritate
rather than encourage as is the case of the ubiquitous, perfidious ‘smiley’
to indicate a joke or other attempt to be friendly or light-hearted. Attempts
to make materials for learners more attractive may require a clear
statement at the beginning of how these symbols will be used in any text.
If learners skip past these early explanations in the textbook, they may be
in danger of missing many of the connections the author intends to make.

3.1.9.2 Titles, Subtitles, Summaries and Conclusions.

Van Dijk (1997:10-11) claims that discourse topics define the overall
‘unity’ of discourse and are “typically expressed in such discourse
segments as headlines, summaries or conclusions.” He also claims that
they “also happen to be the information that we usually remember best of
a discourse,” which, if the case, means that these features are especially
important for study purposes.

Increasingly in modern textbooks the flow of the text is divided into

small sections or paragraphs with titles or subtitles used to indicate the
topic discussed. These also serve to allow topics to be easily located within
the text. The use of a system of colour coding of these titles and subtitles
is also prevalent which together with diagrams and drawings produce a
much more attractive and much less dense appearance than the old dry
textbook styles of some decades ago. Visual display of the kind described
above is now an essential element in making teaching materials attractive
and with a proper pedagogic basis can facilitate understanding of the topic
discussed.

Similarly the statement of aims and objectives at the start of a topic

and summaries at the end of topics help students to focus on what they
are going to learn and what they should have learnt from the texts that

139
went before. In this way they serve to prepare the learner for the task and
serve as a check on learning. These types of activities are known as ‘wrap
around exercises’ to assist with text processing and to enable the learner
to monitor progress on their own at home.

3.1.9.3 Formulae, Numbers, Equations and Tables.

The use of formulae, numbers, equations and tables in scientific

texts features are extremely important. They are the means whereby an
alternative or additional representation of the information contained in the
text is provided. Certain conventions need to be understood such as the
arrow or equals sign, but, fortunately, these conventions enjoy
international standing and recognition. Lemke (1998:96) warns however
that particularly in tables “readers are expected to supply the canonical
semantic relations of thematic terms which are often underspecified or
omitted”.
Similarly, the chemical symbols for elements and compounds are
standardised through international convention and should not prove to be
a stumbling block provided that they are understood in Portuguese,
especially as the formulae follow the English order; for example NaCl,
sodium chloride, in Portuguese would be reversed and read chloride of
sodium (chloreto de sódio). Thus, it can be argued that the number and
range of formulae should add to understanding rather than obscuring it,
provided that there is shared background scientific knowledge. However,
some mathematical working could prove confusing (see 5.4 Mathematics ).
Lemke (1998:90) found that in most of the theoretical physics
articles that he studied the running verbal text would make no sense
without the integrated mathematical equations “which could not in most
cases be effectively paraphrased in natural language” even though they

140
were meant to be read as part of the verbal text “in terms of semantics,
cohesion and frequently grammar”.

3.1.9.4 Diagrams and Drawings.

As with the use of other textual features, the use of diagrams and
drawings should enhance the understanding of the surrounding text,
provided that the referencing to these is understood. In general, visual
material in the text was seen as a form of redundancy as it reiterates what
is being discussed. However, Lemke (1998:104) disagrees with this
position and claims that visual figures and mathematical expressions add
important or necessary information and so complement or complete the
main text. Modern discourse analysis sees multi-modality in texts as an
essential feature of study in discourse semiotics. Kress, Leite-García and
Leeuwen (1997:257) say that “producers of texts are making greater and
more deliberate use of a range of representational and communication
modes which co-occur within the one text” and that the reader has to take
these into account in order to “read texts reliably”.
Van Dijk (1997:6) suggests that “in these times of multi-media
communication ...an analysis of the visual dimension of discourse is
indispensable.” Van Dijk is much more interested in non-verbal signs or
semiotics but, nevertheless, the visual element in student’s textbooks
should be an aid to understanding the discourse of the text if the students
can interpret them accurately. Lemke (1998:87) argues that semiotic
systems such as language, tables, graphs, images and diagrams do not
just “add-on” meaning to a text but actually create new orders of meaning
thereby “multiplying meaning”. Furthermore, Kress, Leite García and van
Leeuwen (1997) suggest that it is important to see visual images as
independent vehicles for meaning in their own right. If the students can
make the connections between visual images and text or ‘read’ visual

141
images in scientific texts accurately, this would help the students to
ascertain the meaning in those texts. The question of whether students
can do this successfully is taken up again later in 5.2.4.

3.1.10 Comparison with other published data.

Bright and McGregor (1970) claim that technical texts contain a

large number of words that are ‘outside simplified English’. They list
twelve such words: ‘absurd, adequate, adjoining, aggression, alert,
alternative, amateur, ample, apparatus, apprehensive, automatic, available’.
Of these words only six do in fact occur with significant frequency (defined
as 100 occurrences or more) in the Grolier encyclopaedia (see word list in
Table 5.1) and of those six only two could be considered sufficiently
different from Portuguese to warrant attention (apparatus and available). It
would seem that from this small comparison that the claims made by
corpus linguists that intuition and the reality found through empirical
research of computer corpora of naturally occurring texts (as opposed to
specially written examples) are at odds is essentially correct.

3.2 Needs Analysis

3.2.1 The Students’ Level of English

In deciding which words should be excluded from the list, several

considerations were borne in mind. First and foremost was what could
reasonably be expected from the students’ previous contact with English
in the schools. When the new students were tested each year they were
asked to state how many years of study of English they had had. The
results were as follows:

142
Table 3.3 Students’ Number of Years of Study of English
Years of study 1993/4 1994/5 1995/6 1996/7 1997/8
0 4.7%
1
2 2.7% 2.8% 1% 0.32%
2.5 0.7%
3 18% 13% 14% 8% 5.73%
4 3.3% 0.6% 1% 1% 1.91%
5 38% 26.5% 29% 30% 23.25%
6 1.3% 4% 3% 2% 5.09%
7 28% 47.5% 47% 53% 54.78%
8 2.7% 4% 6% 4% 7.01%
9 1% 0.96%
10 0.7% 0.6%
11 0.7% 0.6%
15 0.64%
18 0.32%
(3.57% of those who took the test did not answer this question at all.)

Consideration of the results for the academic year 1993/4 shows

that, as some comments offered by the students remind us, there can be
many instances which do not follow what would have been predicted for
students leaving secondary school and starting university. The 1997/98
results also demonstrate that the question on the test paper may have
been given a cursory glance and understood as Quantos anos tem? and the
essential years of study have been overlooked to produce the answer
‘eighteen’. 3

It was expected that the students would have studied English for an
average of three, five or seven years, with some time gap between the years
in which they studied English and university. What was most worrying for
the academic year 1993/4 was the percentage of students with no English
at all embarking on the course in conjunction with a significant number of
students who had studied English for seven or more years.

3
When a student who gave an answer like this was questioned about it later, she admitted that she had in fact
given information about her age and not her studies. However, this particular student justified her answer
by explaining that she had in fact spent most of her life in America so she felt that the English language had
indeed been part of her entire life.
143
 The results for the academic year 1994/5 were a little more
encouraging in that they show no students with absolutely no English,
nevertheless, there are still a significant number with very little English.4

The figures for the academic year 1995/6 show much more clearly
the expected breakdown into three, five and seven years of English. Some
of the intermediate figures could be accounted for by students who have
had to repeat a year at school, which, if true, would suggest that those
percentages were students who could be considered weaker than others in
the same broad categories.5
The figures for the academic year 1997/98 show how there is a
general trend for students to be stronger in English than before, and the
answer ‘fifteen years of study’ reflects students who had been brought up
abroad in English speaking countries. Certain courses, such as Novas
Tecnologias e Comunicações - New Technologies and Communication
(NTC), appear to be attracting students who are generally stronger in
English which is not perhaps surprising given the nature of this course
which has a slightly more ‘humanities’ or ‘arts’ bias than the other Ano
Comum courses. These students also continue with their English studies
for a further year unlike most of the science and technology students in
the university.

4
Chatel (1999:246) records a similar change from 1988 to 1994 for sociology students in the University of
Coimbra.
5
Drª Maria Adelaide de Araújo Nunes of the University of Evora (1999:258) describes the “uncongenial
environment” for ESP with students who “have had only a few years of English at secondary school and/or
having systematically failed the subject there” and so “feel at a loss and are understandably reluctant to
study a subject that they hoped they would never encounter in their lives again”.
144
Figure 3.1 Pie Graph for the Academic Year 1993/94 showing the Students’ Number of
Years of English
1993/94

> 3 yrs < 7yrs

8% 4%
6-7 yrs
29%

3-5 yrs
59%

Figure 3.2 Pie Graph for the Academic Year 1994/95 showing the Students’ Number of
Years of English
1994/95
> 3 yrs < 7yrs
3% 5%

3-5 yrs
40%

6-7 yrs
52%

145
Figure 3.3 Pie Graph for the Academic Year 1995/96 showing the Students’ Number of
Years of English

1995/96

< 7yrs
6%
3-5 yrs
44%

6-7 yrs
50%

Figure 3.4 Pie Graph for the Academic Year 1996/97 showing the Students’ Number of
Years of English

1996/97
> 3 yrs
1% < 7yrs
5%
3-5 yrs
39%

6-7 yrs
55%

146
Figure 3.5 Pie Graph for the Academic Year 1997/98 showing the Students’ Number of
Years of English

1997/98

> 3 yrs
< 7yrs
1%
3-5 yrs 8%
31%

6-7 yrs
60%

One other factor could be affecting these figures and that is that in
the first year all the students took the test but in subsequent years
students were advised that they need not take the test if they felt that
their English was not of a high enough standard. However, in recent years
students have changed their attitude and appear to treat tests as a kind of
lottery where they hope that through some stroke of luck they will pick the
winning combination of answers. They perceive that at any rate they have
nothing to lose by trying. It may also reflect a change taking place in
schools where an increasing number of students opt for English as their
main foreign language at an earlier age and so feel that they are of a
higher level. There are also more and more private language schools
opening up all over Portugal and some of their students must now be
coming through to university in increasing numbers.
Over the years some of the students have felt compelled to add
comments to the question they were asked about how many years they
had studied English. Rather like the example above, some students gave
an explanation for their answers. This could be because they had repeated
a year, as surmised above, or that they had done all of their studies in
147
English in another country, for example, Australia, South Africa or
America. However, a small number of students gave value judgements
about the quality of the teaching they had received, one student replied
“três anos e pessimos” whilst others inadvertently showed the difficulties
they had with English by answering “I am three years” to this question.
Other students explained that their studies had taken place a number of
years before they had taken up their university place and so their English
was ‘rusty’ and, yet others, that they had studied both at school and at
private language schools, thus completing ‘double years’ or that they had
taken the Cambridge University examinations in English.
These figures suggest that assumptions made about the level of the
students’ English could be wildly inaccurate, although there is a general
trend for the students to have studied English for longer in the secondary
school6. One other consideration is that, although the students may have
studied English for seven years, most are unlikely to have studied it in the
final year of their secondary school course as they will have chosen to
follow science subjects and not the humanities. Given the kinds of
problems explained above and the increasing pressure on grades for
university entrance, it is also possible that some of the students had not
studied English for more than two years because of repeating the final
year to improve their grades.
The fact that students have studied English in secondary school
does not negate the fact that their knowledge of language is limited to
what was taught on a general English course which is likely to have
concentrated on spoken English and more ‘literary’ kinds of
comprehension and composition, and to have dealt little, if at all, with the

6
Drªs Ana Maria Ferreira, Dulce Ramos and Fátima Braga da Silva from the University of Porto in their
paper “Evaluation des Curricula de FLE au Portugal” (1999:333-337) show the numbers of students
studying French, English, German and Spanish in the central region of Portugal in the academic year
1994/95. The figures for English demonstrate clearly that a significant number of students continue their
language studies into the final three years before university (approximately half of those studying English in
the 3rd cycle - 7th to 9th years of school).
148
language of science and technology. Langkilde (1982:523) describes the
barriers students in the Copenhagen School of Economics were found to
have because “unless they are made aware of the necessity of developing a
particular method for dealing with specialized texts they will for a long
time go on treating an economic text in the same way as they treated a
chapter from Balzac or a scene from Molière in grammar school.” Tavares,
Valente and Roldão (1996:62) say that the English Programmes for
schools do mention types of texts but these are given as “dialogue,
interview and advertisement” (I:42) and discourse organisation as
“descriptive, narrative and argumentative”7 (I:48)P. These authors suggest
that cultural identity and understanding, within the general development
of the pupil as a responsible citizen, are the main concerns in the
programmes for modern languages in Portuguese schools at the moment.
They also point out that teachers need to be up-to-date with their training
if they are to be able to cope with the requirements of the programmes, an
issue that has been mentioned many times in relation to teaching science
and technology. The school teachers themselves usually come from a
‘literary’ or humanities background and are, therefore, unlikely to feel
comfortable teaching English for science and technology.
Research carried out with students of the fourth year of the teaching
degrees in Portuguese and English, and English and German,
demonstrate that these future teachers have difficulty with numbers in
English just as the students entering the university have (see Test Results
for New Students, Chapter 4). This situation would therefore seem to be
self-perpetuating as teachers are generally unwilling to teach something
that they themselves find difficulty with. Swales (1973:9) describes how
teachers found it “almost impossible to view their Science students’
interests as different from their own” and therefore assumed the students
would find boring what bored them.

7
My translation of “dialogo, entrevista, anúncio” and “descrição, narração, argumentação”
149
 Overall these results would suggest that increasingly the students
could be expected to have an intermediate level of English but with no
science subject specialisation in English. The structure or form words
mentioned earlier should be quite well known to the students but, as will
be shown later in the test results, some discourse markers are less well
understood. The syllabus design implications of these findings are to
complicate the issue of the level at which to pitch the instruction. The
needs of those students in the bottom 1% with less than three years of
English can hardly be met, and this will lead to their virtual exclusion
from most of the activities designed for the majority of 60% with six or
seven years of English. Equally well the top 8% may find the level pitched
beneath their capabilities and so lose motivation. These more able
students must be included in the activities carried out in such a way that
they feel stretched and that they are also making progress. It might be
possible to engage these students in helping their classmates to reach a
higher standard and incidentally help to create bonds between students in
this new environment which is seen to be necessary for successful
learning (Tavares et al.1996).
The knowledge of science and technology that the students bring
with them to the first year of university is also variable. Some students
will have chosen to study physics in their final year at school and some
chemistry, some will have studied more mathematics than others and so
on. This implies that homogeneity in terms of subject knowledge cannot
be guaranteed either in the students entering the foundation year
disciplines. This fact will have repercussions on all of the strategies and
skills that these students require in order to be able to perform well in
their studies of the subject matter in a foreign language.

3.3 Biber’s Methodology of Variation Studies and Corpora

Analyses

150
 The corpora from the Physics and Chemistry textbooks on the
students’ bibliographies will be examined using Biber’s (1988)
methodology of text variation to try to see what must be taught to the
students in the university that is specific to this text-type and, thereby, to
make the course relevant and to ‘fill in the gaps’ that the students bring
from their studies in school. Biber was conducting research into variation
between speech and writing but he provides a very explicit methodology
for the description of the linguistic characteristics of the range of genres in
English that he included in his study, which will allow comparison of the
physics and chemistry texts under study here with his results for
Academic Prose.
Biber’s goal was to include all the ‘potentially important linguistic
features’ of the different genres included in his study in order to identify
the ‘linguistic parameters along which genres vary, so that any individual
genre can be located within an ‘oral’ and ‘literate’ space, specifying both
the nature and the extent of the differences and similarities between the
genre and the range of other genres in English’. It is this identification of
the differences that needs to be studied in order to identify those areas to
be included in a syllabus for undergraduate science and technology
students. Biber claims that it is ‘bundles’ of linguistic features that occur
together in texts that ‘work together to mark some common underlying
function’.
Biber identifies 67 features from previous research, which can be
grouped into sixteen major grammatical categories: (A) tense and aspect
markers; (B) place and time adverbials; (C) pronouns and pro-verbs; (D)
questions; (E) nominal forms; (F) passives; (G) stative forms; (H)
subordination features; (I) prepositional phrases, adjectives and adverbs;
(J) lexical specificity; (K) lexical classes; (L) modals; (M) specialised verb
classes; (N) reduced forms and dispreferred structures; (O) coordination;
and (P) negation. He gives very precise definitions of each of these features

151
(see Appendix A) together with the functions that other researchers have
ascribed to these features. For example in Tense and Aspect Markers he
suggests that past tense forms are usually taken as the primary surface
marker of narrative; perfect aspect forms are associated with
narrative/descriptive texts and certain kinds of academic writing, and that
these co-occur with past tense forms as markers of narrative; and that
present tense verbs can be used in academic styles to focus on the
information being presented and remove focus from temporal sequencing.
By using Biber’s features it will be possible to analyse both the
register and discourse of some of the texts in the undergraduates’
bibliographies in order to apply the results to designing an appropriate
syllabus for these students.

152
Chapter 4 Test Results for New Students
 Chapter 4

Test Results for New Students

4.1.1 Student Numbers

In the first academic year of the Ano Comum and of the Preliminary
Test, that is, 1993-1994, there were approximately 1350 students
studying English in the Ano Comum. and in 1994-1995 there were
approximately 1200 new students entering the Ano Comum and about
180 repeating this discipline. These numbers have continued much the
same for the academic years 1995/96, 1996/7 and 1997/8.
As I showed earlier most of the students entering this discipline
could have studied English for either three, five or seven years in their
secondary schools. What the students have learned, have learned
incompletely or have not learned at all in their secondary schools is
crucial for syllabus design, so the results of the Preliminary Test were
analysed for it to be possible to decide what needs to be given particular
attention in their proposed syllabus.

4.1.2 The Preliminary Test

Using local knowledge and experience, we realised that some of our

students were likely to have studied in an English speaking country
(probably 1-2%) or to have studied English at one of a number of
reputable language teaching institutes (up to 15%), and, therefore, be
generally proficient in the language. Some students, in replying to the
survey mentioned above, volunteered the information that they had

155
already obtained the Cambridge University Certificate of Proficiency in
English which is a qualification regarded as a minimum English teaching
qualification by the Ministry of Education in Portugal. An innovative
decision was therefore taken to test all the potential students
immediately at the beginning of the academic year and give all those who
were deemed to have a sufficient knowledge already the opportunity of
being excused from the discipline altogether. This decision was
applauded by the student body who suggested unsuccessfully that they
would like it extended to other core disciplines. The effect of the decision
to innovate in such a way was to reduce class size a little in an attempt to
give the less proficient students more time and attention in class and to
permit those students with greater proficiency to concentrate their efforts
in other areas where they might not be so proficient. Many students
started learning English (mainly in private schools) whilst they were still
in primary school and this early teaching has also come to be seen as
beneficial in state education in Portugal. Changes have been introduced
in the curriculum to permit different schemes of foreign language study
often also extending this to the final years of secondary schooling for all
children no matter what their core curriculum. Innovation for
undergraduates on science and technology courses has also been the
focus in tertiary education. Students who were found to have great
competence in English were also considered to be likely to be demotivated
by being in a mixed ability class with over forty other students. In actual
fact some students who had expected to be excused from the discipline
were surprised to learn that it was their knowledge of science that let
them down in the test. The specific English being tested went beyond the
mundane day-to-day usage of children and required a more mature,
informed view from students
As was mentioned earlier, all the new students coming in were
tested to see if some of them could be offered the chance of not taking

156
this discipline at all because their English was considered to be of a
sufficiently high level. This level would correspond to already knowing
enough English to be able to pass comfortably the kind of test that they
would be given at the end of the year after studying specific English for
science and technology in large mixed ability groups for two hours per
week for one academic year. In other words, a proficiency test was needed
to evaluate the student’s knowledge. It was decided that an adequate
initial standard of English would equate with a mark of fifteen or above
out of twenty.1
The test had to be one that could be administered and marked
easily, given the numbers involved. A multiple choice format was chosen
as an objective test and so that a template could be used for ease of
marking. A short paragraph was also included to verify the results of the
multiple choice test. This was changed to a reading comprehension test
in the fifth year of the test as it was felt that this area of competence
needed to be checked so that we could feel reasonably confident that the
students who passed the test well were capable of coping with the
reading that they would have to do in English on their courses. The
ability to write well in English was also considered to be less important to
the students’ immediate curricular needs. Four versions of the same test
were produced in order to avoid copying, this was later changed to two
versions of the test because this was found to be both much easier for
the writers of the test and yet sacrificed nothing of the security aspect of
the testing. The test was made up of both grammatical and vocabulary
items as both of these areas were deemed pertinent and specific to the
language of science and technology which the students would need to
cope with in their studies.

1
It was considered that the students would be disinclined to accept our offer if their mean mark was lower
than this, which would defeat the object in view of reducing class size and not wasting student study time.

157
 The results of the test were analysed and those students who
obtained a grade of fifteen or more were duly informed that they need not
attend classes and indeed had already obtained a final mark for the
course (fifteen and above). This does not mean that all the students thus
informed (about 10% of those who took the test) decided to accept this
result. No bar was placed on these students attending the course, if they
so wished, and indeed some did choose to attend. The students could
also choose to take the examinations at the very end of the academic year
if they felt that they could do better than they had initially. Some
students felt that this was possible after they had had access to the
materials used in the discipline from which they could then study some
of the relevant scientific and technical English which they perhaps felt
they were unsure about initially.
This test was analysed and refined for use in future years but
results for the first years show that the major discriminators were
specific vocabulary and grammatical items such as the present perfect,
second conditional, gerunds, phrasal verbs and specific lexical items.
Increasingly items have been included in this preliminary test which
reflect the syllabus of the first year, items on pronunciation and
numerical knowledge for example.

4.1.3 Test Results 1993/94

The test that was used in the first academic year, 1993-1994,
consisted of fifty multiple choice questions and a short, 100 word,
paragraph on a given topic. The reason for this format was first and
foremost that it would be very time-consuming to administer any other
sort of test to such a large body of students and be able to publish the
results early enough so as not to take up too much of limited teaching
time. The written part of the test acted as confirmation of the result
158
obtained in the multiple choice test. The topics given on the first test
were:
i) The importance of computers for students at university.

ii) Why science students should study English at university.

iii) The importance of the course you have chosen to study.

As the students had to write only 100 words on one of these topics,
they had to be extremely concise. Writing such a short amount is often
more difficult than permitting the students to write as much as they
wish. Indeed, many students attempted to go beyond the specified limit
whilst others did not even attempt the written section at all. The
questions also required an expository or argumentative style of writing.
Although discrete item tests are not considered very valid, they do
have the advantage of being reliable. As Weir (1988) points out, the test
can also be made more valid by taking into account the needs of the
students on their individual courses. The different departments took the
optimal view that students would need all four skills of reading, writing,
speaking and listening in order to pursue an academic career (see later
4.2 Needs Analysis by University Department) but the constraints imposed
by the length of time available for the discipline meant that the goals
would have to be somewhat more short-term and reflect the arguably
more receptive skills of reading and listening. The latter could only
represent a small percentage of the whole syllabus even so. Therefore, the
syllabus that the students’ would subsequently pursue could not be
considered communicative in any modern sense of that term. Reading
and some listening would form the bulk of the syllabus and these would
be approached in a way that could give the students ‘enabling skills’ in
the hope that given time they might build on what could be
taught/learned in such a small space of time. In other words, the
methodology used would be as learner centred as possible in order to

159
meet the needs of the individual students as far as this could be
achieved. The test then had no reason to reflect other methodological
aims. A greater allocation of course time and resources would have
behoved a more comprehensive test.
The questions on this first test started with the simple present
tenses, negatives and question forms and went on to modal
constructions, conditionals, phrasal verbs and passive constructions. In
other words the accepted ‘easy structures’ to the more complex. Some
specific vocabulary questions were also included. The results showed that
approximately ten per cent of the students enrolled had achieved a mark
of fifteen or more and could then be released from the discipline.
However, evaluation of the test results also showed that ten per cent of
the students could not competently handle what are considered basic
structures. For example, present simple question forms, present
continuous and present perfect tested in the following way:
Questions of the type:

3. ____________________ coffee?
A Do she like B Likes she C Does she like D Like she

caused nine per cent of those tested to make an error and eleven point five

per cent were caught out by:

4. I ____________________ English.
A am study B studying C studies D am studying

Most of the students tested, that is 96%, could not answer the following

correctly:

34. This is the first time I ___________________ Aveiro.

A am visiting B visited C visit D have visited

160
Certain items like specific vocabulary, prepositions and the subjunctive

caused more than 75% of the students trouble, for example:

13. You _________________ take an aspirin for your headache.

A had better B would better C will better D have better
30. It’s time we __________________ .
A go B went C going D goes
41. I will have to phone later his number was ________________ .
A occupied B engaged C talking D speaking
38. They congratulated their cousin _______ passing his driving test.
A at B by C on D with
50. The car tried to ________ the lorry while it was waiting at the pedestrian
crossing.
A overtake B overlook C overpass D overcome
44. After the mission the space shuttle lands on a _______ in the same way as an
aircraft.
A path B highway C motorway D runway
37. Her cupboards are full of clothes, most of _______ she never wears.
A them B which C those D that
21. They _______ in love at first sight.
A fell B felt C feel D fall
25. He took ____ Keith at once and they became firm friends.
A after B up C to D on
36. __________ the high price of meat, the family bought lamb or beef every week.
A Because B Besides C Despite D In spite

More than 50% of the students could not manage questions on the
second conditional (67%), indirect question forms (60%), the future perfect
(62%), “suggest” with a direct and indirect object (70%), the phrasal verbs

161
“get over” and “put in for” (54% and 61%) and a further five vocabulary
questions including such items as “traffic jam” and “experiment” (50% and
62%). With such generalised difficulty, the syllabus must obviously take
such language deficiencies into account as teaching syllabi always
consider the average student. Extreme positions whether higher or lower
are inevitably for a smaller number of students and so those who
represent the middle ground, median or more properly the standard
deviation of + or – 1 on the normal curve found from testing are always
those taken as the ‘average’ students for whom any course is designed.
This is contrary to many of the older systems, particularly of higher
education, which aimed to teach an elite group with all others falling by
the wayside. The numbers of students involved in modern education in
developed countries necessitates an attempt to raise the general level of
education of all of those involved in the education process and
necessitates new methodologies to achieve this aim. Therefore those test
results that show widespread difficulty but not almost total impossibility
for students are taken as items that need to be included in the syllabus
in order to raise the standard of English of the majority of the
undergraduates.

4.1.4 Test Results 1994/95

The test that was used in the second academic year, 1994-1995,
also consisted of fifty multiple choice questions and a short, 100 word,
paragraph on a given topic. The reasons for this being exactly the same
format as that in the first year was that the numbers involved continued
to prohibit almost any other practical possibility. However, this time the
test items were altered to incorporate some of the items considered
fundamental to the course as it had been taught in the first year of
operation of the Ano Comum. Other items that were considered to be
inadequate discriminators, after the test results had been studied for this
162
purpose specifically, were eliminated. Thus further validation of the test
was incorporated without sacrificing either any of the reliability of the
test, its objectivity or, above all, its speed of administration and
correction.
Overall results were now also available about pass rates and grades
of the first year on this foundation course and these results also
validated the test in that the percentage for allowing students to choose
not to take the course at all equated well with that of all the students of
the year reaching a high grade that is, 15 or more (approximately 12%).
The items considered fundamental that were now included in the
test covered both the specific vocabulary that had been taught during
1993-1994 and an attempt to assess the student’s awareness of
pronunciation. The results this time showed that approximately 13% of
the students had reached a level which was considered adequate, and
could be released. The proportionate increase in the number of students
released was most likely to be due to the fact that, when the answers to
the query about the numbers of years they had studied English were
collated, it was found that 57% of the students had studied six years or
more (43% had studied five years or fewer) whereas in 1993-94 33.4% of
the students had studied six years or more and 66.7% had studied five
years or fewer (see 3.2.1 The Students’ Level of English).
Nevertheless, when the answers to the multiple choice questions
were once again analysed, it was found that the students continued to
have difficulties with modals (97%), direct and indirect objects after
“suggest” (63%), phrasal verbs (71% and 83%), reciprocal pairs (63%) and
the subjunctive (89%).
It was perhaps less surprising to find more than 75% of the
students having problems with those more specific vocabulary items that
had been introduced. Questions like:

163
When light enters another medium it ________________ .
A reflects B absorbs C bends D glows

caused 89.5% of the students to choose an incorrect answer. This could

have been because of either a problem with English or a problem with the
basic scientific concept included in this question. However, the research
carried out on the physics and chemistry texts suggest that this item is
very frequent in just such a context.
84% could not identify the pronunciation of regular past tense
verbs in a question like:

32. The ed ending in “showed” is different from the one in

A studied B remembered C caused D asked

More than half of the students (52%) could not identify the sounds
of the alphabet in the following:

The letter “A” does not contain the same sound as __________ .
A “J” B “K” C “H” D “Q”

This item was considered fundamental because of the various

formulae and descriptions of scientific objects which include reference to
the shape of letters such as a “T-shaped lamina” and “J curves”. Added to
this, if the European Threshold Level (van Ek 1976) for English in schools
is taken into consideration, this is found to be one of the items that has
been singled out as essential learning for secondary school students.
The items which caused the students the least number of
difficulties (25% and less) were pronouns (23%), questions using an
auxiliary (23%), telling the time (24%), present perfect (22% and 25%), the
superlative (20%), the past continuous (21%), future perfect (17% - the
lowest number of errors on this test) and the phrasal verb “get over”
(21.5%).

164
 There would appear to be some strange discrepancies here. If
students find phrasal verbs as easy as (or easier than) telling the time or
the present perfect, and if the future perfect is easier than questions
using an auxiliary, something is apparently going wrong somewhere,
given that, in the first test, 62% of students had difficulty with the future
perfect question and 54% and 62% with the phrasal verb questions.
Although it is difficult to say what the exact cause of these phenomena is,
it may be attributable to the fact that more emphasis may have been
placed on what is considered difficult in previous teaching/learning
situations and so the students have fixed these items better. The
similarity or difference between English and Portuguese in some of these
structures, such as the future perfect, may explain that what is difficult
for other students is not necessarily so for Portuguese students because
of similarities between the languages and vice versa with other structures
such as the present perfect which is not used in the same way in the two
languages.
It is even possible that the idea of what is difficult for students to
learn is in fact incorrect. McDonough (1980:311) says

“psychologists have objected that there is no reason to assume that

linguistic complexity is itself a cause of learning difficulty because many
constructions that appear complex in terms of counts of elements or
underlying rules are used by native speakers with no hesitation or greater
difficulty in execution than apparently simpler ones, in appropriate context
…. This is not to deny that constructions do differ in complexity and
learnability, rather it is to claim that the only measure of learnability is
actual learning and not predictions derived from linguistic description
alone.”

165
 Nevertheless, in terms of course design it does indicate that certain
“basic” items cannot be ignored if 25% of the students cannot cope with
them adequately, nor should we assume that time has to be spent on
teaching lists of phrasal verbs when the students are, in the majority,
able to cope with the more common ones adequately. Indeed, work done
on corpus studies for the COBUILD project suggests that six common
verbs account for nearly 30% of the ‘most important’ phrasal verbs as
mentioned earlier. The use of corpus studies to decide many such
questions for course and materials design is becoming increasingly
important (see Wichmann, Figelstone, McEnery, Knowles (eds. 1997),
Biber, Conrad and Reppen (1998), McCarthy and Carter (1994), Stubbs
(1996))

4.1.5 Test Results 1995/96

The style of the test continued as before with adjustments being

made in the light of the previous year’s test and with a change in the
topics for the paragraph. The topics for this test were:

i) The most important discovery this century.

ii) The changes we will see next century.

Only two topics were offered, because when three topics had been offered,
one was invariably largely ignored. These topics required the students to
use either the past tense as in (i) or the future as in (ii).
The fifty multiple choice questions gave the following error percentages:

Table 4.1 Analysis of 1995/96 Test Results by Item.

Test Item Percentage Test Item Percentage Error
Error
Pronouns 33% Conditional 52%
Present Perfect 40% wish + past perfect 37%

166
 Question inversion 24% Adjective + enough 27%
Reciprocal Verbs 25% Future Perfect 17%
Present Perfect 22% Subjunctive 91%
Telling the Time 21% Passive 24%
Present Perfect 11% Pronunciation 81%
Simple Past 16% Modal verb 60%
Comparative 17% Pronunciation 47%
Adjective Alphabet
Time Clause 25% Conditional 32%
(advice)
First Conditional 34% Reciprocal Pairs 84%
Superlative 15% Relative Pronoun 25%
Advice (had better) 94% Preposition 83%
Direct Object (lack 74% Phrasal Verb 50%
of)
Modal 21% Conjunction 36%
Second Conditional 47% Vocabulary 54%
Infinitive 56% Vocabulary 84%
Indirect Question 36% Vocabulary 67%
Modal (past) 32% Possessive Pronoun 37%
Future Continuous 73% Preposition 61%
Irregular Verbs 74% Vocabulary 15%
Past Continuous 18% Vocabulary 32%
Past Perfect 46% Vocabulary 49%
Past Tense 39% Reciprocal Pairs 59%
Phrasal Verb 67% Reciprocal Pairs 44%

These results continue to demonstrate the difficulties that the

majority of the students have with specific vocabulary, prepositions,
phrasal verbs, the conditional, giving advice, infinitive (as opposed to

167
gerund), modals, and pronunciation recognition. Perhaps more surprising
on this test is that one of the questions on the present perfect proved to
be an inadequate discriminator in that it was correctly handled by almost
90% of the students taking the test. This particular question was the
following:
7. She ____________ to England
A. have never been B. has never be C. has never been D. have never be

However, the following present perfect question caused 40% of the

students to err:
2 They ___________ their cousin since last year.
A. haven’t seen B. aren’t seeing C. didn’t see D. don’t see

Giving advice using “had better” was the question 94% of the students
got wrong. The difficulty here is almost certainly the fact that the full
form “had better” as opposed to the contracted “’d better” was given. This
allowed confusion between “would better” and “had better” in the
distractors. The question was the following:

13. You _________ take an aspirin for your headache.

A. had better B. would better C. will better D. have better

4.1.6 Test Results 1996/97

The test for this academic year continued much as before with one
or two modifications such as the topics for the paragraphs which were
changed to:
i) The worst dangers of pollution.
ii) What the world will be like after the year 2000.

168
Table 4.2. Analysis of 1996/97 Test Results by Item.
Test Item Percentage Test Item Percentage Error
Error
Pronouns 7% Conditional 38%
Past Tense 35% wish + past perfect 36%
Question inversion 27% Adjective + enough 25%
Reciprocal Verbs 29% Future Perfect 22%
Numbers in Words 78% Subjunctive 89%
Telling the Time 38% Numbers 29%
Present Perfect 58% Pronunciation 87%
Simple Past 25% Modal verb 50%
Comparative 31% Pronunciation 53%
Adjective Alphabet
Time Clause 23% Conditional 32%
(advice)
First Conditional 33% Reciprocal Pairs 67%
Spelling 19% Graeco-Latin Plural 69%
Advice (had better) 89% Preposition 92%
Direct Object (lack 76% Phrasal Verb 48%
of)
Possessive Pronoun 15% Conjunction 28%
Second Conditional 43% Vocabulary 52%
Infinitive 50% Vocabulary 92%
Indirect Question 58% Vocabulary 74%
Modal (past) 26% Possessive Pronoun 39%
Future Continuous 70% Preposition 64%
Irregular Verbs 72% Vocabulary 53%
Conjunction 31% Vocabulary 24%
Past Perfect 44% Vocabulary 46%
Comparative 35% Reciprocal Pairs 64%
Phrasal Verb 56% Reciprocal Pairs 43%

169
 Many of the results on this test continue to confirm what had been
found in previous tests but the inclusion of new items, such as numbers
in words, tested competence in other areas which pertain ever more
closely to the students’ future studies. These most likely were not taught
at all in school. The question on numbers in words includes the different
use of the comma and full stop between Portuguese and English. The
comma represents a decimal point and the full stop division into
thousands in Portuguese and vice versa in English.
The following question caused 74% of the students difficulty:

5. The number 1,711 reads ______________

A. one comma seven hundred and eleven
B. one point seven double one
C. one thousand, seven hundred and eleven
D. one comma seven double one

Similarly, the Graeco-Latin plurals were tested and showed that the
students had difficulty here too. Once again this is probably because
these plurals had not been specifically taught.
The following question caused 69% of the students difficulty:

37. Individual teachers may use different ________ when marking the test.
A. Criterions B. criteria C. criteriae D. criterii

4.1.7.Test Results 1997/98

A change was made in the Preliminary Test in the academic year

1997/8. It was decided that, although the students who were allowed not
170
to take the course undoubtedly had a good command of English, there
was little proof that these same students could indeed manage the
specific English being taught on the course and could show adequate
comprehension of a science text. Therefore, the Preliminary Test was
written in a similar way to the final examination for the Ano Comum and
contained forty multiple choice questions on the same type of items
tested at the end of the year together with a text for comprehension.
The results of this test showed overwhelmingly that the students
who were exempted from further study either did better or at least as well
on the comprehension passage as on the multiple choice questions. 75%
did better and 25% the same, of those students who obtained a grade of
fifteen or more on the Preliminary Test. The students at the bottom of the
scale (those scoring seven or less) showed the exact opposite of the above,
that is 75% did worse on the comprehension than on the multiple choice
questions and 25% achieved the same score.

Table 4.3. Analysis of 1997/98 Test Results by Item.

Test Item Percentage Error Test Item Percentage Error
Punctuation 6% Pronunciation 93%
Fahrenheit Scale 47% Comparative 46%
Numbers in Words 71% Passive 28%
Conjunctions 55% Modal 53%
Passive 35% Past Tense 54%
Fractions 35% False Friends 90%
Formulae 44% Pronunciation 22%
Alphabet
Dates 35% Question Form 20%
Conjunction 55% Pronouns 31%

171
 (contrast)
Conjunction 15% Graeco-Latin Plural 79%
(reason)
Conjunction 48% Relative Pronoun 17%
(reason)
Conjunction 23% Vocabulary 77%
(contrast)
Conjunction (cause 32% Vocabulary 75%
and effect)
Pronunciation 35% Conditional 26%
Vocabulary 43% Metric/Imperial 69%
Equivalence
Conjunctions (cause 39% Reciprocal Pairs 67%
and effect)
Superlative 52% Conjunction 17%
(contrast)
Adjectives 71% Translation (false 51%
friends)
Pronunciation 25% Vocabulary 85%
Comparative 80% Conjunction 28%

Other changes can be noted on this test. The pronunciation

question on the alphabet only caused 22% of the students to stumble
compared with 53% in 1996/7.
There is some evidence to suggest that after so many years
administering this test and the course itself there is a change taking
place within the academic community. Teachers from outside the
university have asked for copies of the Preliminary Test in order to
prepare students to take it, an idea which is contrary to the whole
philosophy of the test, which was to find those students whose English
was already of a high level and who could already cope competently with
the demands made upon them by their bibliographies in English in their

172
subjects areas. This shows the detrimental competitive nature of
educational systems which can lead to emphasis often only being given to
the grade obtained, rather than to ability or performance. A change of
emphasis would focus on cognitive knowledge and would preclude
specific ‘teaching for the test’ to be given for a certain test score to be
achieved.
One other area that has obviously continued not to be given much
stress in school curricula is numbers2. As was the case in 1996 with 74%
of students mentioned earlier, 71% of the students in 1997 could not
handle the correspondence between a number and its form in words.
Of almost equal difficulty (69%) was metric to imperial
equivalences, in this case recognising the nearest equivalent to 100 yards
in metres. This, as was mentioned earlier, may have nothing at all to do
with the language involved but be much more a question of cultural
knowledge, recognising the difference between measurement systems in
different countries. Although students studying in the Ano Comum were
not expected to learn the equivalent measurements and conversions of
metric to imperial measurements and vice versa, because this
information is readily available in a good dictionary, students were
expected to have some idea of the relative sizes so that logical
assumptions could be drawn. To take an example: If a student were faced
with the sentence Scientists can calculate the distance of the earth to the
moon to within six inches. The student should recognise that six inches is
not the distance from the earth to the moon as this is equivalent to a
distance of approximately 15 cm. and, therefore, this sentence must
rather be a discussion of accuracy of measurement and not the actual
measurement of the distance mentioned.

2
A similar test was tried on the students in the fourth year of their teaching degrees which showed that these
future teachers also had difficulty with numbers and were often unaware of the contrasts between the
English and Portuguese use of the comma and point in numbers. It is not surprising, therefore, that those
students who had only studied English in school should find this item difficult, a situation which is
unlikely to change significantly in the near future.
173
 Similarly, 80% of the students could not distinguish between the
relative sizes of a British billion and an American billion, (1012 and 109
respectively). Students need to have the ability to question such items
and not merely to assume that the same thing is meant by all those using
the same word, in this case the word billion. However, it is arguable that
this item would in fact cause any difficulty for these undergraduates
because in Portuguese um bilião stands for a thousand million exactly
like the American measurement and the aberration here is the British
measurement of a million million which may be on the point of fading out
of use3.
Other frequent items like the adjective wide caused considerable
difficulty (71% of students got this item wrong). This item appears in the
frequency list 1515 times in 1876 articles and in both the physics and
chemistry corpora studied here, which would suggest that this adjective
is essential for undergraduates and has not been learned by almost three
quarters of those entering the university.
False friends (eventually - eventualmente) and specific vocabulary
(clerical work associated with offices not the clergy) items were the worst
overall items causing 90% and 85% respectively of the students to make
errors. Graeco-Latin plurals also caused considerable error (79%)
although these are often seen to be significant in scientific writing (see
later results 5.1.6 Plurals from Latin and Greek for a further discussion of
this).
Conditional sentences and question forms did not appear to cause
undue difficulty for most of the students who took this test with 26% and
20% making errors on these items. Some linking devices and deictic
pronouns caused more difficulty than others; this could be attributed to

3
Recently (BBC World Business Report 8/4/99) even British use (certainly in economics) has tended to
favour the American thousand million. The answer, aired on the BBC World Business Report programme,
to an e-mail inquiry from a viewer confirmed that the BBC were in fact using the American billion in their
reporting.

174
their being relatively unusual. Only 17% of the students had difficulty
with these but 55% had difficulty with thereby and whereas.
It cannot be taken for granted that the students do not know some
items included in ‘advanced’ grammar books, nor can it be assumed that
the basic structures have been learned soundly. What these results do
suggest however is that it would be of more use, for example, to include a
list of irregular verbs for the students to take away and learn rather than
a list of phrasal verbs. In other words, the students’ proficiency profile
must guide the choice made about what language should be taught in
the first year.
In addition, corpus analysis (on the LOB and the Brown corpora)
for a syllabus for use with students in Germany by Mindt (and Tesch
1990) (1997:40-50) has shown that with careful grading students can
learn a higher percentage of the most frequent and therefore important
irregular verbs (apart from be, have and do) even if they stop learning
after a short period of time. They contrast their list with alphabetical lists
of the kind normally presented for students to learn and show that after
learning five of the verbs (say, make, go, take, come) on their list the
students will be “familiar with 27 per cent of all irregular verb forms of
English”. The corresponding figure for the alphabetical list is 3.6 per cent
(beat, become, begin, bet, bite). After learning ten of the verbs on their list
(say, make, go, take, come, see, know, get, give, find) the student “has
mastered 45.6 per cent of the verb patterns of irregular verbs”. The
combination of what the student already knows and the results of
corpora analyses like the one described here can provide a much more
reasoned syllabus that aims to make maximum use of the students’
study time. Mindt (1997)and Tesch (1990) like Renouff (1984) also find
that their corpora analyses show the discrepancies that occur between
actual language use and what is presented in coursebooks. (Tesch
studied some and any, Mindt modal will and would and Renouff see).

175
These are taken up in more detail later in Chapter 7. Similar research on
the corpora shows that certain irregular verbs are more suitable for the
students of science and technology (see 6.3) and would cover the
difference in the transition from secondary to tertiary education for
undergraduates.

4.2 Needs Analysis by University Department

In order to meet the demand for English language teaching for the
Ano Comum in accordance with what the departments whose students
are involved in the foundation year feel is necessary, a simple needs
analysis was requested from colleagues in other university departments.
Colleagues were asked to indicate their views as to why they thought it
necessary for the students to study English language for their courses.
This was the first stage in the needs analysis, the results of which are
given below.
Colleagues in other departments asked that the students be able to
speak fluently, read fluently and listen effectively. Below is a
representative sample of what our colleagues in other departments
perceive as the English needs of their students taken from the replies
received to the initial simple needs analysis which consisted of a letter
from the English Area Co-ordinator to the different departments involved
in the Ano Comum asking for their comments. Nine replies were received
relating to different courses included in the foundation year. Not all
departments replied. The replies received were from the co-ordinators
responsible for the following courses: Mathematics (Teaching of); Applied
Mathematics and Computation; Geological Engineering; Biology; Biology
and Geology (Teaching of); Ceramics and Glass Engineering; Materials
Engineering; Engineering and Industrial Management.

176
• Inglês coloquial

• Inglês técnico (Matemática, Informática)

• Tradução corrente de textos científicos, uma vez que grande parte da

bibliografia que é indicada encontra-se em inglês.

• O domínio da língua inglesa é essencial na formação dos nossos engenheiros.

Consideramos que a disciplina de inglês dever ter como objectivo, entre outros,

a preparação dos alunos para as seguintes situações: - Dialogo fluente com

profissionais da área das engenharias nos contactos internacionais; - Leitura

de manuais, livros e catálogos técnicos, etc.

• Competência na língua de modo a permitir os alunos: -"enfrentarem-se" com

textos técnicos -conseguirem dialogar.

• Os alunos devem dominar os termos técnicos das áreas de Turismo, Economia e

Gestão. Porque temos frequentemente docentes estrangeiros, surge já durante o

curso a necessidade dos alunos de comunicar em inglês nas aulas.

• Leitura, interpretação e conversação nas vertentes, ciências naturais

(geológica), física e química

The third point given in this needs analysis, that of being able to deal
with bibliographies that were largely in English, was taken as particularly
pertinent to all of the students involved in the foundation year. The aims
put forward by these replies are obviously ideal and would stand the
students in good stead for their futures both in terms of further study
(possibly abroad) and in their careers. However, reaching the ideal is

177
limited by a number of constraints which few of these same respondents
cared to acknowledge in the curriculum they had created.

4.3 Constraints

There were three major constraints on the teaching of the students

in the first year of the university, which are described below.

Several administrative decisions were taken about the size and

number of groups to be taught in the first year. These constraints were
common to all of the subject areas included in the first year course as far
as was possible (some differences existed for practical classes in physics,
chemistry and computer studies).
The first decision was about the size of the groups. There were to be
at least twenty-five groups of forty-five students each. This soon became
at least twenty-eight groups because of the students who were repeating
the disciplines. Different systems were adopted to accommodate these
students, from separate classes being created to distance-learning
courses on computer being provided in 1998-99 for those students
repeating the year4.
Secondly, all of the English language students were to receive two
hours of language teaching per week as theoretical-practical classes.
Previously some courses, such as Tourism and Management, had had far
more English over two or three years, so that this represented a
considerable reduction in time for some courses. In essence, the non-
pure or physical sciences received less language instruction time.

4
The distance-learning courses are run by the University of Aveiro through the Internet and are open to all
students who must enrol and obtain a login to be able to access the relevant material. Working students
are obviously targeted by this scheme of work besides those repeating disciplines. Other systems, like
course tutor support schemes have been instigated to help students to plan their studies and cope with the
psychological strain of the move to a university environment which studies (Tavares et al. 1996)
mentioned earlier had shown to be a reason for lack of success in the first year.

178
 The third constraint was that the students on the course were not
homogeneous, either in terms of the number of years they had studied
English (see earlier Chapter 3, 3.2.1 The Students’ Level of English) or in
terms of the subject which they had chosen to pursue for their degree.
The latter was again a change, as in the past some courses had had
English as a discipline and these groups had been homogeneous in terms
of their degree subject, for example, in Management and Electronics. This
meant that subject specific coursebooks such as English for Electronics
could be used for the appropriate group but that now this was no longer
possible.
The overall percentages of students with more than five years of study
of English has gradually been creeping up from the initial 67% with five
years or fewer of study and 33% with more than five years of study, to a
complete reversal of this situation in the academic years 1997/98 with
32% with five years or fewer of study and 68% with more than five years
of study as mentioned earlier in section 3.2.1. This trend could be
accounted for in a number of ways; the information about the foundation
course year may well have become better known among those students
who were hoping to avoid language study by opting to take up other
courses, in other places, although this seems unlikely; it may be that the
a general trend in the secondary schools to teach more English has
filtered through to higher education; or it may simply be a reflection of
the fact that all the students entering the science and technology courses
have now decided to try their luck at the preliminary examination and so
these figures are a much more complete representation of the student
intake. A trend towards more years of English in secondary schools is a
positive development as it will only help students at university to come to
grips with studying through English language textbooks.
Nevertheless, the aim of the foundation year to give all of the student
intake into science and technology courses a sound basis on which to

179
build their studies in later years and to maximise their learning potential
appears to be being subverted. It seems to have become another hurdle
for students to jump so that strategies like evasion, gamesmanship and
cramming are encouraged. It is interesting to speculate whether students
would voluntarily study English if their departments recommended but
did not insist on it. It is possible that quite sophisticated translation
services would be set up on the periphery of the university if certain texts
were seen to be essential and were only available in English. Students
traditionally claim that they have too little time for their studies and this
would be one means to avoid having to study English at all. This line of
thought leads directly on to the subject of motivation in the students
once again. If the subject is only seen as a hurdle to be jumped, the
students’ focus will necessarily be on test results rather than on
achieving their maximum potential. This means that the course will have
to subversively achieve this end against the wishes of those students who
take this position. Students who are at the bottom of the scale with few
years of English may also be discouraged from the outset or may decide
to take further courses in English outside the university. One means of
persuading the students to focus on learning more English is first and
foremost by encouraging them to attend most of the classes. This has
been successfully achieved by linking evaluation with class attendance.
Students are offered different evaluation schemes depending on whether
they attend a minimum of two thirds of the classes given. This is quite a
normal procedure in the university for practical classes and those that
use continuous evaluation of students so there is no real difficulty in
applying continuous assessment to these classes. One other effect of this
scheme is that the need for hundreds of individual interviews,
traditionally taking twenty minutes each with all the practical
administrative difficulties that this implies in a short examination period,
is avoided.

180
 The other means to try to engage the students is to make the work as
appropriate as possible for them. This can be approached through the
analysis of the language of science and technology and the language
requirements of the books in English on the bibliographies which they
have to deal with. Chapter five will take up this aspect of the students’
language needs.

181
Chapter 5 Scientific English for
Undergraduate Learners

Analysis of Results
 Chapter 5

Scientific English for Undergraduate

Learners

5.1 Analysis of Results

This Chapter will examine the English of science and technology

the undergraduate students will need to cope with from the baseline
corpus and the specific corpora developed from the undergraduate
textbooks in physics and chemistry. The features, outlined in Chapter 3,
that are seen as being salient in these texts and for the profile of the
undergraduate analysed in Chapter 4 will also be discussed in the light of
the data obtained. The results will be compared and contrasted with
Biber’s findings for academic prose and the significant features will be
highlighted. A more detailed analysis will be made of the sub-corpora
from the Physics and Chemistry textbooks and consideration will be
given to the role of mathematics in these texts.

5.1.1 The Baseline Corpus

The Grolier Multimedia Encyclopaedia provided the following words

from approximately 31,000 texts; a total of approximately 150,000
different words from a corpus of several million words. The figures given
after the word refer to the number of articles containing the word, the
range of the word, and the frequency of the word, respectively. Thus, it
can be seen that the actual number of words contained in the
Field Code Changed
183
encyclopaedia is a multiple of the 150,000 different words listed. Wilks et
al (1996:189) describe the Grolier as having 10 million words.1
A ‘law of diminishing lexical returns’ has been put forward which
says that as the size of a corpus increases, the average incidence of word
types decreases. The one million word LOB corpus yields about 50,000
different word types and the Birmingham main corpus of 7.3 million
words mentioned earlier contained only 132,000 different word types.
Only frequencies and ranges of 100 or more have been included
and all proper names have been excluded where possible. However, there
are instances where a proper name and a noun are not immediately
distinguishable, for example, hill and Hill. Those words considered
equivalent or close to a similar word in Portuguese with the same
meaning, that is cognates, have been marked with an asterisk. The
Grolier encyclopaedia does not include a number of words on the
grounds that they are ‘too common’2. A sample with respect to the letter
‘a’ is : about, across, after, again, all, along, also, although, among, an,
and, any, are, around, as, and at. Two other 50,000 word corpora, taken
from the Chemistry and Physics textbooks Chemistry (Chapters 2, 3, 4
and 5 Pages 37-219) and Physics for Scientists and Engineers with Modern
Physics (Chapters 1-5, Pages 15-216)3 in the students’ bibliographies,
have also been examined to produce frequency counts of the same kind
and then compared with the multimedia encyclopaedia in order to
examine the similarities and differences between them.

1
Minugh (1997:79) in his use of Newspaper CD-ROMs for teaching at the University of Stockholm also
mentions this lack of information about the actual size of published CD-ROM material to be one of the
disadvantages, although he hopes that the companies involved in producing them can be persuaded to
incorporate this information in the future. He further laments the fact that “the most frequent words are
classified as “noise” and cannot be searched for.”
2
These are sometimes referred to as structure words of which it is estimated that there are about two
hundred (see Bowen, Madsen and Hilferty (1985:194)) or form words (see earlier 3.1.7 Bright and
McGregor 1970)) or function words (see Biber et al. (1998:29)). However, it is clear that the Grolier
includes any very common word and not only words such as articles, prepositions, and pronouns.
3
Sinclair recommends that the same areas of books are not studied in case they demonstrate only one
specific variety of English, for example the English associated with introductions or first chapters. See
later bibliography for a similar justification of Sinclair’s hypothesis.
184
 When an item is to be found in these texts the word is marked with
the text which contains it. For example, able occurs in both the
chemistry and the physics corpora and so this word is marked with the
letters C and P after the range and frequency figures given in the
multimedia encyclopaedia listing.
Some interesting anomalies occur with the words listed including
those given in the Grolier encyclopaedia as ‘too common’. Under the letter
‘a’ across and around do not appear in the 50,000 words of the chemistry
text at all, under ‘b’ be, became, become, been, begun, being, and
bibliography do not appear in either of the corpora and only by appears in
both corpora. It is not surprising that the word bibliography should not
appear in these corpora as the bibliographies were not included in the
corpora as they traditionally occupy a position at the end of the textbook
which was not taken to be representative of either physics or chemistry
texts per se.
The Grolier’s idiosyncratic pronunciation scheme is also to be found
in the list. Each of these items is identified by the abbreviation (pronun),
for pronunciation, immediately after the entry.

Table 5.1 Grolier Frequency and Range List

Field Code Changed

185
able 827/1066CP *acids 227/483C advances 289/376P
ability 753/996CP acknowledged 129/135 advantage 283/343CP
abilities 111/127 acquire 101/111P advantages 142/165
*abnormal 116/173 acquired 425/519P advent 183/193
aboard 131/204 acquisition 111/135 adventure 114/140
abolished 177/207 acres 167/242P adventures 170/204
abolition 110/127 *act 1197/2198CP advertising 111/216
above 192/1615CP acted 166/186 advice 116/133
abroad 293/357 acting 327/445C advisor 160/170
absence 262/296P *action 981/1452C advocate 249/265
absent 123/136P *actions 297/379 advocated 267/293
*absolute 308/486CP *active 1094/1384CP aesthetic 224/300
*absorb 147/182CP *actively 130/135 affair 175/205
*absorbed 308/408 *activities 826/1119 affairs 546/689
*absorption 191/328 *activity 937/1356CP *affect 236/282CP
*abstract 328/554 *actor 300/502 *affected 367/450C
*abundance 137/155C *actors 124/236 *affecting 123/136P
*abundant 405/535C *actress 168/253 *affects 158/184P
*abuse 126/221 *acts 561/786CP *affiliated 116/120
*academic 273/389 actual 388/449CP afterward 128/134
*academy 756/1120 actually 537/625CP against 2860/4578C
*accelerated 150/175P acute 165/232 age 2141/3573P
accept 282/320C ad 1039/1563CP aged 106/139
acceptance 219/247 *adaptation 163/186P *agencies 241/378
accepted 649/768P *adaptations 120/151 *agency
*access 283/378 *adapted 413/517CP 364/600
*accessible 109/116 add 140/174CP *agent 280/386
*accident 137/169 added 798/1015CP *agents 255/425C
*acclaim 174/182 *addition 1334/1818CP ages 847/1172
*acclaimed 195/229 *additional 518/603CP *aggressive 174/213
*accommodate 111/118 address 105/143 aging 101/167
*accompanied 382/424P addressed 101/111 ago 528/919P
accompanying 112/123 *adequate 196/231 agree 134/146P
*accomplished adjacent 310/368P agreed 301/378
339/376CP *administered 250/320 agreement 307/426CP
*according *administration 876/1271 agreements 108/149
1117/1392CP *administrative 452/602 *agricultural 892/1501
account 617/824CP *administrator 151/162 *agriculture 817/1584P
accounts 311/398 *admiral 122/174 ah (pronun) 405/415
*accumulated 116/130 admired ahead 104/117P
*accumulation 113/129 285/310 ahl (pronun) 106/107
accuracy 185/256P *admission 122/162 ahn (pronun) 167/169
accurate 248/309CP *admitted 185/206 ahr (pronun) 101/102
accurately 130/150CP *adopt 110/118 aid 681/1014
accused 214/238 *adopted 802/1001 aided 274/317
achieve 411/468C *adoption 143/168 aids 146/277C
achieved 972/1196 *adult 544/802 aim 155/175
achievement 320/384 *adults 289/383 aimed 197/214
achievements 246/278CP advance aims 111/133
achieving 118/120 277/384 air 1581/3580CP
*acid 503/1213C advanced 632/826P aircraft 304/885P

186
airfields *animals 1122/2441CP *architectural 392/607
177/186 *annexed 186/207 *architecture 1028/2856
airplane 115/182CP announced 233/277 *archive 978/2752
airport 153/206P *annual 1033/1512C *area 2756/4964CP
*album 118/191 *annually 410/474C *areas 1820/3630CP
*alcohol 221/440C another 2209/3383C argued 327/394
*algae 152/340 *anthology 113/134 argument 111/148P
alive 132/140 *anthropology 102/218 arid 201/298
alleged 123/134 *anti 448/599 arise 183/205P
*allegorical 101/132 *antibiotics 119/197 *aristocracy 123/145
*alliance 316/472 *anticipated 133/143 *aristocratic 157/184
allied 407/685 *antiquity 124/136 arm 215/270P
allies 282/465 anything 114/121 armed 311/413
allow 362/411CP apart 320/368CP armies 249/454
allowed 589/708C apparatus 117/152CP arms 362/635
allowing 231/253CP *apparent 368/431 army 1255/2496P
allows 280/319CP apparently 429/488P arose 276/318
alluvial 108/149 appeal 210/261 aroused 154/159
almost 1735/2419CP appeals 106/134 arranged 361/410C
alone 421/471C appear 697/872CP arrangement 269/335C
*alphabet 101/247P *appearance 647/777C arrangements 180/217CP
already 546/653CP *appearances 102/107 array 155/181
alter 111/123CP appeared 1049/1335C arrest 112/139
*altered 198/223C *appearing 125/129C arrested 190/204
alternate 126/137 appears 553/643C arrival 233/261
alternating 145/190 *application 362/470C arrived 338/420CP
*alternative 230/276CP *applications 359/553CP *art 2573/7631CP
*altitude 239/359CP applied 1011/1315CP article 183/278
*altitudes 116/143C applies 103/107CP articles 314/421
aluminum 287/495CP apply 159/183CP *artifacts 155/192
always 783/959CP applying 147/158CP *artificial 456/672
amateur 146/227C appointed 920/1078 *artisans 117/141
*ambassador 149/181 appointment 168/187 *artist 683/
*ambitious 157/172 approach 551/783CP *artists 665/1236
amendment 251/691 approaches 205/265CP *arts 1051/1704
amino 103/286 appropriate 327/380CP ash 136/185
amount 657/1036CP *approval 150/194 aside 152/169
amounts 490/690CP *approved 246/296 asked 152/162CP
*analysis 572/894CP *approximately *aspect 224/255P
*analytical 106/129C 842/1073P *aspects 498/617CP
analyzed 124/144C Apr 1651/1796 *assassinated 183/217
*anatomy 199/306 April 508/678 *assassination 177/234
ancestor 152/173 *aquatic 170/259 assembled 106/122C
ancestors 154/192 arc 170/314P *assembly 529/862
*ancestral 105/136 arch 112/209 asserted 116/134
ancestry 120/144 *archaeological 263/349 assigned 211/230C
ancient 1900/3090 *archaeologists 117/152 assist 114/121P
angle 190/357P *archaeology 189/368 assistance 237/326P
angles 165/264P *archbishop 147/201 assistant 259/289P
*angular 126/188 *architect 508/761 assisted 157/165
*animal 1002/1793CP *architects 236/382 *associate 182/198
Field Code Changed

187
*associated 1197/1526CP attempted 560/658P *autobiographical 217/264
*association 763/1080 attempting 150/157P *autobiography 496/562
*associations 150/193 attempts 613/723P *automatic 133/219
assume 142/151CP attend 113/137 *automatically 115/139
assumed 548/646CP attended 327/343 automobile 289/420CP
assumption 123/154CP *attention 689/842 automobiles 186/237C
*astronomer 185/253P *attitude 116/155 *autonomous 190/252
*astronomical 193/310P *attitudes 177/230 autumn 153/199
*astronomy 302/633P *attorney 209/280 availability 100/108
athletic 104/151 *attract 205/242C available 825/1150CP
*atlas 171/281 *attracted 397/436 avant 176/217
*atmosphere 469/977C *attraction 161/187C average 975/1647CP
*atmospheric 241/377C *attractions 101/109C averages 226/250
*atom 214/594CP *attractive 173/204C averaging 116/130
*atomic 440/1037CP *attributed 271/296 avoid 228/266CP
*atoms 342/971CP audience 266/394 award 468/655
attached 461/606CP audiences 161/227 awarded 411/488
attack 526/754C Aug 1710/1875 awards 214/291
attacked 339/403 August 562/799CP aware 126/138CP
attacking 109/124 *author 732/909 awareness 138/155
attacks 328/425 *authorities 310/366 away 670/825CP
attain 126/130CP *authority 646/1000 axis 230/454CP
attained 172/187 *authorized 147/169 ay (pronun) 250/256
attempt 703/847C *authors 186/254

The lists obtained for the rest of the alphabet are given in Appendix
B with these Portuguese cognates, as defined above, removed. The words
not included, on the grounds that they are too common, have been shown
in italics at the end of the respective list for each letter of the alphabet.

5.1.2 Range and Context

The search facilities mentioned can provide some relevant data on

the range of an item. If a comparison is made between the adjectives long,
wide, broad, deep, tall and high and the nouns length, width, breadth,
depth and height it is possible to demonstrate the considerable differences
in the topic areas they are most frequently to be found in. Here is the data
for the most frequent uses of these adjectives and their respective nouns
in the CD-ROM encyclopaedia:

188
Table 5.2 Frequency and Range Results for Abstract Nouns and Adjectives.
Length Long
1133 Articles - 1573 Occurrences 4041 Articles - 6701 Occurrences
20 – measurement 46 - plant
17 - metric system 45 - mammal
13 – lens 40 - bird
11 – fish 24 - flower

Width Wide
175 Articles - 216 Occurrences 1515 Articles - 1876 Occurrences
7 – dendrochronology 10 - sound recording and reproduction
5 - river and stream 9 - ice hockey
8 - television
7 - mammal

Breadth Broad
38 Articles – 38 Occurrences 651 Articles - 802 Occurrences
1 – anthropology 11 - plant
1 – dimension (mathematics) 7 - antibiotics
5 - Antarctica
5 - mammals

Depth Deep
350 Articles – 496 Occurrences 829 Articles - 1177 Occurrences
13 – depth charge 36 - deep sea life
10 – perception 13 - ocean and sea
8 – water wave 13 - syntax
7 - gulf and bay 9 – geosyncline

Height High Tall

563 Articles – 745 Occurrences 3145 Articles - 365 Articles -
5679 Occurrences 485 Occurrences
14 – plant 49 - secondary 25 - plant
education
14 – tree 30 - middle schools 11 - tree
& junior high school
10 – atmosphere 22 - sound recording 10 - flower
& reproduction
10 – statistics 21 - nuclear reactor

Field Code Changed

189
 At least two of these areas will be unknown to the average
humanities trained teacher: dendrochronology and geosyncline.
Dendrochronolgy is the science of using tree rings to date structures and
events or to reconstruct past environmental conditions. A geosyncline is a
large, usually elongate depression in the crust of the earth which during
subsidence has accumulated very great thicknesses (thousands of meters)
of sedimentary and usually also volcanic, rocks. The latter is therefore a
part of geology.
From these results it is possible to conclude that the presentation
and practice of these adjectives and nouns in teaching materials would
have to be through different contexts or settings if a ‘natural’ use of such
items was to be given. Length should, from this perspective, be presented
in a physics context for example, whilst long would be more ‘naturally’
presented in a context of biology. The noun breadth has such a low
frequency that it could be ignored but broad should be in a biology context
while wide seems more ‘at home’ in an electronics context. The results for
tall and high as ‘corresponding’ adjectives for height once again show that
different scientific settings are used with each; biology for tall and height
and physics and electronics for high, although height could also be
included in physics, chemistry or maths.
It would be possible for any item or sets of items to be examined in
this way to contextualise the setting of any of the vocabulary or syntax
that has been identified for the syllabus. This would then show the use
and meaning of these items in scientific settings, which as was discussed
earlier in Chapter 2 , does not necessarily correspond to the use or
meaning in everyday contexts.

5.1.3 American Words and Spellings.

It is interesting to note that several American words and spellings

(afterward, aging, aluminum, airplane, analyzed, attorney, authorized and
190
automobile) exist alongside some obviously British words and spellings
(advisor, aesthetic, autumn, car). A possible explanation for this is the
lack of a rigorous editorial practice as mentioned below (Burnard 1992).
Another explanation could be that the author of Chemistry, Raymond
Chang, is described as having been born in Hong Kong, growing up in
Shanghai and Hong Kong, studying in London and at Yale and teaching in
America. Perhaps these inconsistencies reflect his background in both
British and American academia. It is surprising however, that the editors
did not impose consistency on the finished work.
Students appear to have limited awareness of the differences
between British and American English (not to mention Australian, South
African, etc., of which only those few students who had studied in these
places would be aware). From the books in the students’ bibliographies
American English dominates and so familiarity with this form should be
encouraged. This may not necessarily be by contrasting it with British
English although some students may ask for such a contrast to be made
explicitly because of their previous learning.
This mixing of American and British English preferences can also be
seen in the earlier Thorndike and Lorge’s (1944) word list for teachers (in
America) which is meant to have an American bias but where words such
as the British trousers are more frequent than the American pants.

5.1.4 Abbreviations.

Three letter abbreviations for the months of the year appear much
more often than the full words for these in the frequency lists. Such
findings would obviously not imply that the students should only be
taught these abbreviations but they do imply that these features are the
natural ones to be included in the students’ study materials. An obvious
application here is in Tables, Figures or Graphs where abbreviations of
this type are to be expected. Field Code Changed

191
5.1.5 Pronunciation Conventions.

The Grolier encyclopaedia uses such sound groups as ‘ah’, ‘ahl’,

‘ahn’ and ‘ahr’ to illustrate the pronunciation of names and never uses the
phonetic alphabet. This finding seem to run counter to the teaching of
phonetics for referencing skills and the use of dictionaries, in particular
with the international phonetic alphabet being used to indicate
pronunciation in most dictionaries.4 It is also doubtful if these
pronunciation aids would really help Portuguese speakers as they would
not be pronounced in the way that this encyclopaedia obviously imagines.
In other words, this feature reinforces the idea that the encyclopaedia was
designed with a native speaker reader in mind.
The problem of whether to teach the phonetic alphabet so that
students’ can make use of it when using dictionaries revives the perennial
problem of how to make the most of the study time available. Dictionaries
usually present the phonetic alphabet at the beginning and simply
drawing attention to this fact (and conventions of representing stress)
allows those students who wish to, to pursue this potentially fruitful area
further. It is recognised however that unless some attempt is made to
engage the learners with this data most non-linguists will ignore it
altogether and thereby lose an opportunity to enhance and enlarge their
knowledge of English whenever the need arises.

5.1.5 Plurals from Latin and Greek.

Just as Sinclair (1991:68-9) discovered that one form of a lemma is

much more common than others, research into Latin and Greek plural

4
It is interesting to note that phonetic transcription is not usually included in bilingual and Portuguese
dictionaries which may be a reflection of the fact that there is a general idea that Portuguese is a ‘phonetic’
language and that, therefore, transcription is not necessary. This is an increasingly debatable proposition
especially with new spelling conventions coming into being.

192
forms in the Grolier encyclopaedia reveals that there is usually a
substantial difference between the frequency of either the singular or the
plural of Latin and Greek root words. Some simply fail to appear at all,
parentheses (10) is a case in point where the singular parenthesis does not
appear at all in the corpus. Occasionally there is a regular ending applied
to a Latin or Greek root word such as indexes together with indices and of
almost the same frequency (36 and 31 respectively). However, in the case
of formulae and formulas the latter is much more frequent than the former
(4 and 88 respectively). There are very few singulars and plurals that
appear in almost equal numbers; nova (10), novae (10) and novas (10) and
stimulus (146) and stimuli (122), all the others encountered show marked
preference for one or the other form.
Sinclair (1991:67ff) demonstrates that the of singular and plural
forms of nouns are not equivalent, by documenting the different
patterning of eye and eyes. He finds that “There is hardly any common
environment” between the two word forms and they “do not normally have
the capacity to replace each other”. The plural co-occurs with adjectives
such as blue, brown, covetous, manic. The singular hardly ever refers to the
anatomical object, except when talking about injury or handicap. The
singular and plural also occur in different sets of fixed phrases (all eyes
will be on, rolling their eyes, turn a blind eye, keep an eye on). It is this sort
of analysis which highlights the fact that lexis and syntax are totally
interdependent.

Similarly, in this encyclopaedia, the plural algae (339) is found to be

more frequent than the singular alga (28) and yet axis (452) is found to be
more frequent than axes (142). These findings reflect the fact that the
former is being used as a generic term whilst the latter is being used to
specify a particular part of a graph. It is essential to be able to make this
form of distinction so that these terms can be used in their most likely
contexts.
Field Code Changed

193
 Many authors have recognised the importance of Latin and Greek
terms in scientific texts but these have tended to be taught in lists of
singular and plural forms rather than in specific contexts where either the
singular or the plural would be most appropriate. The kind of language
manipulation exercise where students produce the singular or plural of a
Graeco-Latin word is generally considered unsuccessful as a
teaching/learning strategy and would ignore the semantic differences
inherent in scientific contexts. For example, in the Physics and Chemistry
corpora only two Latin and Greek singular and plural forms exist together.
These are: bacteria - bacterium and axis - axes. Formula - formulas are to be
found in the Chemistry corpus but no formulae. All other Graeco-Latin
words are only found either in the singular or in the plural only. For
example, analysis, apparatus, appendix, criterion, data, parentheses.
The Latin and Greek roots and affixes described by Strevens
(1978:193) given earlier in 3.1.8.4 are not found in their entirety in the
Physics and Chemistry corpora either. His cyto, plasma, pyro, ante, and
post, are found in neither the physics nor the chemistry corpora and many
of the other prefixes like hydro and poly are in combinations such as
hydrogen and polygon which are cognates with Portuguese. Similarly with
his suffixes, -ite is most often found in words such as white and write and -
valent in equivalent. The Grolier does present examples of all of the
prefixes and suffixes mentioned by Strevens but once again closer
examination shows that the majority of entries for the prefixes and
suffixes are not of the type anticipated by him. For example under the
prefixes hydro- and anti- the largest number of entries refer to hydrogen
and antiques. It would seem therefore from these results that these items
need not be heavily focused upon in the syllabus.
White (1998:275) argues for a distinction to be made between
science and technology texts. He (ibid.) says that science texts show
preference for non-vernacular Latin/Greek borrowings, whereas

194
technology texts prefer elaborated nominals where all the elements are of
vernacular derivation together with acronyms or provisional or ‘proto-
nouns’. A proto-noun is a word that is now commonly used as a noun but
was originally an acronym such as scuba (self-contained underwater
breathing apparatus) or laser (light amplification by simulated emission of
radiation). However, White (1998:285) claims that the “fact that classical
scholarship is no longer so widespread may offer a part explanation as to
why Greek/Latin coinings have declined” in science, even though “they
remain the norm”. The textbooks studied show a combination of both the
science and technology features that White found with the Latin/Greek
borrowings mentioned above, together with vernacular and proto-nouns
such as scuba.

5.1.7 Word preferences.

The word attempt is found to be approximately five times more

prevalent in the list than try and the word change is ten times more
frequent than alter. These choices, although apparently arbitrary, may,
according to Quirk (1995), indicate formality in the texts and be
representative of this particular genre

Bright and McGregor (1970:29) claim that technical texts contain

both technical vocabulary and a large number of ‘general’ words that are
“outside simplified English”. They further claim that there is “good reason
to suppose that anybody who is going to continue his education in English
will find them (general words outside simplified English) useful”. They list
twelve such words: ‘absurd, adequate, adjoining, aggression, alert,
alternative, amateur, ample, apparatus, apprehensive, automatic, available’.
Of these words only six do in fact occur with significant frequency in the
Grolier encyclopaedia (see word list in Table 5.1 above: adequate,
alternative, amateur, apparatus, automatic and available) and of those six
Field Code Changed

195
only three could be considered sufficiently different from Portuguese to
warrant attention (amateur, apparatus and available). It would seem from
this small comparison that it is true that the results found through
intuition and those found in reality through empirical research are
different. However, it could be that the genre studied here differs
considerably from that used by Bright and McGregor about which,
incidentally, they give no specific information.
As mentioned earlier, Hoffman (1981), Darian (1981), Weber (1981)
and White (1998) all argue that even the same lexical item can (and
usually does) take on new meaning when used in a scientific text. White
(1998:285) further suggests that the use of acronyms in technological
texts is because these are “eminently well equipped to meet the constant
need for new lexis to map the ever-unfolding reality of technological
development”. The fact that only part of the lexis of science and technology
can be seen to be in any way stable suggests that any study such as this
will only be representative as long as the textbooks are considered
sufficiently up-to-date for use with undergraduates. The textbooks that
have been selected for use in this study continue to be in the students’
bibliographies. Despite the fact that newer editions have recently been
published, these have not undergone any significant changes.

5.2 Other Features of the Text

The features that must be taken into consideration (already

discussed in Chapter 3), as they contribute to understanding and
interpreting texts, cannot be seen in the frequency lists of the corpora.
These are typographics, formulae, numbers, equations and tables and
diagrams and drawings these will be discussed below with reference to the
undergraduate textbooks studied here.

196
5.2.1 Typographics.

The use of typographics in the texts must be explored because as

Kress, Leite Gárcia and Leeuwen (1997:270) say “If humans make and
communicate meanings in many modes, then language no longer suffices
as the focus of attention for anyone interested in the social making and
remaking of meaning.” Therefore they point out that it is “highly
problematic to read only linguistically carried meaning”. Modern
textbooks, like modern newspapers, are colourful and have many different
sizes and styles of type and punctuation together with pictures and
drawings and are altogether different from the dense text typology of the
past. These changes bring with them the complication of interpretation
and discovering meaning referred to by Kress, Leite Gárcia and Leeuwen
above and which they (ibid.) believe to be “a normal condition of reading,
whether of a textbook page, of any page of a newspaper, or of a television
screen” in the modern world.
In Chang’s Chemistry technical terms are explicitly marked
orthographically the first time they are used. Orthographic marking
involves printing the word in boldface within the body of the text. A
marked term is accompanied by a definition or rather technically an
elaboration in the nearby text. Typically, once the term has been
elaborated, it is no longer highlighted. In addition, abstracts or ‘asides’ are
printed in the left-hand margin in blue type. For example Chang
(1991:170-1):

A barometer is an instrument that measures atmospheric pressure. A simple barometer

can be constructed ……

See Section 1.7 to review In SI units, pressure is measured in pascals (Pa), defined as
the definition of a newton
one newton per square meter:

Field Code Changed

force
pressure = 197
area
This equation gives the SI
1 atm = 1.01325 x 102 kPa
definition of 1 atm.

In Serway, Physics for Scientists and Engineers with Modern Physics,

the orthographic marking involves printing the word in boldface within the
body of the text and the elaboration is indicated by means of italics in the
nearby text.
For example in Serway (1992:202) there is the following:

“Effusion
Whereas diffusion is a process by which one gas gradually mixes with another, effusion is
the process by which a gas under pressure escapes from one compartment of a container to
another by passing through a small opening. Figure 5.20 shows the …..”

White (1998:267) claims that science and technology differ in the

way that they define; science locates definitional structures “in a
systematised set of taxonomic relationships”, while technology “typically
acts to identify the functionality of the items”. The chemistry text
definitions described above do both of these things, the barometer
definition is like technology, as defined by White, in that it describes its
function, and the newton and atmosphere definitions follow his
description of pure science definitions locating these terms in systems and
taxonomies.
It is important that these presentational elements are included in
the syllabus and that the different representations in pure science and
those of technology are expressed in the materials designed for use with
the students.

198
5.2.2 Titles, Subtitles, Summaries and Conclusions

As was mentioned earlier (section 3.1.9 Optical Character

Recognition), the optical scanner or optical character recognition machine
(OCR), which in this study was used to assemble the physics and
chemistry corpora, can only recognise what is visibly present on the page
and that it cannot undertake any kind of editing nor can it distinguish
structurally different components of the printed page, such as footnotes
and headings, even if these are visually distinct (Burnard 1992). Any
corpora searched will not provide these distinguishing features so that a
human/visual, graphical/layout analysis of the texts themselves as
published can often reveal some other important features of those texts
(see later 5.2.4 Diagrams and Drawings).
Both of these textbooks use a system of keywords to emphasise
those items considered important in the text and to allow cross-
referencing to these. Similarly, the text itself can be sub-divided into body
text, text describing figures, footnotes and asides. As mentioned earlier,
the Chemistry text employs the use of marginal notes in blue type and it is
only in the chemistry text that extensive use of footnotes is to be found,
reaching up to 20% of the total page text on occasions.
The division of text into shorter sections is considered important by
Stubbs (1996:84) who suggests that the author is showing by it a
particular attitude towards the expected reader. He notes that school
textbooks are traditionally divided up into small sections which may
suggest that the author expects a limited attention span from readers.
Division into smaller sections is also a technique of the Physics and
Chemistry textbook authors under study here to improve the readers’
interpretation and understanding of their essentially instructional texts
and for ease of finding relevant sections of the textbook. Chang (1991:xxii)
describes this as ‘readability’ and says that he tried to provide flexibility
Field Code Changed

199
for in-class assignments whilst making smooth transitions from topic to
topic.

Each Chapter in these textbooks is summarised and there are lists

of keywords on the topic of the chapter at the end of every chapter. One of
the most distinguishing features of these textbooks is the long section of
questions based on the theory presented and exemplified throughout the
chapter. It is the sections which exemplify, in different coloured boxes in
the body text and the individual sections with ‘real world’ applications of
the theories discussed, and the questions at the end of the chapter which
distinguish this genre and make it essentially instructional in nature.
These are the sections nevertheless that often display more culturally
loaded information with their references to particular American sports and
activities which are an attempt by the authors to present the topics
through activities that the (American) students can relate to their everyday
lives (see 5.3.3 The Physics and Chemistry Sub-Corpora).

5.2.3 Formulae, Numbers, Equations and Tables.

In science texts the use of formulae, numbers, equations and tables

are essential features, as they often represent the only clear
representation of the information contained in the text or an alternative
interpretation of the text. Certain conventions need to be understood,
such as the arrow or equals sign in order to interpret formulae and
equations correctly but, fortunately, these conventions enjoy international
standing and it can be assumed that the students had been taught their
use in school in Portugal so that shared background knowledge exists in
respect of these conventions. They do, however, involve the integration of
semantics, cohesion and frequently grammar in order to be read as if they
are part of the text (Lemke 1998). In this way the students’ understanding

200
must go beyond the simplest interpretation of these features. Tarone et al.
(1981:201) find in the astrophysics journals that they studied that:

One of the most striking characteristics of the sentences used in these papers is
the fact that lengthy equations are embedded within them, and must be
arranged in such a way as not to interfere with the reader’s processing of the
basic grammar of the sentence. Because of end-weight such equations are often
placed at the end of clauses, and the use of active or passive verb forms is often
conditioned by this requirement.

In Chang there are a number of equations that follow Tarone et al.’s

findings, being placed at the end of clauses such as:

The molar mass of H3PO4 is given by

3(1.008g) + 30.97 g + 4(16.00 g) = 97.99 g

However, there are other grammatical constructs used such as

conditionals and either... or clauses:

If we had used the empirical formula HO for the calculation, we would have written

1.008 g
%H = × 100% = 5.926%
17.01g
16.00 g
%O = × 100% = 94.06%
17.01g

In terms of factor-label method, we can write the unit factor as

Field Code Changed

201
 In Serway there is a similar situation with an entire paragraph
containing equations which must be read as part of the sentence
grammar:

The defining equation for acceleration,

dv
a=
dt
may also be written in terms of an integral (or antiderivative) as

v = ∫ a  dt + C1
where C1 is a constant of integration. For the special case where the acceleration a is a
constant, this reduces to
v = at + C1

Similarly, the chemical symbols for elements and compounds are

standardised through international convention and should not prove to be
a stumbling block, provided that they are understood in Portuguese
especially as the formulae follow the English order. Thus, it can be argued
that the number and range of formulae should add to understanding
rather than obscuring it, provided that there is shared background
scientific knowledge. Formula would also help to explain any common
term used in English for a particular chemical compound such as
common salt rather than its chemical name. However, some mathematical
working could prove confusing (see later 5.4 Mathematics).
Formulae can represent almost 20% of the text on a page in both
the Physics and Chemistry textbooks on average. Lemke (1998:89) finds
that it is “actually unusual to find high concentrations of both equations
and graphics in the same article or on the same page”. This is untrue of

202
the undergraduate physics and chemistry textbooks being studied here
which often integrate graphics and equations. The example from Serway
given above ends with a graph showing the velocity versus time curve for a
particle moving with a velocity that is proportional to the time which was
represented by the formulae contained in the paragraph quoted and is
followed by a further paragraph extrapolating further and containing
formulae as part of the sentence grammar. Lemke (ibid.) suggests that
experimental-empirical reports tend to have more graphics, whilst
theoretical analyses have more equations. This might explain why the
textbooks have both together in order to survey the experimental work
which has been carried out and also to represent the theoretical
contribution made to the field by that work.
Tarone et al. only considered active and passive voice in the
combination of clauses and equations but in these textbooks many other
structures were found such as: conditionals (see example above),
contrasts and comparisons (Similarly,..., are compared as follows, however,
not etc.) alternative either...or constructions (see example above),
exemplifications (that is, in other words, is given by, as follows etc.), logical
conclusions (we can now write ..., we can write this as ..., we obtain ..., we
would thus write) and the expression of laws as formulae (This is called the
associative law of addition: A + (B +C) = (A + B) + C, and is known as the
commutative law of addition: A + B = B + A, Charles’ law: V ∝ T...etc.). The
syllabus for the undergraduates coming into contact with these types of
texts must take account of the integration of grammar and formulae and
equations, exploring the different types of sentence structure that they are
usually found and the means by which these are expressed in words.
Recognition of the oral expression of formulae and equations is important
for understanding lectures given in English and also for note taking.

Field Code Changed

203
5.2.4 Diagrams and Drawings

The use of diagrams and drawings should enhance the

understanding of the surrounding text or even create new orders of
meaning but, in either case, it is essential that the referencing to these be
understood for this to take place. In general, visual material in the text is
a form of redundancy, as it reiterates or expands upon what is being
discussed. However, some problems may arise when the text and the
diagrams do not correspond exactly. An example of this occurred in some
materials produced for use with undergraduates in the Ano Comum in the
university. An exercise was produced using diagrams and text taken from
an encyclopaedia explaining the four-stroke petrol engine. The students
were asked to match the descriptions of the four strokes with the
diagrams which had been placed in a random order. The relevant part of
the text gave the following list:

1st stroke: induction stroke: while the inlet valve is open, the descending piston draws
fresh petrol-and-air mixture into the cylinder.
2nd stroke: compression stroke: while the valves are closed, the rising piston
compresses the mixture to a pressure of about 7-8 atm.; the mixture is then
ignited by the sparking plug.
3rd stroke: power stroke: while the valves are closed, the pressure of the gases of
combustion forces the piston downwards.
4th stroke: exhaust stroke: the exhaust valve is open and the rising piston discharges the
spent gases from the cylinder.
The diagrams were presented in the order shown below:

204
 The students could have adopted a number of strategies to find the
correct sequence. One of the obvious features was the fact that the
labelling had been retained in image three, which would therefore,
conventionally, make this the first diagram. The answer to this problem as
given by the encyclopaedia in the original was, stroke 4, stroke 3, stroke 1,
stroke 2. The discussion of this order invariably revolved around the fact
that there was inconsistency between the second image and the fourth
image. The second image does not conform to “forcing the piston
downwards” whereas the fourth image does. The students often failed to
take note of the ‘spark’ that image four displays but which is difficult to
detect in a black and white image like those shown above.
In other words the relationship between diagrams and the
supporting text is not as self-explanatory as may at first be thought. This
suggests that the syllabus must reflect this difficulty by presenting a
number of different types of relationships between visuals and the text. As
was mentioned earlier in 5.2.3 this could include equations and visuals
being read as part of a paragraph in the physics textbook with which the
students would need practice and would need to develop strategies to
overcome some of the difficulties encountered like the mismatch described
above. This is an area that is increasingly important with the change
towards much more use of visual representation in modern life both in
textbooks and in computers and specific discourse strategies which
encourage students to explore the relationships set up by visual
representation is increasingly important in education (Carter, Goddard,
Reah, Sanger and Bowring 1997).
Analysis of the Physics and Chemistry textbooks shows that
photographs, diagrams, graphs and drawings of apparatus are all found
together with formulae and examples in highlighted boxes. The relative
composition between pictures and text on each page is, on average, one
Field Code Changed

205
third visual to two thirds text, although these values can range quite
widely from between 70% for pictures and 30% text, and 90% text and
10% diagrams.5
Chang (1991:xxiii) says that their use of a “5-color design” will help
students “to visualise the appearance of compounds and various chemical
processes”. They also mention that in this edition they have added “many
full-color photographs and line drawings” and have introduced “A number
of marginal arts” to “enhance discussion and to accompany worked
examples”. Moreover, they have attempted to be consistent in their use of
colour to illustrate similar concepts “wherever appropriate”. In other
words, a principled approach has been taken by the textbook editors,
which the students need to be aware of, in order to benefit from the
insights these features should bring.

5.3 The Undergraduate Textbooks

The physics textbook recommended for students of the combined

first year (Ano Comum) of the science and technology courses of the
university of Aveiro is written by an American, Raymond Serway,
published in 1992 and entitled Physics for Scientists and Engineers with
Modern Physics, 3rd Edition. There is now a fourth Edition (1996)
available of this textbook which claims that it has been successfully used
by over 700 colleges and universities. The chemistry book for this year is
by Raymond Chang, published in 1991 and entitled Chemistry. Both of
these books appear, to a certain extent, to be trying to overcome the
pedagogic limitations outlined by Halliday and Martin in their (1993) book
Writing Science who claim that science textbooks pay too little attention to

5
It is interesting to note that, even in some translations of textbooks into Portuguese, the labelling of
diagrams remains the same.

206
solid pedagogic principles. Chang (1991:xxii) claims to use a “Problem-
Solving Pedagogy” where students are “asked to examine the
reasonableness of the answer” they give to problems in the end-of-chapter
exercises. Students are expected to explain the “why” of chemistry through
the review questions, the “how” of chemistry through the problems and to
identify the “concept, topic or technique to be applied” through the
miscellaneous problems.
The way that these particular textbooks have attempted to overcome
Halliday and Martin’s ascription of pedagogic limitations is by dealing with
topical issues or everyday situations. Each of the chapters in this physics
textbook contains an essay on a topic of more general interest, but with a
very specific physics focus, and each chapter is followed by a number of
questions or problems based on laws dealt with in the chapter, but which
attempt to put the scientific theories into more popular situations such as
those of travel, sports, nature and so on. Chang (1991:xxii) explains that
“to define complex terms in a clear manner, and to explain difficult
concepts carefully” use is made of analogy, for example downhill skiing
and dynamic chemical equilibrium are used to introduce a chemical
concept. Similarly, the chemistry textbook claims in the preface
(1991:xxii):

“Real-World Applications. One of the joys of learning chemistry is seeing

how chemical principles can be applied to everyday experience. The
Chemistry in Action sections show the relevance of chemistry to biological,
medical, technological, and engineering fields, as well as current news
topics.”
Glaser (1982:78) points out that “analogies from the learner’s
everyday experience are often used by way of comparison” in what she
describes as the didactic style of ESP textbooks and that this has
implications for the linguistic structure of the text. However, Laurillard
(1993:59-60) warns how unsuitable analogies are often used by students Field Code Changed

207
and that this process of imagining concrete analogies is not “a reliable way
of gaining access to the experience of academic knowledge”. She (1993:59)
also claims that “Physics is notorious for alluring concrete analogies that
lead you falsely” and even suggests that teachers themselves use
inappropriate analogies in their teaching. Whilst it is not being suggested
that these textbooks lead students astray with their extensive use of
analogies they may be falling into the trap that Laurillard warns about
because the students are not capable of thinking scientifically and may
therefore draw the wrong conclusions from the analogies given. It is in this
way that even teachers can fall into traps as Martins and Veiga (1999)
explore in their study on training primary school teachers to teach science
to pupils by exploring contexts from their daily lives. They found that the
teachers themselves often needed to learn how to think scientifically in
order to overcome misconceptions before they could help their classes to
do the same.
Some of the titles of the real world or “Chemistry in Action”
analogies in the chemistry textbook show the diversity of the subjects the
students will encounter, for example these range from “The Scientific
Method and the Extinction of the Dinosaurs”, “Salvaging the Recorder
Tape from the Challenger”, “Black and White Photography”, “Breath
Analyzer”, “Scuba Diving and the Gas Laws”, “Fuel Values of Foods and
Other Substances”, “How a Bombardier Beetle Defends Itself”,
“Determining the Age of the Shroud of Turin”, to “The Thermodynamics of
a Rubber Band” and many more. The scope that these provide for
misconceptions is therefore quite broad depending on the ideas the
students already have on these diverse topics.
In addition, this textbook details nine supplements for use with it
including video and computer programs, these are: Student Solutions
Manual, Microscale and Macroscale Experiments for General Chemistry,
Study Guide, Instructor’s Manual, Test Bank, R-H Test, Overhead

208
Transparencies, Chemistry at Work Videodisc, Micro Guide. These
supplements are not available to the undergraduates through the library
which contains multiple copies of the textbooks themselves in English.
The latest edition of Serway references a world wide web site at the
University of Texas which will provide answers to students questions. It is
likely that this site was intended for those hundreds of colleges and
universities in the United States which he claims successfully use the
textbook. The tendency to provide more and more support material
through computers (Serway also has accompanying computer simulations
and spreadsheets) cannot be ignored. The syllabus for these
undergraduates will have to encompass computer literacy in order to keep
in step with the developments that are taking place in educational
materials for science and technology students.

5.3.2 The Physics and Chemistry Algorithms and Functions

compared with Biber’s Academic Prose

Halliday (1991) makes a case for the validity of examining not only
lexical frequency in text but also grammatical frequency. He claims that
grammatical frequency is even more powerful than lexical frequency
because the system is closed and the number of choices is small, so that
significant probabilities can be calculated.
Biber standardised or “normalized” grammatical features found in
texts, that is, he standardised the raw data to reflect the frequency in a
thousand word extract by dividing the number of occurrences of a certain
grammatical feature by the total number of words in the text and then
multiplying by one thousand6. By observing this same level of scientific
rigour any text can be compared with Biber’s results in order to draw
conclusions about its position on the continuum of variation and thereby,

6
In Biber, Conrad and Reppen (1998:33) they refer to this process as a normed count. Field Code Changed

209
for the purposes of this research, to draw conclusions about the nature of
the text concerned and its consequent difficulty for students. The specific
definitions of each of the sixty-five algorithms used by Biber and in this
research can be found in Appendix A.
The values presented in the table below include normalised
frequency values and the chi-square test (χ2) to show which values are
significant. The degrees of freedom value (df) is one and, at the five percent
level, the critical value is 3.84. Yates correction was used on all values of
less than five.

Table 5.3 Normalised Frequencies from the Main Corpora compared to Biber’s Academic
Prose with Statistical Significance Values (chi-square χ2)
Linguistic Feature Physics Text Chemistry text Biber’s
Academic Prose
Past Tense 1.5 1.3 21.9
χ2 = -19.95 χ2 = -20.33
Perfect Aspect 0.9 1.0 4.9
χ2 = -4.13 χ2 = -3.95
Present Tense 44.2 69.2 63.7
χ2 = -5.97 χ2 = 0.47
Place Adverbials 3.7 1.3 2.4
χ2 = 0.26 χ2 = -1.06
Time Adverbials 1.8 1.4 2.8
χ2 = -0.8 χ2 = -1.29
First Person Pronouns 7.6 8.7 5.7
χ2 = 0.63 χ2 = 1.58
Second Person Pronouns 2.5 0.9 0.2
χ2 = 16.2 χ2 = 0.2
Third Person Personal Pronouns 4.5 2.6 11.5
χ2 = -4.89 χ2 = -7.68
Pronoun it 4.5 3.8 5.9
χ2 = -0.61 χ2 = -1.14
Demonstrative Pronouns 8.4 5.3 2.5
χ2 = 13.92 χ2 =3.14
Indefinite Pronouns 0.1 0 0.2
χ2 = -1.8 χ2 = -2.45
Pro-verb DO 0.1 0.1 0.7
χ2 =-1.73 χ2 = -1.73
Direct WH-questions 4.3 3.3 0
χ2 = 14.44 χ2 = 7.84
Nominalizations 51.8 49.8 35.8
χ2 = 7.15 χ2 =5.47
Gerunds 6.8 7.0 8.5
χ2 = -0.34 χ2 = -0.26
Total Other Nouns 155.7 159.1 188.1

210
 χ2 = -5.58 χ2 = -4.47
Agentless Passives 19.5 11.9 17.0
χ2 = 0.36 χ2 = 1.53
By - Passives 3.6 2.8 2.0
χ2 = 0.61 χ2 = 0.05
BE as Main Verb 20.4 19.5 23.8
χ2 = -0.48 χ2 = -0.78
Existential there 0.5 1.0 1.8
χ2 = -5.04 χ2 = -0.94
that Verb Complements 4.7 3.6 3.2
χ2 = 0.31 χ2 = -0.003
that Adjective Complements 0.1 0.1 0.4
χ2 = -1.6 χ2 = -1.6
WH - Clauses 0.3 0.1 0.3
χ2 = -0.83 χ2 = -1.63
Infinitives 0.1 9.2 12.8
χ2 = -13.61 χ2 = -1.01
Present Participial Clauses 2.3 1.2 1.3
χ2 = 0.19 χ2 = -0.28
Past Participial Clauses 0.1 0 0.4
χ2 = -1.6 χ2 = -0.85
Past Participial WHIZ Deletion 3.2 0.6 5.6
Relatives χ2 = -1.5 χ2 = -5.4
Present Participial WHIZ 2.8 1.3 2.5
Deletion Relatives χ2 = -0.02 χ2 = -1.16
that Relative Clauses on Subject 1.5 2.0 0.2
Position χ2 = 3.2 χ2 = 8.45
that Relative Clauses on Object 0.5 0.4 0.8
Position χ2 = -0.8 χ2 = -1.01
WH Relative Clauses on Subject 1.5 1.5 2.6
Position χ2 = -0.98 χ2 =-0.98
WH Relative Clauses on Object 2.1 1.1 2.0
Position χ2 = -0.08 χ2 = -0.98
Pied-piping Relative Clauses 0.8 0.5 1.3
χ2 = -0.77 χ2 = -1.3
Sentence Relatives 0 0 0
Causative Adverbial 0.7 1.2 0.3
Subordinators χ2 = 0.03 χ2 = 0.53
Concessive Adverbial 0.3 0.4 0.5
Subordinators χ2 = -0.98 χ2 = -0.72
Conditional Adverbial 4.2 2.2 2.1
Subordinators χ2 = 1.22 χ2 = -0.08
Other Adverbial Subordinators 2.2 1.5 1.8
χ2 = -0.01 χ2 = -0.35
Total Prepositional Phrases 127.5 125.4 139.5
χ2 = -1.03 χ2 = -1.43
Attributive Adjectives 29.8 36.6 76.9
χ2 = -28.84 χ2 = -21.12
Predicative Adjectives 6.1 3.1 5.0
χ2 = 0.24 χ2 = -1.15
Total Adverbs 6.4 8.5 51.8
χ2 = -39.79 χ2 = -36.19
Field Code Changed

211
 Type/Token Ratio 51.9 38.3 50.6
χ2 = 0.03 χ2 = -2.99
Word Length 5.8 6.0 4.8
χ2 = 0.05 χ2 = 0.1
Conjuncts 2.4 4.5 3.0
χ2 = -0.4 χ2 = 0.33
Downtoners 0.9 1.7 2.5
χ2 = -1.76 χ2 = -0.68
Hedges 0.1 0.1 0.2
χ2 = -1.8 χ2 = -1.8
Amplifiers 1.2 1.6 1.4
χ2 = 0.35 χ2 = 0.06
Emphatics 2.2 2.2 3.6
χ2 = -1.0 χ2 = -1.0
Discourse Particles 0.3 0.1 0
χ2 = -0.04 χ2 = -0.16
Demonstratives 7.2 5.8 11.4
χ2 = -1.55 χ2 = -2.75
Possibility Modals 5.5 5.1 5.6
χ2 = -0.001 χ2 = -0.04
Necessity Modals 1.6 1.5 2.2
χ2 = -0.22 χ2 = -0.65
Predictive Modals 3.8 2.8 3.7
χ2 = -0.04 χ2 = -0.53
Public Verbs 1.8 2.9 5.7
χ2 = -3.4 χ2 = -1.91
Private Verbs 11.5 7.2 12.5
χ2 = -0.08 χ2 = -2.25
Suasive Verbs 0.3 0.3 4.0
χ2 = -4.41 χ2 = -4.41
Seem/appear 0.1 0.5 1.0
χ2 = -2.96 χ2 = -1
Contractions 0.2 0 0.1
χ2 = -1.6 χ2 = -3.6
Subordinator -that Deletion 0.2 0.1 0.4
χ2 = -1.23 χ2 = -1.2
Stranded Prepositions 0 0 1.1
χ2 = -2.33 χ2 = -2.33
Split Infinitives 0.1 0.1 0
χ2 = -0.16 χ2 = -0.16
Split Auxiliaries 0.4 1.0 5.8
χ2 = -5.59 χ2 = -4.31
Phrasal Coordination 22.5 20.1 4.2
χ2 = 75.43 χ2 = 56.46

Independent Clause Coordination 0.7 0.4 1.9

χ2 = -1.52 χ2 = -2.11
Synthetic Negation 0.6 0.6 1.3
χ2 = -1.11 χ2 = -1.11
Analytic Negation 1.6 2.0 4.3
χ2 = -2.38 χ2 = -1.82

212
 The algorithms that are found to differ significantly from Biber’s
Academic prose for each of the corpora are the following:

Table 5.4 The Physics Main Corpus: Significantly Higher and Lower Results

Significantly Lower Results Significantly Higher Results

Past Tense Second Person Pronouns
Perfect Aspect Demonstrative Pronouns
Present Tense Direct WH-questions
Third Person Pronouns Nominalizations
Total Other Nouns Phrasal Coordination
Existential there
Infinitives
Attributive Adjectives
Total Adverbs
Suasive Verbs
Split Auxiliaries

Table 5.5 The Chemistry Main Corpus: Significantly Higher and Lower Results

Significantly Lower Results Significantly Higher Results

Past Tense Direct WH-questions
Perfect Aspect Nominalizations
Third Person Personal Pronouns that Relative Clauses on Subject Position
Total Other Nouns Phrasal Coordination
Past Participial WHIZ Deletion Relatives
Attributive Adjectives
Total Adverbs
Suasive Verbs
Split Auxiliaries

Taking the two main corpora together there are eleven algorithms
found to be significantly lower or higher in both the physics and the
chemistry corpora from the sixty-five examined by Biber, these are:
Significantly Lower in both Main Corpora: Past Tense, Perfect Aspect, Third
Person Personal Pronouns, Total Other Nouns, Attributive Adjectives,
Total Adverbs, Suasive Verbs, Split Auxiliaries.

Field Code Changed

213
Significantly Higher in both Main Corpora: Direct WH-questions,
Nominalizations, Phrasal Coordination.

Biber comments that past tense forms are usually taken as a

surface marker of narrative, and that there is a co-occurrence of the past
tense, third person pronouns and perfect aspect verbs in narrative,
reported styles, typical of fiction. As these features are found to be
significantly lower in the corpora this suggests that overall these texts
emphasise an extreme ‘academic’ text profile. This in turn suggests that
they would be even more difficult for students because dry, academic texts
are seen to be much more abstract and more difficult to understand.
Biber reports that third person personal pronouns mark relatively
inexact reference to persons outside the immediate interaction and in
previous studies (1986) has found that they co-occur frequently with past
tense and perfect aspect forms as a marker of narrative, reported, (as
opposed to immediate,) styles. These texts are then more ‘immediate’ in
style than would be expected. Biber (1988:137-8) describes the non-
narrative purposes as including “(1) the presentation of expository
information, …(2) the presentation of procedural information, …(3)
description of actions actually in progress.” These features of texts must
then be respected in the materials used with students in order to prepare
them for their studies. Equally well the combination of these findings
suggest that fiction would have no place in the discipline if only
appropriate texts are to be used with the students. However, this assertion
will be challenged with respect to the sub-corpora findings which follow.

5.3.3.The Physics and Chemistry Sub-Corpora

As mentioned above both the Physics and the Chemistry textbooks

contain essays in every chapter which are meant to illustrate or explain
the subject of the chapter in either a more stimulating or in a real-world
214
application. These essays were analysed using the same criteria given
above because it was felt that they could represent difficulties for the
foreign language student. The full texts are given in Appendix E.
Chapter One of Serway (1992) Physics for Scientists and Engineers
with Modern Physics contains an essay entitled Scaling - the Physics of
Lilliput written by Philip Morrison of the Massachusetts Institute of
Technology. The principal objective of the essay is to examine the
hypothesis that it is possible to have either very much larger or very much
smaller creatures who could look just like us. People are said to come in
‘all shapes and sizes’ but if this suggestion is examined more closely it can
be seen that this variety really only operates within strict upper and lower
limits and the extremes recorded in the Guinness Book of Records have
varied little over the last century (between four feet or one metre and seven
feet or two metres approximately). In other words it is not possible to have
the minute six inch creatures or the giants twelve times Gulliver’s size as
described by Swift in his Gulliver’s Travels. Morrison reaches the
conclusion that “Lilliputians must be a hungry lot, restless, active,
graceful, but easily waterlogged.” because of the necessity to produce
enough energy they would be constantly searching for and eating food but,
if it rained, their relative surface area would cause them to be weighed
down by the water covering their bodies, rather like a fly.
The essay itself is about two and a half thousand words long and
gives a ratio of the total number of words to the number of different words
or token/type ratio of 55.4. Biber (1988:104-5) in Variation across speech
and writing argues that the higher the type/token ratio the higher the
lexical variety and the more abstract the text concerned is. This figure
would therefore represent a difficult text for the student, given that
approximately two in every three words is different.
The Chemistry textbook presents similar but generally shorter
essays and there are more of them in each chapter than the usual one
Field Code Changed

215
long essay per chapter in the Physics textbook. The essay which has been
used for comparative study from the Chemistry textbook is entitled
Salvaging the Recorder Tape from the Challenger and it appears in the
Chemistry in Action section of the third chapter of the textbook and is
thus part of the overall corpus taken from this textbook. There are two
other Chemistry in Action essays in this chapter illustrating the chemical
reactions which have been described in the preceding chapter but these
are extremely short with lots of equations and lots of photographs
respectively. These essays were therefore rejected as being too limited in
scope for the purpose of comparison. This particular essay is concerned
with the crash of the space shuttle Challenger in 1986 and the
subsequent recovery of the tape of the flight and the chemistry used in
order to be able to listen to the seawater damaged recording that had been
made of the fateful flight. The essay is only about 450 words long and
presents a token/type ratio of 51.1, which is considerably higher than for
the chemistry corpus as a whole.
The results for each of the linguistic features as described in
Appendix A for these two essays are given below:

Table 5.6 Normalised Frequencies from the Sub-Corpora compared to Biber’s Academic
Prose with Statistical Significance Values (χ2)
Linguistic Feature Physics Essay Chemistry Biber’s
Essay Academic Prose
Past Tense 7.6 20.4 21.9
χ2 = -9.34 χ2 = -0.1
Perfect Aspect 4.9 0 4.9
χ2 = -0.05 χ2 = -5.95
Present Tense 35.0 24.9 63.7
χ2 = -12.93 χ2 = -23.63
Place Adverbials 1.7 0 2.4
χ2 = -0.6 χ2 = -3.5

Time Adverbials 2.8 4.5 2.8

χ2 = -0.09 χ2 = 0.51
First Person Pronouns 17.3 0 5.7
χ2 = 23.61 χ2 = -5.7
Second Person Pronouns 5.2 2.3 0.2
χ2 = 101.25 χ2 = 12.8

216
Third Person Personal Pronouns 16.3 6.8 11.5
χ2 = 2.0 χ2 = -1.92
Pronoun it 8.3 4.5 5.9
χ2 = -5.32 χ2 = -0.61
Demonstrative Pronouns 2.1 4.5 2.5
χ2 = -0.32 χ2 = 0.9
Indefinite Pronouns 0 0 0.2
χ2 = -2.45 χ2 = -2.45
Pro-verb DO 0.7 0 0.7
χ2 = -0.36 χ2 = 0.06
Direct WH-questions 0.7 0 0
χ2 = 0.04
Nominalizations 13.9 38.5 35.8
χ2 = -13.4 χ2 = 0.2
Gerunds 5.9 6.8 8.5
χ2 = -0.8 χ2 = -0.34
Total Other Nouns 164.8 226.2 188.1
χ2 = -2.89 χ2 = 7.72
Agentless Passives 10.4 22.6 17.0
χ2 = -2.56 χ2 =1.85
By - Passives 2.8 4.5 2.0
χ2 = 0.05 χ2 = 2.0
BE as Main Verb 23.2 18.1 23.8
χ2 = -0.02 χ2 = -1.37
Existential there 1.0 2.3 1.8
χ2 = -0.94 χ2 = 0
that Verb Complements 3.8 6.8 3.2
χ2 = 0.003 χ2 = 3.0
that Adjective Complements 1.0 0 0.4
χ2 = 0.03 χ2 = -2.03
WH - Clauses 3.8 0 0.3
χ2 = 30.0 χ2 = -2.67
Infinitives 7.6 20.4 12.8
χ2 = -2.11 χ2 = 4.51
Present Participial Clauses 1.0 2.3 1.3
χ2 = -0.49 χ2 = 0.192
Past Participial Clauses 0 0 0.4
χ2 = -2.03 χ2 = -2.03
Past Participial WHIZ Deletion 1.7 11.3 5.6
Relatives χ2 = -3.46 χ2 = 5.8
Present Participial WHIZ 1.4 2.3 2.5
Deletion Relatives χ2 = -1.02 χ2 = -0.2
that Relative Clauses on Subject 0.7 2.3 0.2
Position χ2 = 0 χ2 = 12.8
that Relative Clauses on Object 1.4 0 0.8
Position χ2 = 0.01 χ2 = -2.11
WH Relative Clauses on Subject 4.5 0 2.6
Position χ2 = 0.75 χ2 = -3.7
WH Relative Clauses on Object 2.1 9.1 2.0
Position χ2 = -0.08 χ2 = 21.78
Pied-piping Relative Clauses 0.3 0 1.3
χ2 = -1.73 χ2 = -2.49
Sentence Relatives 0.7 0 0 Field Code Changed

217
 χ2 = 0.04 χ2 = 0
Causative Adverbial 1.7 0 0.3
Subordinators χ2 = 2.7 χ2 = -2.13
Concessive Adverbial 0 0 0.5
Subordinators χ2 = -2.0 χ2 = -2.0
Conditional Adverbial 3.5 0 2.1
Subordinators χ2 = 0.39 χ2 = -3.22
Other Adverbial Subordinators 1.4 0 1.8
χ2 = -0.45 χ2 = -2.94
Total Prepositional Phrases 112.4 131.2 139.5
χ2 = -5.26 χ2 = -0.49
Attributive Adjectives 30.2 76.9 76.9
χ2 = -28.36 χ2 = 0
Predicative Adjectives 14.9 6.8 5.0
χ2 = 19.6 χ2 = 0.65
Total Adverbs 14.9 15.8 51.8
χ2 = -26.29 χ2 = -25.02
Type/Token Ratio 55.4 51.1 50.6
χ2 = 0.46 χ2 = 0.01
Word Length 5.6 6.0 4.8
χ2 = 0.02 χ2 = 0.1
Conjuncts 5.9 9.1 3.0
χ2 = 1.92 χ2 = 10.45
Downtoners 1.7 2.3 2.5
χ2 = -0.68 χ2 = -0.2
Hedges 0.3 0 0.2
χ2 = -0.8 χ2 = -2.45
Amplifiers 6.6 2.3 1.4
χ2 = 19.31 χ2 = 0.29
Emphatics 6.6 2.3 3.6
χ2 = 1.74 χ2 = -0.9
Discourse Particles 0.3 0 0
χ2 = 0.04
Demonstratives 5.9 4.5 11.4
χ2 = -2.65 χ2 = -4.18
Possibility Modals 10.4 4.5 5.6
χ2 = 4.11 χ2 = -0.46
Necessity Modals 4.2 0 2.2
χ2 = 1.02 χ2 = -3.31
Predictive Modals 8.3 2.3 3.7
χ2 = 4.54 χ2 = -0.98
Public Verbs 3.1 0 5.7
χ2 = -1.69 χ2 = -6.74
Private Verbs 7.3 2.3 12.5
χ2 = -2.16 χ2 = -9.16
Suasive Verbs 0.7 0 4.0
χ2 = -3.61 χ2 = -5.06
Seem/appear 0.3 0 1.0
χ2 = -1.44 χ2 = -2.25
Contractions 1.0 0 0.1
χ2 = 1.6 χ2 = -3.6
Subordinator -that Deletion 1.0 0 0.4
χ2 = 0.03 χ2 = -2.03

218
 Stranded Prepositions 0 0 1.1
χ2 = -2.33 χ2 = -2.33
Split Infinitives 0 2.3 0
χ2 = 3.24
Split Auxiliaries 1.4 6.8 5.8
χ2 = -4.14 χ2 = 0.17
Phrasal Coordination 17.0 0 4.2
χ2 = 36.02 χ2 = -5.26
Independent Clause 1.4 22.6 1.9
Coordination χ2 = -0.53 χ2 = 214.76
Synthetic Negation 2.8 4.5 1.3
χ2 = 0.77 χ2 = 5.61
Analytic Negation 8.0 0 4.3
χ2 = 2.38 χ2 = -5.36

The algorithms that are found to differ significantly from Biber’s

Academic prose for each of the sub-corpora are the following:

Table 5.7 The Physics Sub-Corpus: Significantly Higher and Lower Results

Significantly Lower Results Significantly Higher Results

Past Tense First Person Pronouns

Present Tense Second Person Pronouns
Pronoun it WH-clauses
Nominalizations Predicative Adjectives
Total prepositional Phrases Possibility Modals
Attributive Adjectives Predictive Modals
Total Adverbs Phrasal Coordination
Split Auxiliaries Amplifiers

As the object of the study of these sub-corpora is to see how far they
differ from the main corpora (and Biber’s findings), the results which show
a significant difference from both the main corpus and Biber’s findings are
those of interest. The algorithms which differ from Biber’s findings in both
the main physics corpus and the physics sub-corpus are the following:

Significantly Lower: Past Tense, Present Tense, Attributive Adjectives, Total

Adverbs, Split Auxiliaries.
Significantly Higher: Second Person Pronouns, Phrasal Coordination. Field Code Changed

219
 Nevertheless, the fact that the sub-corpus shows a greater
occurrence of certain features makes the essay significant in terms of
syllabus design, where certain features should be included in the syllabus
because of their presence in typical materials that the students will come
across in their studies. McCarthy and Carter (1994:112) say that
“whatever aspects of lexico-grammar we choose to look at, we cannot
really separate them from the concerns of creating discourse”. In other
words, these features make up the whole and cannot be taken out of
context without misrepresenting natural language use, in this case the
style of the science textbook in exemplifying real-world situations. If we
want students to cope with these kinds of texts, we must bring the
students into contact with the specifics of those texts.
The features that differ in the physics sub-corpus need not be
compared to those found for the chemistry sub-corpus as they are, in
themselves, a deviation from the norm of the main (physics) corpus and so
deserve study in their own right.
The results for the chemistry sub-corpus are given below in Table
5.8.

Table 5.8 The Chemistry Sub-Corpus: Significantly Higher and Lower Results

Significantly Lower Results Significantly Higher Results

Perfect Aspect Second Person Pronouns

Present Tense Total Other Nouns
First Person Pronouns Infinitives
Total Adverbs Past Participial WHIZ Deletion Relatives
Public Verbs that Relative Clauses on Subject Position
Private Verbs WH Relative Clauses on Object Position
Suasive Verbs Conjuncts
Phrasal Coordination Independent Clause Coordination
Analytic Negation Synthetic Negation
Demonstratives

220
 The chemistry sub-corpus differs from the chemistry main corpus in
one crucial way; it contains many more significantly high features than
the main corpus. That is to say, it contains many more examples of
features that are not so prevalent in either the main corpus or Biber’s
findings for Academic Prose.
The algorithms that the main and sub corpora share are the
following:

Significantly Lower: Perfect Aspect, Total Adverbs, Suasive Verbs

Significantly Higher: that Relative Clauses on Subject Position

Although the points in common are few, this finding is even more
significant as it shows the wide degree of difference between the text as a
whole and the essay studied here. This can only reinforce the conviction
that this kind of differentiation in the text will cause some students to
have greater difficulty than ever with the attempt by the author to
exemplify what is being studied through ‘real world’ situations. That is to
say, what is intended by the author to provide pedagogical enlightenment
can prove to be linguistic obfuscation for the non-native speaker learner.
Like Sinclair, Biber (1988:238-9) comments on the fact that the
longer the text the fewer new word types there are to be found, so that if
the entire length of text is considered, as in the figures calculated for the
main corpora above, such an accurate description of the difficulty of a
particular text especially for comparative purposes, is not demonstrated.
Furthermore, Biber himself (1988:48) suggests that “academic prose is
contextualized in that it crucially depends on shared (academic)
background knowledge for understanding”. However, as Bloor and Bloor
(1991:2) point out there is a “false expectation that educational structures
and systems do not differ internationally”, which means that we would do
well to anticipate differences in the students’ academic background from
Field Code Changed

221
the background assumed in the textbook being examined. Halliday and
Martin (1993:2) suggest that native speaker students of science are
“alienated” by the language of science. If all of these conclusions are true,
how much more alienated will the foreign language learner be by both a
combination of the (foreign) language of science and the lack of a shared
academic background to the subject. This is especially the case with the
essays under discussion, with their dense text and exophoric appeal to
native speaker background understanding in scientific, general and
literary knowledge.

5.3.4 The Physics Sub-Corpus: Gulliver’s Travels

A simple frequency listing of the text immediately highlights a

number of lexical items in the text which we would not have predicted in a
physics textbook despite the fact that technical discourse is supposed to
use repeatedly “a small set of technical vocabulary to refer to the exact
concepts and entities intended” (Grabe 1984).
In the physics text there are words relating to animals (not counting
man, family, human and so on) such as: animal (1), animals (1), bee (1),
bison (3), cattle (1), deer (2), dinosaurs (1), dog (1) and dogs (1), elephant (2),
fish (1), fly (2), frogs (1), gazelle (5), horse (1), insects (1), lamb (2), mammals
(1), mouse (2), whale (3), whale’s (1) and whalelike (1) and others relating to
plants: agriculture (1),grass (1), trees (1). Other ‘biological’ references
include; body (11), bone (8), bones (7), breathing (1), digest (1), fingers (1),
flesh (1), forearm (1), leg (5), limb (1), limbs (1), muscles (1), ribs (1), senses
(1), skeleton (1), skin (3), tendons (1), and warm-blooded (2) not to mention
the much more difficult to classify giant (1) and giants (5).7 Readers could
be forgiven if at this point they became confused as to whether this was
indeed a physics text or whether they had not come across the discourse

7
The figures given in brackets refer to the number of times that each word appears in the text, that is, their
frequency.

222
of biology by mistake. The questions which immediately follow this essay
also continue in the same vein and hummingbird, elephant and guinea pig
are used in the follow-up work.
The fact that certain items are used many times over in a text leads
to a lower type/token ratio, despite the fact that they may be difficult in
themselves, especially for non-native science students8. This is further
complicated when these human biological structures are compared to the
structures of buildings and braces, columns and cables are encountered
being compared with muscles and tendons, with an earlier analogy being
drawn between the strength of a wire or a rope. The collocations for braces
are as follows:
the skeleton - supported by various braces and cables which are muscles and tendons
the strength of his columns and braces is proportional to their cross-sections

The specific vocabulary and specific grammar of texts are now seen
to be inseparable. Halliday and Martin (1991:4) point out that

“technical terms are an essential part of scientific language, it would be

impossible to create a discourse of organized knowledge without them. But they
are not the whole story. The distinctive quality of scientific language lies in the
lexicogrammar (the ‘wording’) as a whole, and any response it engenders in the
reader is a response to the total patterns of the discourse.”

Halliday ( in Ajmer & Altenberg eds. 1991:32-3) goes on to attack a

school of thought that suggests that grammatical frequency has no
validity. He argues that:

8
It is debatable if many native speakers would be able to draw an adequate distinction between gazelle and
deer. The Oxford Advanced Learner’s Dictionary gives the following definitions “gazelle small, graceful
antelope, deer any of several types of graceful, quick-running, ruminant animal, the male of which has
antlers”
Field Code Changed

223
 “it does not make sense to condone relative frequency in lexis but deny its
validity in grammar (...) the concept of the relative frequency of positive:
negative, or of active: passive is no more suspect than the concept of the
relative frequency of a set of lexical items. It is, on the other hand, considerably
more powerful, because the relative frequencies of the terms in a grammatical
system, where the system is closed and the number of choices is very small
(typically just two or three), can be interpreted directly as probabilities having a
significance for the language as a whole.”

He goes on to say that these “grammatical choices may mean different

things in different registers”. This is obviously the case here where not
only is there a very specific (and unusual for a physics text) set of lexis but
this is coupled with significant grammatical variation from other academic
prose. The syllabus must therefore consider these aspects and include
suitable practice in both.

5.3.5 Comparison with other Genres in Biber’s Variation Studies

Biber (1988) identified a number of grammatical features in texts

and tried to demonstrate that the variation between different text types is
on a continuum rather than consists in any absolute distinction.
Nevertheless, he also demonstrates that it is both the absence as well as
the presence of certain grammatical features that distinguish different
genres. This leads into areas that Biber has discussed in relation to what
he called ‘Academic prose’ (a very diverse grouping including texts from
the humanities to texts from medicine), specifically, the use of pronouns.
Biber (1988:193) claims that academic prose, particularly those texts
dealing with technology/engineering and the natural sciences, does not
generally contain third person pronouns as this genre shows “non-
narrative concerns”. (He finds humanities prose to be something of an

224
aberration in this respect “showing a topical concern for concrete events
and participants”). Lemke (1990:440) disagrees and claims that

“Technical discourse is also dominated by third person forms. No “I” speaks to

a “you”, no space for dialogue, disagreement, or differing points of view is
opened in this way either. Even the solidary (inclusive) “we” is absent, and
only the authoritative authorial (exclusive) “we” of multiple authorship is
allowed. The world of technical discourse is a closed world which admits no
criteria of validity outside its own.”

There are eight examples containing he in the Gulliver’s Travels text

which would suggest that it is untypical of its genre according to Biber’s
results as mentioned above. Lemke is probably referring to the use of “it”
rather than “he” or “she” when he says ‘third person forms’, as later he
goes on to say (1990:440) that technical discourse is ‘independent of the
particular human agent who has happened upon “the facts”.’ Biber,
however, ranks the ‘pronoun IT’ as a separate feature from ‘third person
pronouns’ in his text analysis.
On examining other pronouns in this short essay, one finds that
there are two examples of I (both of which appear in a quotation) 30
examples of we, four examples of us, eleven examples of our and two
examples of ours. Biber explains that first person pronouns are usually
treated as demonstrating personal involvement in a text and are often
associated with cognitive verbs and ego-involvement. Moreover, they have
been used as a comparison between spoken and written texts. McCarthy
and Carter (1994:14) have developed a scale based on Smith’s (1986) work
which allows examination of mode variation between reader-listener and
writer-speaker where 2nd person pronouns indicate the presence of the
reader and listener and first person pronouns indicate the presence of the
writer and speaker. Passive constructions and third person references are
Field Code Changed

225
seen to indicate the ‘absence of reciprocity’ of senders and receivers. As
the examples of I are both from quotations, it would be wrong to suggest
that this shows Philip Morrison’s ego involvement in the text, but the
unusual presence of 30 occurrences of we, eleven examples of our and
four examples of us in a scientific text should alert to what this implies in
terms of involvement. This text shows features that would more usually be
associated with other genres than the one being studied and suggests a
greater degree of informality in the text.
There are seven examples of you, which as Biber suggests requires a
specific addressee and indicates a high degree of involvement with that
addressee. This is perhaps less surprising in a text that is meant to be
instructional. Biber has used second person pronouns as a marker of
register differences and so once again there is evidence of involvement in
this text. There is one example of one as a pronoun, If one wishes.
McCarthy and Carter (1994:15) find the pronoun ‘one’ to be a marker of
‘absence of intimacy’ in either the spoken or the written mode.
There are sixteen examples of it. In a previous study Biber (1986)
suggested that a high frequency of this pronoun marked a relatively
inexplicit lexical content due to strict time constraints and showed a non-
informational focus, in other words it in high frequencies is used in text-
types like telephone conversations, face-to-face conversations, personal
letters, spontaneous speeches and interviews. Others (Kroch and Hindle
1982) have also associated greater use of this pronoun with spoken
situations which is clearly not the case here. This text breaks away from
the general situation in the textbook being studied which suggests that
there must be provision made in the syllabus for a sufficiently wide range
of text-types so that these features can be studied in an appropriate
context, which would nevertheless not be face-to-face conversation.
However, McCarthy and Carter (1994) recommend finding texts that
combine the discourse features that students need to study even if these

226
are in another text-type which can be used as an appropriate vehicle for
studying those particular discourse features. Halliday and Martin (1993)
say that there is a lot of discussion in the science classroom to clarify the
language of the scientific textbook being studied even though the students
usually only write short sentences and definitions. The authors of the
textbooks being analysed here are anticipating that their books will be
used mostly in classes with teachers. Non-native speaker undergraduates
however are expected to have to read these textbooks alone. It may
therefore be entirely appropriate for texts from these textbooks to be used
in language classrooms where discussion of the texts can take place with
a language teacher rather than a science teacher. This would go some way
towards reproducing the expected mode of use of the textbooks and
perhaps thereby help the students to examine both the language and the
scientific discourse they need to cope with.
There are 21 examples of his and 9 examples of their. Biber reports
that third person personal pronouns mark relatively inexact reference to
persons outside the immediate interaction and, in previous studies (1986),
has found that they co-occur frequently with past tense and perfect aspect
forms as a marker of narrative, reported (as opposed to immediate), styles.
There are no examples at all of either she or her. The gender deficiency
found in this work confirms the findings of linguists like Halliday, Martin
and Beaugrande who claim that the language of science is the domain of
white, middle-class, adult males. However, the use of one and you in the
same essay points to a certain confusion of usage of pronouns and the use
of one as a pronoun is not included in Biber’s analysis at all. Once again
this essay would appear to be atypical of its genre according to Biber’s
findings. Serway suggests that he uses informal language in order to make
his work clear and penetrable for students but this confusion does not
support his proposition.

Field Code Changed

227
 The relevant normalizations for pronouns in this essay are as
follows:
First Person Pronouns 17.3
Second Person Pronouns 5.2
Third Person Pronouns 16.3
Pronoun it 8.3
Biber’s averages for these features were 5.7, 0.2, 11.5 and 5.9 respectively.
Examination of Biber’s findings reveals that for first person
pronouns the Gulliver’s Travels text is closer to the averages Biber found
for Religion (16.6), Biographies (22.1), and Science Fiction (22.2). The use of
biographical data on the scientists whose theories are discussed is
common in scientific textbooks and given the subject of Gulliver’s Travels
fiction is also included to some extent in this essay.
For second person pronouns Prepared Speeches (5.2) and Hobbies
(4.2). For third person pronouns Hobbies (14.1).
For pronoun IT Humor (8.2) and Prepared Speeches (8.9) and Press
Reviews (7.9).
If these texts provide an accurate picture of the use of these
features, it would be possible to widen the scope of the materials used
with students by including some of these as alternatives and making the
work more varied and interesting whilst sacrificing none of the relevancy.
Biber, Conrad and Reppen (1998:61) find that “academic prose uses
nominalizations to treat actions and processes as abstract objects
separated from human participants.” and that academic prose “more often
refers to a process with a stative nominalization, where fiction and the
spoken corpus describe a specific person’s action with a verb or adjective.”
In other words (1998:75) academic prose shows “a preference for static
rather than dynamic packaging of information.” They find that six different
nominalised words are very common in academic prose, that is with
frequencies of over 500 per million words. These are movement, activity,

228
information, development, relation and equation. In contrast, no
nominalisations were found to occur in fiction or speech this frequently.
All of the nominalisations mentioned above are found in both the physics
and the chemistry corpora studied here.
Abstraction is seen by Halliday and Martin (1993) to be one of the
reasons that science writing is so different from other writing that
students come into contact with and it is this abstraction factor that leads
to the difficulty experienced with science texts9. However, Biber’s
definition of nominalisations is somewhat different from Halliday and
Martin’s. Biber includes all words ending in -tion#, -ment#, -ness#, or-ity#
(plus plural forms) only, whereas Halliday and Martin allow anything which
can function as an element in another clause. Halliday and Martin
(1993:15) say

“Isolated instances of this (nominalization) would by themselves have little

significance; but when it happens on a massive scale the effect is to reconstrue
the nature of experience as a whole. Where the everyday ‘mother tongue’ of
commonsense knowledge construes reality as a balanced tension between
things and processes, the elaborated register of scientific knowledge
reconstrues it as an edifice of things. It holds reality still, to be kept under
observation and experimented with; and in so doing, interprets it not as
changing with time (as the grammar of clauses interprets it ) but as persisting -
or rather, persistence - through time, which is the mode of being of a noun.”

This is confirmed by Sinclair (1997:36) in his work on corpora.

He has found that if a word exists as both a noun and a verb the
more concrete meaning will be associated with the noun and the
more abstract with the verb. He gives the example of combat which

9
Halliday and Martin argue that students actually enjoy the technical terminology of science texts and do not
have difficulty with it as long as it is presented systematically.
Field Code Changed

229
as a noun means “actual physical fighting” and as a verb “means
something like ‘struggle against’”.
Other features identified by Biber as representative of
academic prose are passives and the use of the past tense and, as
mentioned earlier, it is the co-occurrence of some of these features,
(for example third person pronouns plus past tense plus perfect
aspect forms), that is important in positioning the text on the
continuum of the genre. In order to compare these factors with those
Biber obtained, it is necessary to normalise the text to a standard
1,000 words as Biber did in his research. The figures obtained from
this normalisation process are as follows:

Past Tense = 7.6 Biber found a mean of 21.9 for this feature.
This feature is only a third as frequent as in the Biber findings
putting it more on a par with Professional Letters (10.1) in Biber’s study.
Biber comments on the fact that his category of academic prose contains
wide variations and he suggests that Humanities prose shows a high score
because of its ‘topical concern for concrete events and participants’ while
engineering/technology prose reflects ‘concern with abstract concepts and
findings rather than events in the past’ and therefore has a low score. This
is borne out by the results in the main physics corpus (1.5) which is even
lower and not matched by any of Biber’s categories. It is interesting to
note that the Chemistry sub-corpus is very close to Biber’s finding (20.4 to
21.9 respectively) but that overall the main corpus is even lower (1.3) than
the physics main corpus.
Agentless Passives = 10.4 Biber found a mean of 17 for these
Passive Voice = 2.8 Biber found a mean of 2 for this feature
Passives are taken as characteristic of writing and when the agent is
dropped there is a static, more abstract presentation of information. In the
case of agentless passives this text is more on a par with Press Reportage

230
and Popular Lore in Biber’s study (11 and 10.6 respectively). Svartvik
(1966) calculated the number of passive clauses per 1,000 words of
running test for his 320,000 word corpus of eight text types. His results
showed an average of 11.3 and a range of 3.0 in advertising to 23.0 in
science. A comparable average in the physics text under inspection would
be 13.2, once again considerably lower than that referred to by other
investigators. It could be argued that it is the attempt by the authors to
reach (involve) the readers (students) that causes this finding.
Perfect Aspect Verbs = 4.9 Biber found a mean of 4.9 for this
Biber notes that these verbs have been associated with
narrative/descriptive texts and with certain types of academic writing. It is
interesting that this text is exactly the same as Biber’s finding for
academic prose, whereas the main physics corpus is only 0.9.
Nevertheless, this is the lowest mean score for this feature found in any of
the text types examined by Biber which makes academic prose in a
category of its own as regards the use of the perfect. For syllabus purposes
this is particularly significant and must be explored.
Nominalizations = 13.9 Biber found a mean of 35.8 for these.
This score is almost matched by that for Hobbies in Biber’s study (13.1),
closely followed by Science Fiction (14.0) and then Humor (12.1) and
General Fiction (10). Perhaps this finding is not so unexpected, given that
the discussion contained in the text examined here deals with the science
in Gulliver’s Travels.
Nouns = 164.8 Biber found a mean of 188.1 for this feature.
In Biber’s study Adventure Fiction most closely matches this mean score
(165.6) followed by Mystery Fiction (165.7) and General Fiction (160.7). This
may reflect the nature of the subject matter once again or the attempt to
make this physics text more amenable to its audience.
Prepositions = 114.4 Biber found a mean of 139.5 for these

Field Code Changed

231
The prepositions examined by Biber are taken from Quirk et al. (1985:665-
7). Biber finds that prepositions tend to co-occur frequently with
nominalizations and passives in academic prose and other informational
types of written discourse. The closest mean score for this feature in
Biber’s study is for Hobbies (114.6) and Popular Lore (114.8).
These results would place this text in rather different company than
that given in Biber’s results for academic prose, however, these are mean
scores and considerable variation has to some extent been integrated into
Biber’s study by including all kinds of academic prose and not only the
academic prose of science and technology which is of prime interest for
the students who have to study such texts in the University of Aveiro.
Nevertheless, the precise description of the study undertaken by Biber has
allowed a number of features in this physics text to be compared with his
findings and so allows a scientific comparison and interpretation to be
made, which in turn can be the basis for a reasoned approach to the
relative difficulty of such study material for our students and,
consequently, a clearer definition of the approach that needs to be
adopted in teaching such students to cope with their textbooks. In this
case, some of the features of abstraction are not present to any significant
degree, as defined above, but that the attempt to be more accessible will,
in fact, lead to even greater difficulty for the foreign language student of
physics at university level.

5.3.6 The Chemistry Sub-Corpus: Salvaging the Tapes from the

Challenger

The relevant normalizations for the same features discussed above

for the physics sub-corpus, together with the nearest text-type mean
found by Biber are as follows:
First Person Pronouns 0 Academic Prose (5.7)

232
Second Person Pronouns 2.3 Broadcasts (2.7) and Religion (2.9)
Third Person Pronouns 6.8 Professional Letters (8.7)
Pronoun it 4.5 Press Reportage (5.8) and Academic Prose (5.9)
Past Tense 20.4 Academic Prose (21.9) and Broadcasts (18.5)
Agentless Passives 22.6 Official Documents (18.6) and Academic
Prose (17)
Passive Voice 4.5 Official Documents (2.1) and Academic Prose (2.0)
Perfect Aspect Verbs 0 Academic Prose (4.9)
Nominalizations 38.5 Official Documents (39.8) and Academic Prose
(35.8)
Nouns 226.2 Broadcasts (229.8) and Press Reportage (220.5)
Prepositions 131.2 Biographies (122.6) and Broadcasts (118.0)

As with the Physics textbook, the Chemistry textbook includes

essays which purport to bring the science which is being taught in the
chapter into the realm of the everyday world. As with the physics textbook,
these essays would appear to be more complicated for the foreign
language student of science and technology, although the findings in this
case conform much more often with Biber’s findings for Academic Prose.
In some cases, such as Pronoun it and Perfect Aspect Verbs the results
found for this essay were much lower than in Biber’s study but
Prepositions were much higher, suggesting that this essay is still outside
the usual range found in Biber’s work.
McCarthy and Carter (1994:91) argue that “pronouns … stand in a
direct relationship with noun phrases and the demonstratives at the
discourse level” and that noun phrases and demonstratives ‘topicalize’
entities whereas pronouns “simply continue topics already raised to the
status of current focus”. The combination of higher and lower features
found above therefore help to define the identity of this sub-corpus of
chemistry, distinguishing it from the main chemistry corpus and
Field Code Changed

233
highlighting the even greater difference found in the physics corpora
studied here.
One of the features that has to be taken into consideration in this
sub-corpus is the cultural and historical aspects of the case of the crash
of the Challenger space shuttle. Americans could be expected to
‘remember’ this event as it was a tragedy for a nation who have
traditionally found failure difficult to accept. Foreign students could not be
expected to share such a collective consciousness on this topic and indeed
Chang’s essay does not appear in a Portuguese translation.

5.4 Mathematics

Whilst there is no separate mathematical corpus, as this was

deemed to represent little in terms of language that could be analysed to
provide significant data for materials writing, mathematics does represent
extra-linguistic features which must be taken into account to understand
the learning difficulties posed by mathematical formulae and numbers in
texts. Lemke (1998:104) says mathematics is “more powerful than
visualisation, even though it is less intuitive, because it can represent
patterns that cannot be visualised, and allow them to be compared,
manipulated, combined, etc.”
Both the Physics and the Chemistry corpora produced
approximately four pages of four columns of figures before the word lists
presented in Appendices C and D. The frequency of numbers in the
corpora suggest that they are of some significance in the textbooks under
analysis. Particularly in physics Serway recommends that students
coming to his book should have already have studied calculus and if they
have not then at the very least they should be studying it concurrently.
Therefore some observations on numbers and their significance for
the students’ understanding of text must be made. As was mentioned
before, the use of diagrams, figures and pictures usually reinforce the

234
content of the text they are associated with and the same is usually true of
numbers and formulae included in the texts; they reiterate the
commentary of the text, exemplify or complement the meaning in some
way. However, English and Portuguese do not follow the same resolution
of mathematical problems. A simple example will illustrate this difference.
If one number is divided by another the ‘working’ of the calculation will be
different even though the result should turn out the same. Take, for
example, 1526 divided by 32. In English this would appear as follows:

47.6875
32) 1526
128
246 4/128
224 7/224
220 6/192
192 8/256
280 5/160
256
240
224
160
160
000

As can be seen from the above, the answer to this problem, the
quotient, is given above the line at the very top of the calculation, 47.687,
the divisor is on the left, 32, and the number to be divided inside the
frame to the right, 1526. Each subtraction is shown below the number to
be divided in a series of steps. The indication that the result is a decimal
is given by the punctuation ‘full stop’ between the whole numbers and the
decimals and the necessary ‘working’ is given to one side of the calculation
itself (in this case on the right although there is no hard and fast rule
about this positioning).

Field Code Changed

235
 In Portuguese this calculation would look something like the
following:

1526.000) 32
246 47,6875
220
280
240
160
00

The number to be divided is on the left with a number of zeros

added to allow for the decimal places to be represented, 1526.000, and the
divisor, 32, on the right, the quotient is given under the line on the
right,47,6875, and only the results of the various steps of subtraction are
given on the left in the ‘working’ of the problem. In addition, the
punctuation used to distinguish whole numbers from decimals is a
comma, but a point is used for thousands.
Most Portuguese students when faced with the English style of
presenting the ‘working’ of such mathematical calculations find it difficult
to work out what is going on with each of the steps, although they are of
course essentially the same, the wealth of information provided seems to
confuse rather than enlighten.
As will be discussed later in relation to the students and can be
observed from both television and film subtitles, numbers are often not
readily translated from English into Portuguese. This observation holds
true even though the numbers are not being used in expressions such as
six of one and half a dozen of the other, which would be understandably
more difficult to translate as this would obviously require the translator to
interpret the phrase to suit the action in the film or television programme.
Added to these difficulties are those of the metric and imperial systems of
measurement which will be especially marked in these American texts

236
because America still adheres to the Imperial system. Despite the fact that
Britain is now almost completely metric, other differences still remain
between British and American measurements an American “ton” is lighter
than a British “ton” and is, of course, different again to its metric
equivalent “tonne” and a British gallon is more than an American gallon.
International scientific convention, and European Union regulations,
would require metric measurements to be given for everything but, as was
described above for the corpora, the fact that the authors have attempted
to bring their observations to bear on the everyday world and common
American pursuits invites the use of imperial measurements which make
up part of that world. The words “foot” and “feet” and “inches”, and “miles”
and “acres” and “pounds” and “tons” are indeed found in the corpora and
will probably cause difficulty as with the confusing billion and ton which
are also there. Serway says that he uses metric measurements in all but
the engineering sections which he nevertheless keeps to a minimum.

5.4.1 Mathematics in the Gulliver’s Travels Text

As mentioned above, one area of potential confusion for the foreign

student of science is the use of the imperial system of measurements. In
this essay these are evidenced by the use of “inch”, “inches” and “foot” as
in:

Lilliputians were a little under 6 inches high, on the average, and

all built on the scale of one inch to the foot

The idea of the ‘magnitude of things’ in the imperial system will

continue to be of major importance while America is a major trading
nation and world power.
Other mathematical expressions found in the text which could
conceivably cause confusion is the use of “square” as in:

Field Code Changed

237
In other words, the breaking strength of a wire or rope is proportional to its area of cross-
section, or to the square of its diameter
Because the strength of his columns and braces is proportional to their cross-sectional area
and thus to the square of their linear dimension

These items are often much more accessible to foreign language

students if they are expressed as mathematical formulae when it becomes
obvious that these uses of ‘square’ are referring to a number multiplied by
itself and not to the shape of an object.
Biber suggests that although “all academic sub-genres are
characterized by the features of highly informational production” (frequent
nouns, prepositions, attributive adjectives, long words, and high lexical
variety), mathematics texts have an even higher score because their
subject matter is technical and often non-linguistic, using mathematical
expressions instead. With respect to visualising meaning, the Gulliver’s
Travels text does contain some illustrations of relative bone sizes, which
should help the student to envisage the relative proportions being
described in the text, but the level of abstraction is nevertheless still
extremely high as illustrated by the examples given for square above.

238
Chapter 6 Discussion of the Results
 Chapter 6

6.1 Discussion of the Results

“There is a widespread consensus that language is never neutral and texts are
never innocent. Things can always be formulated differently, any linguistic
expression of the facts chooses some aspects of reality and downplays others,
and all choices are political (Martin, 1985). Representations are always from
a point of view, and express group interests. Such points of view are not
usually explicit, are often denied and may not be directly observable, because
they are often a matter not of individual words, but of patterns of distribution
and frequency.” Stubbs (1996:235)

Just as the results from the frequency studies and corpora

analyses will only be as relevant to students as the material they are
based on is relevant to them, the results of the tests on the students’
level of English may well vary with time. The readability levels of scientific
texts have been found by Hayes (1992) to have been increasing this
century. A study carried out at the University of New York by Dr Linda
Hirsch (reported in Rosenthal 1996:116-7) also found that the reading
skills necessary to read recent introductory science texts “required
reading skills at the college level or beyond, levels difficult to attain for
many ESL or post-ESL students” in American universities. One of the
introductory science texts that Dr Hirsch included in her comparison was
the physics textbook written by Serway studied here. She gave this
textbook a reading level of 14 which corresponds to grade 14 or the
second year of an American college-level course. Despite the fact that Dr
Hirsch considered the physics textbook to be somewhat easier than that

Field Code Changed

241
of chemistry, this level of readability will be extremely difficult for the
students considered by this thesis.
The fact that the reforms being carried out in schools will have an
effect on the scientific knowledge the students bring with them to their
undergraduate studies must also be taken into consideration. Added to
this is the manner in which those subjects were taught. Textbooks
produced for students from other cultures build on the understanding of
the educational background that the students are perceived to have. If
science subjects in schools were taught in isolation or as integrated
subjects, such as earth science rather than geography and chemistry and
so on, this will have an effect on the type of textbook written. Older books
and textbooks from different educational cultures may therefore be at
odds with the student profiles which will cause a subsequent increase in
difficulty for the undergraduate in a different cultural setting.
Many of the changes that are taking place in education have
underlying principles that are based on Europe-wide perceptions of what
skills and knowledge the educated person of the future will possess. As
students come through to the university from these reforms in the
schools both the students’ backgrounds and expectations will have
changed and the university has to be prepared to meet these different
expectations and aims. Moreover, the learning styles and previous
educational experiences of the students coming in to the university must
be taken into account. If the students’ had not studied science in a
collaborative way with discussion of the issues raised, this will provide a
mismatch with the style of the American textbooks studied which build
on this educational background in students rather than one in which
rote learning and memorisation are the norm with students not expected
to ask questions in class. Flexibility and the ability to reflect on and
study these various parameters must therefore remain one of the most
important aspects of university education.

242
 The findings from the frequency analyses and corpora studies must
therefore be examined in order to suggest what implications they bring to
the teaching of undergraduates at university.

6.2 Coursebooks and Multimedia Encyclopaedia Frequency

and Range Results.

The results of the comparison between the multimedia

encyclopaedia and the Physics and Chemistry coursebooks show some
surprising variations, particularly in the area of common words. Some of
the words excluded by the Grolier as being too common are not to be
found at all in the Physics and Chemistry corpora. Similarly, certain word
forms are to be found only in one or other of the Physics and Chemistry
corpora. These findings are consistent with the view that distinctions
between texts are caused by the relative frequency of use of certain
features rather than absolute and unique differences1. It must also be
remembered that only part of each of the textbooks was included in the
corpora so that there is always the possibility that these findings would
have a slightly different emphasis if the whole of the textbooks had been
included. For example, the frequency of the noun ‘bibliography’ would
have been greater if the bibliographies from the textbooks had been
included in the corpus. The results therefore show tendencies and not
absolute values.
These frequency lists can be used to show which items are
appropriate to include in materials for students and which require
particular emphasis. For example, the differences in the usage of the
Latin and Greek singular and plural forms found would suggest that

1
Biber et al. (1998:136) gives the example of ‘balls’ and ‘strikes’ being used as countable nouns only in
broadcasts of baseball games, as the exceptional, rare situation where these features are found only in that
one register rather than being shared with other registers to a greater or lesser degree.
Field Code Changed
243
teaching materials should reflect this difference rather than being
prescriptive and suggesting that only one plural is ‘correct’ usage when
the corpora suggest that actual usage is other than this2. The advantage
of having access through the corpora to the context of these forms and
their range across texts also provides information on the most useful
items to be used in each particular situation. The differences found
between Physics and Chemistry on words that would have been predicted
to be essential for science provide clear guidelines for the context that
these should be presented and studied in as was described in 5.1.2.
Stubbs (1996:40) reaches a number of conclusions about work on
lexico-grammar that are relevant here. He makes the following points:

1. Any grammatical structure restricts the lexis that occurs in it, and
conversely, any lexical item can be specified in terms of the structures in
which it occurs.
2. Such restrictions are typically not absolute, but clear tendencies:
grammar is inherently probabilistic.
3. Meaning is not constant across the inflected forms of a lemma.
4. Every sense or meaning of a word has its own grammar and each
meaning is associated with a distinct formal patterning. Form and meaning
are inseparable.
5. Words are systematically co-selected: the normal use of language is to
select more than one word at a time.
6. Since paradigmatic choices are not made independently of position in
syntagmatic chain, the relation between paradigmatic and syntagmatic has to
be rethought.
7. Traditional word-classes and syntactic units also have to be rethought.
Native speakers have only limited intuitions about such statistical
tendencies. Grammars based on intuitive data will imply more freedom of

2
Peters (1998:6-12) reports on the Langscape Project of Macquarie University, Sydney, Australia on the
Langscape 1 questionnaire on spelling by age group and nationality. The Langscape 4 questionnaire is

244
 combination than is in fact possible. Grammar is corpus-driven in the sense
that the corpus tells us what the facts are. Some of these facts may seem
intuitively obvious in retrospect. But they cannot be predicted in advance
and they certainly cannot be exhaustively documented from intuition.

The implications of these points is that the use of the corpus

material for syllabus design becomes essential rather than merely
desirable if the students are to be taught the actual meanings and usage
of scientific English and thus helped to cope with their bibliographies in
these areas. It is difficult to find a substitute for scientific texts that show
the particular syntax and semantics of scientific English which could be
used for undergraduate study materials without confounding these forms
and thus misrepresenting what scientific English is and means. The
frequency lists would be a starting point but collocations and referring
back from the lists to the actual texts concerned would be essential. A
teaching methodology based on the analysis of comparable or contrastive
uses of lexis in physics and chemistry for example would help to
highlight the different uses of lexical items in scientific texts. White
(1998:268) argues that there are differences in the specialist lexicons of
science and technology, science is characterised by lexicon
“revaleurisation” that is, after Martin (1993), it establishes categories
which reconstrue common sense experiences of reality; technology on the
other hand is characterised by lexicon extension which neither
challenges nor displaces the vernacular system. In the textbooks studied
both of these features are found together, suggesting that these textbooks
really represent the two discourses of both science and technology at the
same time. This situation is probably a result of the different influences
that there have been in science education policy in the United States this
century. Matthews (1994) describes this as moving from the practical

investigating the issue of the preferences for the plurals from Latin.
Field Code Changed
245
application of technology and ‘general’ science giving way to discovery
learning then moving on to a much more elitist ‘pure’ form of science
study and finally recently to a more liberal study of science which
includes the history of science and discussion of the moral, social,
cultural and ethical aspects of the application of science. Matthews also
describes how the pedagogic aspects of the pure science curriculum were
not taken into account and teachers were not involved in the design of
the school curricula which were dictated by scientists alone. This was
especially the case after America felt that it lagged behind the USSR
when Sputnik was launched in 1959. However, this has since been
superseded, as mentioned above, and there is no longer such a
centralised curriculum as prevailed at that time and the twin technology
(applied science) and pure science elements are now integrated in the
modern curriculum.
Furthermore, White (1998:276) argues that technology extends the
everyday sense of terms which are possible because “the polysemous
nature of much vernacular lexis means that different phenomena may be
referenced by the same lexical item”. The use of polysemy to extend the
sense of lexis makes it an important area to concentrate on in teaching to
demonstrate and sensitise the students to this phenomenon in their
reading of such textbooks as these.

McCarthy and Carter (1994:102) advocate teaching by means of a

contrastive approach to specialised and general texts in order to bring to
light the important distinctions and conventionalised patterns in the use
of tense, voice and aspect in specific genre. They suggest developing
student awareness by analyses of the discourse of the texts presented as
a teaching strategy which leads into useful activities that students can
employ themselves when faced with a number of similar types of textual
pattern like that of the report. In other words they argue that it is
possible to teach students how to deal with particular kinds of texts and
246
when they have mastered the techniques taught they will have developed
suitable strategies which they can apply to other texts that they need to
understand.
The use of the corpora to develop materials for use with students
also helps teachers to overcome any preconceptions they might have
about what is, or is not, scientific English. The actual instances of use
are what is important with no need to invent examples which may be
entirely inappropriate. The teaching strategies recommended by
McCarthy and Carter described above would also increase the teacher’s
own understanding of the nature of scientific English, which is often an
area that is overlooked. Furthermore, these insights into usage and
semantics must be used in the testing that is carried out on the students
to ensure that they truly have a command of the English of science and
technology rather than a command of the perceived ideas of what the
English of science and technology is, from a humanities trained teacher’s
point of view.

6.3 Textual Features Compared with Biber’s (1988) Variation

Results.

Biber was dissatisfied by the search for absolute distinctions

between speech and writing. He (1988:93) decided that there were no
absolutes between these but it was possible to grade texts on a number
of factors that were characteristically present or absent in particular
contexts. He gave these groups of factors the headings;

Dimension 1 ‘Involved versus Informational Production’,

Dimension 2 ‘Narrative versus Non-Narrative Concerns’,
Dimension 3 ‘Explicit versus Situation-Dependent Reference’
Dimension 4 ‘Overt Expression of Persuasion’
Field Code Changed
247
 Dimension 5 ‘Abstract versus Non-Abstract Information’
Dimension 6 ‘On-line Informational Elaboration’

Dimension 1 is concerned with to what extent a text contains those

elements that can be associated with spontaneous speech. A high score
on this dimension would show discourse that was interactive,
fragmented, affective and generalised, associated with not having time to
prepare what is going to be said. This is contrasted with text that has a
lot of information which has been carefully integrated into it with very
precise lexical choice which would require careful planning and time to
accomplish and be much more clearly pre-prepared which would be
shown as a negative score.
Dimension 2 is very straightforward and deals with the presence or
absence of the characteristics of narrative, defined as describing past
events and referring to participants in those events. The opposing,
negative factors on this dimension are those of a more static, descriptive
or expository discourse.
Dimension 3 measures the extent to which a text shows devices for
the explicit, elaborated identification of referents. The negative factors on
this dimension show reference to places and times outside of the text
itself. In other words this dimension distinguishes between highly
explicit, context-independent reference and non-specific, situation-
dependent reference.
Dimension 4 shows features that function together to mark
persuasion: either explicit marking of the speaker’s own persuasion
(point of view) or argumentative discourse designed to persuade the
addressee.
Dimension 5 examines the factors that show a text to be abstract,
formal and highly technical.

248
 Dimension 6 distinguishes discourse that is informational but is
produced under real-time conditions so that it displays fragmented
presentation of information with tacking on of clauses rather than
carefully integrated presentation of information.
Biber (1988:94) explains how he standardised the frequencies of
the features in each factor so that those features that occurred with great
frequency would not have an inordinate influence on the factor score.
Applying the same calculations to the results obtained from the physics
and chemistry main and sub-corpora, these texts can be compared with
the corpus as a whole which Biber examined. All of the features were
standardised to a mean of 0.0 and a standard deviation of 1.0. The
results are as follows:

Table 6.1. Mean scores of each of the Dimensions compared with Biber’s Academic Prose
corpus results
Physics Main Chemistry Physics Sub- Chemistry Sub- Biber’s
Corpus Main Corpus corpus corpus Academic
Prose
Dimension 1 - 7.65 - 8.03 - 1.8 - 5.96 - 14.9
Dimension 2 - 5.06 - 6.02 - 3.23 - 2.41 - 2.6
Dimension 3 5.34 2.75 1.89 - 1.07 4.2
Dimension 4 - 5.42 - 4.79 1.56 - 2.37 - 0.5
Dimension 5 5.72 3.53 4.45 11.44 5.5
Dimension 6 - 0.73 - 1.57 0.94 - 1.31 0.5

The following figures show the results of the corpora examined in this
study compared with Biber’s main texts mean scores for each of the six
dimensions.

Field Code Changed

249
Figure 6.1 Dimension 1 ‘Involved versus Informational
Production’
40 |
|
| telephone conversations
|
|
35 | face-to-face conversations
|
|
|
|
30 |
|
|
|
|
25 |
|
|
|
|
20 | personal letters
| spontaneous speeches
| interviews
|
|
15 |
|
|
|
|
10 |
|
|
|
|
5 |
| romantic fiction
| prepared speeches
|
|
0 | adventure fiction
| mystery fiction
| general fiction; physics sub-corpus
| professional letters
| broadcasts
-5 |
| science fiction; chemistry sub-corpus
| religion; physics main corpus
| humor; chemistry main corpus
| popular lore
-10 | editorials; hobbies
|
|
|biographies
|press reviews
|academic prose; press reportage
-15 |
|
|
|
| official documents
-20 |

250
6.3.1 Discussion of Dimension 1 ‘Involved versus Informational

Production’

Factor 1 contains features that are present and absent in certain

types of texts. The features that make up Factor 1 on the positive scale
are:
private verbs
THAT deletion,
contractions
present tense verbs
2nd person pronouns,
DO as pro-verb,
analytic negation,
demonstrative pronouns,
general emphatics,
1st person pronouns,
pronoun IT,
BE as main verb,
causative subordination,
discourse particles,
indefinite pronouns,
general hedges,
amplifiers,
sentence relatives,
WH questions,
possibility modals,
non-phrasal co-ordination,
WH clauses and
final prepositions.

The negative features on this scale are:

nouns,
word length,
type/token ratio, and
attributive adjectives.

Biber explains that the negative features on this factor are all
associated with careful, precise presentation of informational content,
which is not usually a characteristic of speech, whereas the positive
features are characteristic of “on-line” information that is to say, Field Code Changed
251
information that is produced immediately or what teachers usually refer
to as ‘thinking on one’s feet’ and show involvement and interactive or
affective purpose. Biber (1988:132) does not however see this dimension
as a distinction between speech and writing per se but rather as “the
interpretation of involved real-time production versus informational,
edited production”. White (1998:289) argues that ‘hedges’ mark one of
the differences in the lexico-grammar of the scientific and vernacular
technological systems of valeur which would suggest a much more subtle
refinement would have to be made between the corpora included in
Biber’s study in order to separate the scientific from the technological.
Figure 6.1 shows that the results for both the main and the sub-
corpora are generally in the same direction as Biber’s findings. This is not
surprising as the texts are highly informational. There are some features
included in this Dimension however, that might help to explain why all of
the results are higher than those Biber found for academic prose. For
example, one of the factors was WH questions which were found to be
significantly higher in both of the main corpora (see Chapter 5, Tables 5.4
and 5.5). This is one of the features found in large numbers in both of the
textbooks for undergraduates studied here, which contain several pages
of problems for the students to solve at the end of each of the chapters.
The Physics sub-corpus shows a significantly higher frequency of
pronoun it and analysis done by McCarthy (1994-98:275) provides a
tentative conclusion on the uses of it, this and that in texts, seen in this
dimension as demonstrative pronouns and pronoun it:

(1) It is used for unmarked reference within a current entity or focus of attention.
(2) This signals a shift of entity or focus of attention to a new focus
(3) That refers across from the current focus to entities or foci that are non-current,
non-central, marginalizable or other attributed.

252
 McCarthy (ibid.) sees this kind of finding as raising fundamental
questions about “how writers (and speakers) structure their arguments,
create foci of attention in texts and signal desired interpretations.” The
tentative interpretation I would make here is that the physics sub-corpus
displays different argument structures than the other corpora.
In three of the corpora; the Physics main corpus and the physics
and chemistry sub-corpora, second person pronouns are also found to be
significantly higher than in Biber’s findings (see Chapter 5, Tables 5.3,
5.6 and 5.7). This is because of the essays that are used to demonstrate
real-world applications of the theories discussed in the chapters (see
5.3.4 and 5.3.6 for discussion of the essays used in the sub-corpora). The
intention of the authors is more ‘involved’ and ‘affective’ in order to teach
the reader. Biber, Conrad and Reppen (1998:149-150) suggest that “first-
and second person pronouns, wh-questions, emphatics, amplifiers, and
sentence relatives can all be interpreted as reflecting interpersonal
interaction and the involved expression of personal feelings and
concerns.” Glaser (1982:78) found that “emotive features and figures of
speech alongside with the visual code are predictable characteristics” of
the ESP style of using analogies from the learner’s everyday experience.
On the other hand, factors such as nouns and attributive
adjectives, which were seen as negative factors for this dimension, were
significantly lower in both of the physics and chemistry main corpora
with the exception of both the sub-corpora for nouns and the chemistry
sub-corpus for attributive adjectives. The effect of this would be to raise
the result more towards the centre of the scale, as can be verified in
Figure 6.1. The physics sub-corpus then shows affinity with fiction rather
than academic prose which is understandable given the topic of Gulliver’s
Travels as mentioned earlier.
The implications of these findings for teaching and syllabus design
is to reconsider whether some other text types and cultural topics should
Field Code Changed
253
not be included in science and technology courses both for the subjects
covered and for the textual attributes that pertain to them. Sports are
used consistently as a means to involve the (student) reader, as can be
seen from the following extracts from the corpora. In the physics corpora
there are references to American Football, Golf and Baseball as in:
Physics Text

PROBLEMS

34. A quarterback takes the ball from the line of scrimmage, runs backward
for 10 yards, then sideways parallel to the line of scrimmage for 15 yards.
At this point, he throws a 50-yard forward pass straight downfield
perpendicular to the line of scrimmage. What is the magnitude of the
football's resultant displacement?
36. A novice golfer on the green takes three strokes to sink the ball. The
successive displacements are 4 m due north, 2 m northeast, and 1 m 30º
west of south. Starting at the same initial point, an expert golfer could
make the hole in what single vector displacement?
4. A golf ball is hit off a tee at the edge of a cliff. Its x and y coordinates
versus time are given by the following expressions:
In addition, the spin of a projectile, such as a baseball, can give rise to
some very interesting effects associated with aerodynamic forces (for
example, a curve thrown by a pitcher).

However, the use of the terms and expressions “quarterback”,

“scrimmage”, “golfer”, “green”, “strokes”, “sink the ball”, “tee”, and “cliff”
would be particularly dense for foreign language students of science and
technology.
The Chemistry corpus contains the following reference to Table
Tennis which also requires considerable processing to understand:

5.28 A dented (but not punctured) Ping-Pong ball can often be restored to
its original shape by immersing it in very hot water. Why?

This corpus also contains references to basketballs and tennis balls

but without the same level of difficulty displayed above. The Chemistry
corpus also contains such items of a ‘real-world’ type as the following:

254
 5.29 Discuss the following phenomena in terms of the gas laws: (a) the
pressure in an automobile tire increasing on a hot day, (b) the "popping" of
a paper bag, (c) the expansion of a weather balloon as it rises in the air, (d)
the loud noise heard when a light bulb shatters.
5.30 Nitric oxide (NO) reacts with molecular oxygen as follows The heat
generated in this reaction helps melt away obstructions such as grease, and
the hydrogen gas released stirs up the solids clogging the drain.

The predicted areas of difficulty in the questions illustrated above

are American “tire” (British “tyre”), “light bulb shatters”, “grease” and
“solids clogging the drain”.
As mentioned above, these items occur most often in the ‘problems’
section at the end of the chapter and are characteristic of this kind of
textbook. As such, they need specific attention to help students to cope
with this written but interactive style. The fact that these are nothing to
do with the language of science implies that the undergraduate textbooks
do not conform to what many expect in terms of scientific language. The
idea that the language of science excludes other types of discourse is a
failing of many ESP coursebooks which must be remedied in the
syllabus.

Field Code Changed

255
Figure 6.2. Dimension 2 ‘Narrative versus Non-Narrative Concerns’
| romantic fiction
7 |
|
|
|
|
6 | mystery, science and general fiction
|
|adventure fiction
|
|
5 |
|
|
|
|
4 |
|
|
|
|
3 |
|
|
|
|
2 | biographies
|
|
| spontaneous speeches
|
1 | humor
| prepared speeches
|
| press reportage
| personal letters
0 | popular lore
|
|
| face-to-face conversations
| religion; editorials
-1 | interviews
|
|
| press reviews
|
-2 | telephone conversations
| professional letters
| chemistry sub-corpus
| academic prose
| official documents
-3 | hobbies
| broadcasts; physics sub-corpus
|
|
|
-4 |
|
|
|
|
-5 | physics main corpus
|
|
|
|
-6 | chemistry main corpus

256
 6.3.2 Discussion of Dimension 2 ‘Narrative versus Non-

Narrative Concerns’3

The results for Dimension 2 are once again in keeping with the
general tendency for academic prose, although the results for the main
physics and chemistry corpora are an exaggeration of the tendency
towards non-narrative concerns as Figure 6.2 shows. The features that
Biber grouped under the heading ‘Narrative versus Non-narrative
Concerns’ were:
past tense verbs,
third person pronouns,
perfect aspect verbs,
public verbs,
synthetic negation and
present participial clauses.

There were no significant negative features used in the calculation

of this dimension. Although present tense and attributive adjectives were
found to be negative weights, these were not included in the calculation
for this dimension.
The features that contributed to the exaggerated negative effect of
this result were the significantly lower results found for past tense,
perfect aspect verbs and third person pronouns on both of the main
corpora, past tense on the physics sub-corpora and perfect aspect verbs
on the chemistry sub-corpus. In addition, public verbs were significantly
lower on the chemistry sub-corpus. It is interesting to note that if Biber
had retained the present tense and attributive adjectives as negative
factors on this dimension4, both the chemistry and the physics main

3
In Biber, Conrad and Reppen (1998:148) this dimension is relabelled “Narrative versus non-narrative
discourse”.
4
The factors for these two were eliminated by Biber because he included each feature on only one factor
score in order to maintain their independence although he (1988:89) found them to have factorial scores
of - . 47 for present tense verbs and - .41 for attributive adjectives which he regarded as salient in his
calculations.
Field Code Changed
257
corpora for attributive adjectives and the physics and chemistry sub-
corpora for present tense would have the effect of emphasising the
tendency towards lower results than those Biber found. This implies that
the effect of including these features in the calculations on this
dimension would have been to produce negative weightings on this
dimension with the result that the corpora would have shown an even
more extreme negative trend and would have increased the distance from
any of Biber’s findings even further.
Biber (1988:137-8) sees non-narrative purposes as

“(1) the presentation of expository information, which has few verbs and
few animate referents; (2) the presentation of procedural information,
which uses many imperative and infinitival verb forms to give a step-by-
step description of what to do, rather than what somebody else has done,
and (3) description of actions actually in progress.”

These findings would conform to the more conventional view of

scientific method and therefore of ‘scientific’ discourse. However, Trimble
(1985:126) presents a much more semantic analysis of the use of these
tenses in scientific discourse:
“if writers use the past tense in reporting research done previously by
themselves or by others then that research is of secondary importance to the
current work being reported on. If, on the other hand, the writer uses the
present perfect or the present tense, then the research is of more direct and
primary importance to the writer’s current work. Also, the present tense is
often chosen when a discussion follows the initial citing of a reference to
their own or the others’ research and/or when important generalisations are
being expressed.”

Trimble (1985:123-4) suggests that there are three areas where the non-
temporal use of tense regularly occurs in written EST discourse and
258
these are: 1. when apparatus is described, 2. when reference is made to a
visual aid, and 3. when previously published research is referred to.
Points 1 and 2, describing apparatus and making reference to a
visual aid are significant in the corpora studied with such exhortations in
the physics corpus as:

“See Problem 54 for definition”

“see Appendix B.2”
“we see from Equation 3.14 …”
“Figure 1 shows the …”
“Fig. 2.14 summarizes the signs of …”

But we also have the exhortation to:

“Try it and see!”

Which would seem to be a somewhat less formal style than would be

expected from scientific prose. The implications of these findings are that
the relationship between the stylistic representation of information in
academic prose like this needs to be given particular attention. Laurillard
(1993-97:27), discussing the fact that academic education is concerned
with ‘mediating learning’, points out that it ‘relies heavily on symbolic
representation as the medium through which it is known’ and although
this medium is often language it may also be ‘mathematics symbols,
diagrams, musical notation, phonetics, or any symbol system that can
represent a description of the world, and requires interpretation’. Aiding
students in this interpretation process in the scientific and technical
fields, as well as with the language itself as it is a foreign one, is
particularly important. Laurillard (ibid.) laments the fact that although
Field Code Changed
259
interpretation has been subject to research at secondary school level,
very little has been done at university level in this respect. Further
research is necessary to try to clarify the issue of how students interpret
graphical and symbolic information, as it is apparent that
misconceptions do occur and with understanding of the problem suitable
solutions could be found.

260
Figure 6.3 Dimension 3 ‘Explicit versus Situation-Dependent

Reference’

| official documents
7 |
| professional letters
|
6 |
|
| physics main corpus
5 |
|
| press reviews; academic prose
4 |
| religion
|
3 |
| chemistry main corpus
| popular lore
2 |
| editorials; biographies; physics sub-corpus
| spontaneous speeches
1 |
|
| prepared speeches; hobbies
0 |
| press reportage; interviews
| humor
-1 | chemistry sub-corpus
| science fiction
|
-2 |
|
|
-3 |
| general fiction
| personal letters; mystery & adventure fiction
-4 |
|
|
-5 |
| telephone conversations
|
-6 |
|
|
-7 |
|
|
-8 |
|
|
-9 | broadcasts

Field Code Changed

261
6.3.3 Discussion of Dimension 3 ‘Explicit versus Situation-

Dependent Reference’5

Dimension three is derived from both positive and negative features

on factor three. The positive features are:
WH relative clauses on object position,
pied-piping constructions,
WH relative clauses on subject position,
phrasal coordination and
nominalizations.

The negative weights are:

time adverbials,
place adverbials and
adverbs.

The features that are found to be significantly at variance with

Biber’s academic prose on this dimension are Total Adverbs,
Nominalizations and Phrasal Co-ordination. Total adverbs are found to be
significantly lower in all of the corpora while Nominalizations and Phrasal
Co-ordination are both found in significantly higher numbers in the main
corpora. Phrasal co-ordination is more complex with a discrepancy being
found between the physics and the chemistry sub-corpora. The physics
sub-corpus has a significantly greater number of these, whilst the
chemistry sub-corpus has a significantly lower number. The chemistry
sub-corpus also has less analytic negation. The combination of analytic
negation, be as main verb and non-phrasal co-ordination according to
Biber (1988:106) can all be associated with “a fragmented presentation of
information, resulting in low informational density.” This result then
argues for high information density in the chemistry sub-corpus.

5
In Biber, Conrad and Reppen (1998:148) this dimension is relabelled “Elaborated versus situation-
dependent reference” because it is characterised by “highly explicit, context-independent reference versus
situation-dependent reference”.

262
 Biber (1988:110) says that WH relative clauses together with
phrasal co-ordination and nominalization show referentially explicit
discourse which is usually integrated and informational. He (ibid.)
suggests that this dimension distinguishes between endophoric and
exophoric reference (Halliday and Hasan 1976). This would place the
chemistry sub-corpus in a different category from all the other academic
prose categories and from the main corpora.
Biber, Conrad and Reppen (1998:153) describe the use of wh-
relative clauses (including pied-piping constructions) as specifying “the
identity of referents within a text in an explicit and elaborated manner”
whereas time and place adverbials “are used for text-external references
to the physical context of the discourse. The following extracts are taken
from the beginning and the end of the sub-corpus essay to demonstrate
these features:

When the space shuttle Challenger exploded in flight on January 28, 1986,
the crew cabin separated from the rest of the orbiter and broke up when it
hit the water. The cabin was equipped with tape recorders to collect shuttle
data and record conversations among the crew. However, there was no
"black box" to protect the tapes as is used in airplanes. Thus, when the
tapes were found six weeks later in 90 feet of water they were considerably
damaged by exposure to seawater and resultant chemical reactions. The
tapes were described as "a foaming, concretelike mess, all glued together."

The recording showed that at least some of the crew members were aware
in the final seconds that the shuttle was in trouble. The impressive fact
about this tape-salvaging project is that the principle involved is no more
complex than what you would encounter in an introductory chemistry
experiment!

Field Code Changed

263
 The extracts show the frequent use of exophoric referencing for
example: space shuttle Challenger; the orbiter; "black box”; The recording;
an introductory chemistry experiment.
And clauses such as: When the space shuttle Challenger exploded in
flight on January 28, 1986; when it hit the water; when the tapes were
found six weeks later in 90 feet of water; what you would encounter in an
introductory chemistry experiment.
In his discussion of adjectives and adverbs Biber (1988:237) refers
to his earlier work as finding that the distribution of these together with
prepositions and subordinate features were varied. He finds that
prepositional phrases are most frequent in formal abstract styles and
subordination in highly interactive, unplanned discourse. Here again only
the physics sub-corpus is significantly different in this respect with a
higher figure for prepositional phrases than Biber found for academic
prose. All of the corpora show significantly fewer adverbs than Biber’s
corpora did and therefore, a more negative weighting in this dimension.
The implications of these findings are that the sub-corpora are seen
to be difficult to understand if the context is not properly understood.
The need for sufficient background knowledge is made particularly
apparent in this dimension if the texts are to be understood. Whether the
students have sufficient background knowledge to enable them to
interpret these texts correctly is then the problem to be investigated. The
focus for teaching in this case would be in identification of the
referencing in these essays and for learning would be practice with
identifying different types of referencing within texts such as these.

264
Figure 6.4 Dimension 4 ‘Overt Expression of Persuasion’

4 |
|
|
| professional letters
|
|
3 | editorials
|
|
|
|
|
2 | romantic fiction
| hobbies
| personal letters
| physics sub-corpus
|
|
1 | interviews; general fiction
|
|
| telephone conversations; prepared speeches
| spontaneous speeches; religion
|
0 | official documents
| face-to-face conversations; humor; popular lore
| academic prose
| biographies; mystery and science fiction; press reportage
|
|
-1 |
| adventure fiction
|
|
|
|
-2 |
|
| chemistry sub-corpus
|
|
| press reviews
-3 |
|
|
|
|
|
-4 |
|
| broadcasts
|
|
| chemistry main corpus
-5 |
|
| physics main corpus
|

Field Code Changed

265
6.3.4 Discussion of Dimension 4 ‘Overt Expression of

Persuasion’6

The features used for this dimension are:

prediction modals,
necessity modals,
possibility modals,
conditional clauses,
suasive verbs,
infinitives and
split auxiliaries.

Biber (1988:148) sees all of these functions as “overt markers of

persuasion in one way or another”. Biber (ibid.) regards professional
letters and editorials as “opinionated genres”, which therefore have a high
score on this dimension.
Both of the main corpora in this study show extremely negative
scores compared to Biber’s findings. However, the physics sub-genre
shows a much more positive score than most of Biber’s genres, making
this a much more opinionated text. The Gulliver’s Travels text provides
the following extract which shows the use of modals and conditionals in
this sub-corpus.
Within our present technology our scaling arguments are important. If we
design a new large object on the basis of a small one, we are warned that
new effects too small to detect on our scale may enter and even become the
most important things to consider. We cannot just scale up and down
blindly, geometrically, but by scaling in the light of physical reasoning, we
can sometimes foresee what changes will occur. In this way we can employ
scaling in intelligent airplane design, for example, and not arrive at a jet
transport that looks like a bee - and won't fly.

6
In Biber, Conrad and Reppen (1998:148) this dimension is relabelled “Overt Expression of
Argumentation” because (1998:155) this dimension “marks the degree to which persuasion is marked
overtly, whether marking the speaker’s point of view, or the speaker’s attempt to persuade the addressee.”

266
 Predictive modals are used to refer to the future and consider
events that will or will not occur (e.g., what changes will occur, won’t fly),
possibility modals and conditional clauses are used to consider different
perspectives on the issue (e.g., If we design a new large object, may enter,
We cannot just scale up and down blindly, we can sometimes foresee, In
this way we can employ scaling)
In contrast, the other corpora are all to be seen as not involving
opinion or argumentation at all. The fact that science texts neither show
doubts nor allow alternative points of view or argumentation of the facts
presented to the reader may be one of the reasons that they are said to
exclude (see Halliday and Martin 1993). On the other hand, once again
(see Glaser above) the physics sub-corpus is an example of the author’s
attempt to be open-ended and include the student reader in the
discourse.
The problems that need to be dealt with as seen from the main
corpora are therefore similar to the problems that native speakers would
have with scientific texts, which is understanding the concepts developed
by the authors or as Laurillard (1993-7:27) says “the problems stem from
the fact that the two worlds, of everyday knowledge and academic
knowledge, are not as synergistic and inseparable as Vygotsky suggested,
but are contrasting and separate.” Students need to learn what experts
are telling them rather than what they can observe from everyday
experience, thus they need to develop academic knowledge of the world.

Field Code Changed

267
Figure 6.5 Dimension 5 ‘Abstract versus Non-Abstract
Information’
11 | chemistry sub-corpus
|
|
|
10 |
|
|
|
9 |
|
|
|
8 |
|
|
|
7 |
|
|
|
6 |
| physics main corpus
| academic prose
|
5 |
| official documents
| physics sub-corpus
|
4 |
|
|
| chemistry main corpus
3 |
|
|
|
2 |
|
| religion
| hobbies
1 |
| press reviews
| press reportage
| professional letters; editorials
0 | popular lore
|
| humor; biographies
|
-1 |
|
|
| broadcasts
-2 | prepared speeches; interviews
|
| general, science and adventure fiction; spontaneous speeches
| mystery fiction; personal letters
-3 |
| romantic fiction; face-to-face conversations
|
| telephone conversations

268
6.3.5 Discussion of Dimension 5 ‘Abstract versus Non-Abstract

Information’7

The features included in this dimension are:

conjuncts,
agentless passives,
adverbial past participial clauses,
by-passives,
past participial WHIZ deletions and
other adverbial subordinators.

Biber interprets this dimension as distinguishing genres with an

abstract and technical focus from the other genres. In this case all of the
corpora subscribe to this difference, although the chemistry main corpus
and the physics sub-corpus are not as pronounced as the others, they
are nevertheless considerably higher than any of the other genres except
official documents. The chemistry sub-corpus is much higher than any of
the others and the extract given earlier (Dimension 3) demonstrates why,
with the large number of passives: was equipped, were found, were
considerably damaged by exposure to seawater, tapes were described,
were aware and the WHIZ deletion the principle involved is no more
complex. The fact that the agent is dropped or demoted is interpreted as
resulting in a “static, more abstract presentation of information”.
Agentless passives are four times more common than by-Passives in the
corpora which suggests that the writers state facts not “hypotheses for
which some named source is responsible, and which are open to different
interpretations” (Stubbs 1996:147). Biber, Conrad and Reppen (1998:76)
find that there is a preference in academic prose for “generalised states

7
In Biber, Conrad and Reppen (1998:148) this dimension is relabelled “Impersonal versus non-impersonal
style” because this dimension marks “informational discourse that is impersonal, technical, and formal in
style versus other types of discourse.”
Field Code Changed
269
and processes” rather than “the descriptions of specific people performing
actions” which will be found in fiction and conversation.
As was mentioned above for overt expression of persuasion, the
problem to be approached here is the same as that found in native
speaker students in higher education having to learn the academic
representations of the world which are presented as facts rather than
negotiable concepts where argumentation is possible. Many teaching
materials produced for students of science and technology have treated
the use of the passive as a transformation of the active voice in
sentences, which is to misrepresent the meaning of the passive voice.
Discourse analysis has shown how power and authority are confirmed
through the use of the passive and foregrounding of information in texts
(see van Dijk 1997, Stubbs 1996). The use of the passive is deliberate in
science and technology texts to achieve this authoritarian position vis-à-
vis the reader, who is therefore not permitted to question the “laws” put
forward in the text.

270
Figure 6.6 Dimension 6 ‘On-Line Informational Elaboration’

| prepared speeches
|
|
| interviews
3.0 |
|
|
|
| spontaneous speeches
2.5 |
|
|
|
|
2.0 |
|
|
|
|
1.5 | press editorials; professional letters
|
|
|
|
1.0 | religion
| physics sub-corpus
|
|
|
0.5 | academic prose
|
| face-to-face conversations
|
|
0 |
|
|
| bibliographies
|
-0.5 |
|
| hobbies; physics main corpus
| popular lore
| press reportage; official documents; telephone conversations
-1.0 | press reviews
|
| romantic fiction
| broadcasts; chemistry sub-corpus
| personal letters
-1.5 | humor
| general fiction and science fiction; chemistry main corpus
|
|
| mystery and adventure fiction
-2.0 |
|
|
|
|

Field Code Changed

271
6.3.6 Discussion of Dimension 6 ‘On-Line Informational

Elaboration’

The features that have high positive weights on this dimension are:

THAT clauses as verb complements,

demonstratives,
THAT relative clauses on object positions, and
THAT clauses as adjective complements.

Biber interprets this dimension as indicating informational

elaboration under strict real-time conditions, which would explain why
speeches have a high score on this dimension. Biber (1988:159) also
suggests that the features on this dimension “enable a direct encoding of
attitude or stance” which would explain the relatively high results for
professional letters, editorials and religion. The physics sub-corpus is
just below the latter which suggests that it is more attitudinal or takes a
particular stance more than is usual in academic prose. One of the other
characteristic features that Biber found for this dimension but did not
include (because its factorial weight was only .32) was existential there.
This feature is found to be significantly lower in the physics main corpus
than that found by Biber. Stubbs (1996) finds existential there to be used
with an unattributed source which is therefore impersonal and no
responsibility is given for the information presented in this way. The code
sources are of objective, public knowledge. The implication is that the
texts generate their own truth in this way without the reader being
involved in interpretation.
Examples of THAT complements and demonstratives from the
physics sub-corpus are given below:
Whereas, if the size of a body be diminished, the strength of that body is
not diminished in the same proportion, indeed, the smaller the body the
greater its relative strength. Thus a small dog could probably carry on his

272
 back two or three dogs of his own size; but I believe that a horse could not
carry even one of his own size.
If we go far enough toward the very small, surfaces no longer appear
smooth, but are so rough that we have difficulty in defining a surface.
Other descriptions must be used. In any case, it will not come as a complete
surprise that in the domain of the atom, the very small, scale factors
demonstrate that the dominant pull is one which is not easily observed in
everyday experience. Such arguments as these run through all of physics.

It is interesting to note the use of “I believe” here which reinforces the

idea of expression of attitude in this sub-corpus. This example appears in
an imagined dialogue with Galileo included in the essay. Quotation is
often a source of unusual examples in corpora as they may refer to some
period of time from which modern language usage has altered.
Recognition of ‘dated’ language, although obvious to the native speaker,
would require advanced knowledge of the foreign language student. This
stylistic device in the text would only serve to confuse the foreign
language student with the use of such features as the subjunctive, which
was seen through the tests to be one of the features that most of the
undergraduate students coming into the university in the first year
science and technology courses were not much acquainted with.
All of the other corpora examined produced negative results on this
dimension which suggests that they neither express attitudes nor
opinions nor had any kind of time constraints on their production. Biber,
Conrad and Reppen (1998:75) find that there is a preference for to-
clauses rather than that-clauses in academic prose because “academic
prose has a preference for static rather than dynamic packaging of
information”. Academic prose presents the idea that ‘this is so and has
always been so’, rather than the idea that change is taking place
constantly.
Field Code Changed
273
6.4 Academic Prose Sub-Genres

Biber found considerable variation amongst the sub-genres of

academic prose. He examines the following sub-genres: Natural Science,
Medical, Mathematics, Social Science, Politics/Education, Humanities,
and Technology/Engineering for the six dimensions. Consideration of the
relationship between the corpora studied here and Biber’s sub-genres will
also provide further insights into where these text-types fit in the pattern
of the academic prose genre as studied by Biber.
Table 6.2 shows the main physics and chemistry corpora compared
with the sub-genre studied by Biber on the six dimensions. Table 6.3
shows the sub-corpora relative to Biber’s sub-genre which were taken
from the major LOB or London-Lund sub-categories. The LOB corpus has
been the subject of criticism for causing difficulties with the studies
carried out to produce the Leuven English teaching vocabulary list
because of its journalistic leanings (and also the Brown Corpus -
Goethals, Engels and Leenders 1990 mentioned earlier). This may also
have an effect here. White (1998:282) finds there are differences between
what would be expected in popular technological magazines in such
things as their use of explications of acronyms, although he found
popular science magazines met expectations in this regard. A word of
caution must therefore be given in the interpretation of these findings,
which may be affected by inappropriate text-type style being included in
Biber’s study taken from the LOB corpus..
Table 6.2 The main physics and chemistry corpora compared with Biber’s Academic
Prose sub-genres.
Physics Main Biber’s sub- Chemistry Biber’s sub-genre
Corpus genre Main Corpus
Dimension 1 -7.65 -4.4/-14.0 -8.03 -4.4/-14.0
Maths/Soc. Sc. Maths/Soc. Sc.
Dimension 2 -5.06 -4.1 Tech/Eng. -6.02 -4.1 Tech/Eng.

274
Dimension 3 5.34 5.1 Soc. Sc. 2.75 2.7 Nat. Sc.
Dimension 4 -5.42 -2.1 Nat. Sc. -4.79 -2.1 Nat. Sc.
Dimension 5 5.72 7.3 Med. 3.53 3.4 Soc. Sc.
Dimension 6 -0.73 -0.8 Nat. Sc. -1.57 -0.8 Nat. Sc.

Table 6.3 The physics and chemistry sub-corpora compared with Biber’s academic Prose
sub-genres
Physics Sub- Biber’s sub- Chemistry Sub- Biber’s sub-
Corpus genre Corpus genre
Dimension 1 -1.8 -4.4 Maths. -5.96 -4.4 Maths.
Dimension 2 -3.23 -3.1 Maths. -2.41 -2.6 Nat. Sc.
Dimension 3 1.89 2.7 Nat. Sc. -1.07 ---
Dimension 4 1.56 2.6 Pol./Ed. -2.37 -2.1 Nat. Sc.
Dimension 5 4.45 3.7 Pol./Ed. 11.44 9.7 Tec./Eng.
Dimension 6 0.94 0.9 Pol./Ed. -1.31 -0.8 Nat. Sc.

The following figures show the results of the corpora examined in this
study compared with Biber’s mean scores for the academic sub-genres
for each of the six dimensions.

Figure 6.7 Dimension 1 ‘Involved versus Informational

Production’ for the Academic Prose Sub-Genres
0 |
|
| physics sub-corpus
|
| Mathematics
-5 |
| chemistry sub-corpus
| physics main corpus
| chemistry main corpus
|
-10 |
|
|
|
|
| Social Science; Technology/Engineering
-15 | Humanities; Politics/Education
|
| Medical
| Natural science
|
-20 |

Field Code Changed

275
6.4.1 Discussion of Dimension 1‘Involved versus Informational

Production’ for the Sub-Genres

Biber finds the differences on this dimension to be relatively small

for the academic prose sub-genres that he considered and indeed the
corpora studied here also fall into a generally negative grouping on this
dimension. Biber (1988:193) argues that Mathematics has a higher score
on this dimension because its subject matter is “technical and sometimes
non-linguistic, using mathematical expressions instead”. All of the
corpora studied here fall closer to the Mathematics sub-genre rather than
any other and the extensive use of non-linguistic information in these
textbooks has already been discussed earlier (3.1.9.3 Formulae, Numbers,
Equations and Tables).
The distinction which White (1998:267) made between science and
technology lexicogrammar does not appear to have had any significant
effect on the overall results for the corpora studied here. However, other
distinctions that White made between science and technology such as the
use of Latin/Greek borrowings in science and the use of acronyms (or
provisional or proto-nouns) in technology are not specifically included in
the parameters used by Biber and therefore were not applied in this
dimension on the corpora studied. This is an area that could be
researched in the future to see to what extent the corpora represent a
balance between the lexicogrammar of ‘pure’ science and the
lexicogrammar of technology as defined by White (1998).

276
Figure 6.8. Dimension 2 ‘Narrative versus Non-Narrative
Concerns’ for the Academic Prose Sub-Genres
0 |
|
|
|
|
-1 |
| Medical
| Humanities
|
|
-2 |
|
| chemistry sub-corpus
| Natural Science
| Social Science; Politics/Education
-3 | Mathematics
| physics sub-corpus
|
|
|
-4 | Technology/Engineering
|
|
|
|
-5 | physics main corpus
|
|
|
|
-6 | chemistry main corpus

6.4.2 Discussion of Dimension 2 ‘Narrative versus Non-Narrative

Concerns’ for the Sub-Genres

Biber finds this dimension to exhibit considerable variation among

the sub-genres. He (1988:193) argues that the higher score found for
Humanities shows “a topical concern for concrete events and
participants” characterised by historical and biographical studies which
describe and analyse events in the past. In contrast, the
Technology/Engineering sub-genre shows “concern with abstract
concepts and findings rather than events in the past” characterised by
philosophical and analytic studies which deal with abstract, conceptual
information. The differences found in the corpora used in this study have
Field Code Changed
277
already been mentioned above, showing the significant difference in the
use of the past tense in these corpora. It is not surprising that the sub-
genre that are most closely associated with the both main corpora
findings and the physics sub-corpus finding are closest to Technology
and Engineering as these areas are specifically mentioned in the
applications that the textbooks discuss to give analogies with the theory
mentioned. The Chemistry textbook says that these real-world
applications are from technology, engineering, biology, and medicine.

Figure 6.9 Dimension 3 ‘Explicit versus Situation-Dependent

Reference’ for the Academic Prose Sub-Genres
6 |
|
| physics main corpus
5 | Social Science
| Technology/Engineering; Politics/Education
|
4 | Medical
| Mathematics; Humanities
|
3 |
| Natural Science; chemistry main corpus
|
2 |
| physics sub-corpus
|
1 |
|
|
0 |
|
|
-1 | chemistry sub-corpus
|
|
-2 |

6.4.3 Discussion of Dimension 3 ‘Explicit versus Situation-

Dependent Reference’ for the Sub-Genres

Biber interprets the difference found on this dimension as relating

to the explicit and elaborated identification of referents as exhibited by
the Technology/Engineering prose, contrasting with the lack of this in

278
Natural Science which shows situation-dependent reference rather than
inexplicit reference. Biber (1988:193) argues that texts taken from
disciplines such as geology, meteorology, and biology deal with “specific
aspects of the physical environment and thus make extensive reference
to that environment.” The main corpora studied here divide equally
between these two positions, the physics main corpus being on a par
with the Technology/Engineering prose and the chemistry main corpus
on a par with Natural Science. This may not be at all surprising but it
will have to be taken into consideration in developing the syllabus. The
subject matter of the Physics sub-corpus Gulliver’s Travels was shown
earlier (5.3.4) to have extensive biological referencing in it which may
explain why it is even further away from the Technology/Engineering
sub-genre and closer to Natural Science than the main physics corpus.

6.4.4 Discussion of Dimension 4 ‘Overt Expression of

Persuasion’ for the Sub-Genres

Figure 6.10 shows the results for dimension 4 ‘Overt Expression of

Persuasion’. Biber (1988:194) finds dimension 4 to show how
Political/Education texts are quite persuasive or argumentative, while
Social Science prose is “more typical of academic exposition in being non-
persuasive”. This difference he attributes to the purpose and style of the
sub-genre. The higher scores show the “extent to which an author
considers alternative points of view and argues persuasively for a
particular perspective.” The more the study relies on experiment or
empirical data the less it depends on “the logical comparison of
alternatives and the use of persuasive form”. Once again the physics sub-
corpus demonstrates the difference in stance adopted by the author
when he relates the experimental or empirical data given in his exposition
with the real world applications of the facts.
Field Code Changed
279
280
Figure 6.10 Dimension 4 ‘Overt Expression of Persuasion’ for
the Academic Prose Sub-Genres
3 |
|
| Politics/Education
|
|
|
2 |
|
|
| physics sub-corpus
|
|
1 |
|
|
|
|
|
0 |
| Mathematics
| Technology/Engineering
|
| Humanities
|
-1 |
|
|
|
| Social Science
| Medical
-2 |
| Natural Science
| chemistry sub-corpus
|
|
|
-3 |
|
|
|
|
|
-4 |
|
|
|
|
| chemistry main corpus
-5 |
|
| physics main corpus
|

Field Code Changed

281
Figure 6.11 Dimension 5 ‘Abstract versus Non-Abstract
Information’ for the Academic Prose Sub-Genres
11 | chemistry sub-corpus
|
|
|
10 |
| Technology/Engineering
|
|
9 |
| Natural Science
|
|
8 |
| Mathematics
|
| Medical
7 |
|
|
|
6 |
| physics main corpus
|
|
5 |
|
| physics sub-corpus
|
4 |
| Politics/Education
|
| Social Science; chemistry main corpus
|
3 |
| Humanities
|
|
|
2 |
|
|
|
|
1 |
|
|
|
0 |

282
6.4.5 Discussion of Dimension 5 ‘Abstract versus Non-Abstract

Information’ for the Sub-Genres

Biber (1988:194) describes the difference between the sub-genres

on this dimension as demonstrating the distinction between the strictly
technical and abstract which therefore “do not deal with specific
participants or events” and those which are less technical in nature. He
(ibid.) argues that this strictly technical style where “empirical studies are
factual, and therefore faceless and agentless” from a linguistic point of
view might be seen to result from the style that all scientists and
engineers are explicitly taught. Humanists Biber argues “are taught (and
teach) that passives are dispreferred constructions and that good writing
is active”, a position that is echoed by word processing grammar
checkers! From this point of view the Chemistry main corpus is less
abstract or technical (or more humanistic) whilst its sub-corpus is
extremely abstract and technical. The wide range to be found within this
one textbook argues for considerable variety in the text-types that should
be used with undergraduates to present and practice language through.
Investigation into the styles adopted by teachers in schools in
Portugal to teach scientific writing would also help to shed some light on
the cultural expectations of undergraduates with regard to the need to
use passives in science and technology. Some undergraduates have
considerable difficulty in reorganising information into a series of logical
steps like those found in the report of an experiment. This could indicate
a difference in report writing styles between different cultures.

Field Code Changed

283
Figure 6.12 Dimension 6 ‘On-Line Informational Elaboration’ for
the Academic Prose Sub-Genres
|
|
|
| Mathematics
3.5 |
|
|
|
|
3.0 |
|
|
|
|
2.5 |
|
|
|
|
2.0 |
|
|
|
|
1.5 |
|
|
|
| Medical
1.0 |
| Politics/Education; physics sub-corpus
|
| Social Science
|
0.5 |
|
|
| Technology/Engineering
| Humanities
0 |
|
|
|
|
-0.5 |
|
| physics main corpus
| Natural Science
|
-1.0 |
|
|
| chemistry sub-corpus
|
-1.5 |
| chemistry main corpus
|
|
|

284
6.4.6 Discussion of Dimension 6 ‘On-Line Informational

Elaboration’ for the Sub-Genres

Biber (1988:195) explains how his earlier description of this

dimension taking the that compliments to be a characteristic of discourse
that could not be carefully planned and integrated does not preclude
their use in logical development in a text or for emphasis which indeed
characterise “all academic prose sub-genres … to some extent” and
demonstrates their primary use in such genres. For this reason the
Mathematics sub-genre is seen to be divergent because of the use of
formulae and argumentation in this sub-genre which requires that
compliments to mark logical relations within the text. The corpora
studied here do not show such an extreme position as that found in the
Mathematics sub-genre but, nevertheless, conform to Biber’s idea that
these features are to be found in all academic prose to some extent.

6.5 The English of the Students in the First Year of University.

Two clear trends are apparent from the test results obtained
for the new students in the first year of the university. First, it seems that
the students are studying more English at school. This result is
confirmed by other factors such as the increase in the number of
students studying the English language in schools particularly in the
fifth and sixth years, despite the changes in demographics that the
schools are undergoing. The second feature that can be observed is that
the more years of English the students have had at school the better their
results are on the preliminary test. This may seem obvious, however the
view that the students are not learning anything, and that the same
material is repeated over and over again in each school year as a result of
this, is often bemoaned at conferences on teaching and meets with
Field Code Changed
285
widespread sympathy from teachers. These results run counter to that
general idea. The students are indeed learning more English when they
have more time dedicated to the study of the language. This is not to say
that the students have learnt the specific language of science and
technology but it does suggest that strategies for dealing with the
comprehension of texts have been developed and that these strategies
can be applied by the more advanced students in other situations.
Rosenthal (1996:19), reviewing the research on second language students
and academic success in further education, finds that the level of
language proficiency necessary to ensure academic success takes five to
seven years to develop. She is discussing the American system which has
both immigrants whose first language is not English and foreign students
in higher education and notes that there are many different systems in
operation within the universities and colleges in the United States to
teach English to such students. However, she also points out that this
time factor cannot be overlooked but is a considerably longer period of
time than that allotted to any of the language programmes in the colleges
and universities. Added to this she recognises that the fact that faculty
members in the other academic disciplines often have no idea how
students learn English which leads to a separation of English and the
study of the academic subject matter. She recognises that this situation
is no longer appropriate because the acquisition of English occurs best
when students are using the new language purposefully but that many
mainstream faculty members have unrealistic expectations of what
students can achieve within the confines of the language classroom.
Despite the very different circumstances between the American system
she is describing and the Portuguese system examined here, the latter
unrealistic expectation holds true. In other words, time is of the essence
but so is contact with the language in purposeful contexts.

286
 The results that show that the students have a greater ability with
some of the structures traditionally regarded as difficult or more
advanced and struggle with those traditionally considered simple call into
question these distinctions. Examination of the irregular verbs in the
frequency lists for the chemistry and physics corpora reveal that the
results of Mindt’s corpora analyses (1997:47-49) are not duplicated in the
results of this study. Mindt’s top ten irregular verbs (apart from do, have
and be) are say, make, go, take, come, see, know, get, give and find. Of
these say, go, and come are not found with any frequency in either of the
main corpora and take and get are only found in one of them. On the
other hand, show and write are extremely common in these corpora.
Halliday (1993:19) working on the original COBUILD corpus provides a
list of the first 25 most common verbs, once again without the most
frequent ones of all be, have and do. This list does not coincide with
Mindt’s list entirely either. The most common irregular verb forms as
given by Halliday (ibid.) include think and tell before find and give. Added
to this is the problem of how many of these verbs are actually found in
the past tense forms. Halliday (1993:21) finds that the use of the present
and the past tenses is approximately the same, however, the corpora
studied show a preference for the present tense. These findings suggest
that the syllabus must be built on the evidence contained in the
frequency listings of the corpora used here, rather than on any other
arbitrary corpus, or indeed other studies conducted for other purposes
on other material, in order to be both useful and relevant for the
students. (see McCarthy and Carter 1994:20).
Similarly, the grammatical structures taught should appear in
actual contexts of use so that they reflect the meanings and common
usage of these kinds of textbooks. Biber, Conrad and Reppen (1998)
mentioned above, Trimble (1985), Stubbs (1996) and McCarthy and
Carter (1994) (among others) have all demonstrated how both the
Field Code Changed
287
grammatical choices and their meanings differ between genres and text-
types. The students were not specifically tested to see if they could
distinguish between general English usage and the specific science and
technology usage of certain grammatical items. Nevertheless, their test
results do show that many of the students have not grasped the general
usages they should have learnt in school.8 Of particular importance is
the specific use of modals, conditionals and irregular verbs and their
meanings in scientific texts.

8
Mindt (1997:43-45) argues that the grammars used in schools (in Sweden) do not show the most common
forms of usage of some and any (amongst others) and so misrepresent the language as used by native
speakers in any case. Therefore he argues for a new approach to didactic grammar based on corpus
analysis which would show the most frequent usages for students to learn in a graded manner.

288
Chapter 7 The Syllabus
 Chapter 7

The Syllabus

The language of science and technology is an area where new

information is being introduced more and more quickly; inevitably new
discourses will also develop quite quickly. Changes are being introduced
in schools because of new information and because of new attitudes to
the world engendered by a new understanding. A more environmentalist
and interdisciplinary approach to learning new discourses (Stubbs 1994,
1996, Kerridge and Sammells eds. 1998, Pope 1998 gives an overview) is
being put forward and these changes will have to be studied by educators
and students alike. Areas such as e-mail can be seen to be an altogether
new, recently invented genre, whereas many other new discoveries can be
seen to have arisen from other genres and share some of the
characteristics of those other genres. For example, in technology there
are many new inventions the descriptions of which nevertheless follow
some of the discourse characteristics of the genre of technological style.
What then can be done to prepare students to study these and other new
genres? Stubbs (1992:206) says
IT extends literacy and therefore requires new study skills for school and
work: the newer technologies can provide access to learning resources,
such as documentary films and databases, and therefore lead to an increase
in collections of authentic language products for teaching about the uses of
language. The key question is therefore accessibility. The diversity of self-
access teaching materials and study packages, and the increased availability

Field Code Changed

291
 of CD-ROM, video disk, satellite television, etc. will require an increase in
study skills to access such materials.
7.1 Study Skills

ESP can be broken down into other sub-divisions like EAP, which
is broken down further by Jordan (1997:3) into ESAP (that is, English for
Specific Academic Purposes and EGAP (English for General Academic
Purposes), although he reports that the more usual model in the USA is
to break EAP down into EAP and EST. EAP is seen by Jordan to cover
both the more specific focus of a subject such as engineering and also the
skills and proficiency in formal academic style and register that the
students need for study purposes. In either division, EAP can be taken as
the short term objective of the undergraduates studying on the science
and technology courses analysed here. This is particularly relevant as a
teaching objective given the constraints imposed by both the size of the
classes, their heterogeneous nature and the paucity of time available. The
objective of the discipline would be to cope with the students’ immediate
course concerns. However, the medium and longer term needs of the
students cannot be ignored and a process whereby the students can be
given the skills to continue learning (or the “learning to learn” dimension)
must not be ignored.

Jordan (1997:7) suggests the following breakdown of study skills:

STUDY SITUATION/ACTIVITY STUDY SKILLS NEEDED

1. lectures/talks 1. listening & understanding
2. note-taking
3. asking questions for: repetition, clarification
and information
2. seminars/tutorials/discussions/supervisions 1. listening and note-taking
2. asking questions - as above
3. answering questions; explaining
4. agreeing and disagreeing; stating points of
view; giving reasons; interrupting
5. speaking with(out) notes: giving a paper/oral
presentations, initiating comments, responding;
verbalising data
292
3. practicals/laboratory work/field work 1. understanding instructions: written and spoken,
formal and informal
2. asking questions; requesting help
3. recording results
4. private study 1. reading efficiently: comprehension and speed
(journals and books) 2. scanning and skimming; evaluating
3. understanding and analysing data (graphs,
diagrams, etc.)
4. note-making; arranging notes in hierarchy of
importance
5. summarising and paraphrasing
5. reference material/library use 1. using the contents/index pages
2. using a dictionary efficiently
3. understanding classification systems
4. using a library catalogue (subject and author) on
cards, microfiche and computer
5. finding information quickly (general reference
works and bibliographies)
6. collating information
6. essays/reports/projects/case 1. planning, writing drafts, revising
studies/dissertations/theses/research 2. summarising, paraphrasing and synthesising
papers/articles 3. continuous writing in an academic style,
organised appropriately
4. using quotations, footnotes, bibliography
5. finding and analysing evidence; using data
appropriately
7. research in addition to 3-6 above:
(linked with 3-6 above) 1. conducting interviews
2. designing questionnaires
3. undertaking surveys
8. examinations: 1. preparing for exams (techniques)
a) written 2. revision
3. understanding questions/instructions
4. writing quickly: pressure of time
b) oral 1. answering questions: explicitly, precisely
2. explaining, describing, justifying

He describes the skills necessary for these activities to be the following:

Skills generally applicable:

1. organising study time efficiently, i.e. time management
2. logical thinking: constructing arguments - use of cohesive markers and connectives;
recognising weaknesses and bias in arguments; balance; critical analysis
3. accuracy
4. memory: recall; mnemonics
5. using computers/word processors

Field Code Changed

293
 Jordan (1997:8) notes that the term reference skills is sometimes
confused with the generic term study skills. In Jordan's description of
study skills there are a number of repetitions e.g. listening, note-taking
and asking questions, in the various situations that students can be
expected to find themselves. Jordan's framework was produced with
students studying in an English-speaking environment in mind, so some
of these items will not be the most crucial for undergraduate students in
Portugal, although in the initial needs analysis from other departments
one mention was made to listening to lectures given in English which
would suggest that in some circumstances there may be greater
similarities with an English-medium educational situation. Moreover, the
use of questions was found to be of particular importance in the
textbooks studied so written questions will need specific attention on the
syllabus for these specific undergraduates. Stubbs (1992:208-9) points
out that Jordan’s point 5 above – reference skills may be different using
CD-ROM material than for conventional library selection of material. He
questions what novel uses and representations of language and
knowledge are brought about by computer systems and concludes that
“Computers require precise and accurate instructions, and the
production and interpretation of clear and precise information is an
important goal of English teaching.”
Jordan assumes that study skills can be transferred from one area
into another, that is, if the students are proficient in study skills in their
mother tongue, they can transfer these skills to the foreign language
situation. He (1997:5) does accept that students might need help with
this and also with adjustment from school to a different academic
environment. He also recognises that there are other elements of EAP
that may need addressing, such as style, and that some students will not
be proficient in study skills even in their mother tongue. In this
particular instance, it is both the textbook which represents and relates

294
to students from a different culture and academic environment and the
fact that most of the students on the first year discipline are also in the
university environment for the first time.
Laurillard (1993:2) argues that at undergraduate level it is
unrealistic to expect students to take control of their own learning, but
goes on to show how she sees the student developing academic
knowledge through mediated learning. Waters and Waters (1992:264) say
that in their experience “what students frequently lack is not only a
knowledge of study skills, but, more fundamentally, the underlying
competence necessary for successful study - self-confidence, self-
awareness, the ability to think critically and creatively, independence of
mind and so on.” The general development that is taking place in
undergraduates who have just made the transition from school to
university must also be addressed together with the more straightforward
aspect of the lack of study skills. One means of achieving this
development in students is through different methodologies in the
classroom which encourage increasing confidence through working on
different levels of material with success and interacting with classmates
in pairs or small groups to avoid shyness and ridicule and to encourage
the sharing of knowledge between students. More general discussion
encouraging students to explore their own ideas and opinions on topics
which are given proper consideration and treated with respect by the
teacher and other students can aid self-confidence and creativity.
Different schemes which allow students to interact with the lecturer on a
more personal basis such as outside the classroom in attendance hours
and through different means of directing questions which may be
personal (through e-mail) or taken up with the whole group by the
lecturer if they are found to be more generalised can also lead to success
and have the added advantage that the lecturer reflects upon what has
been successfully achieved and what has not be grasped by the students.
Field Code Changed
295
Rosenthal (1996:150-174) describes some approaches adopted in the
United States which encourage lecturers to evaluate their own
performance and laments the fact that often the principles that guide
scientific research are not applied to science teaching. She (1996:178)
sums this up thus:

Few researchers would maintain the same protocol if experiment after

experiment failed. They would reevaluate their hypotheses, reassess the
literature, check for errors, develop new procedures, and redesign their
experiments. Yet, when it comes to their teaching, some of the same
individuals resist any form of change, any innovation, which might
improve student interest and achievement in science.

Rosenthal admits that it might seem rather odd to be comparing

and contrasting such different fields as science and second language
acquisition but she feels that these two merge in the students who are
trying to study science (in America) and who have limited English.
Despite the need being greater in students studying science where the
language of instruction is English, many of the problems she outlines
apply equally to foreign language students who need to understand
English for their science studies and so reflection by the lecturer on their
teaching to aid success in science learning is essential.

7.2 Student Needs and the Syllabus

If decisions are taken based on the needs analysis provided by

other departments and the student profiles gained from testing, the
following areas will have to be addressed in the discipline:
i) The classes will have to be taught in English, thus exercising the
students' listening skills and preparing them for lectures given in
English in some of the Departments. The use of video recordings of
296
 actual lectures given in this or other universities in English would be
important to provide the students with an appropriate context for
language study. This practice would also pave the way for attendance
at conferences where English would almost certainly be one of the
conference languages. 1

ii) Both American and British English will have to be included especially
for comparative purposes as students may have come from different
educational backgrounds where one or the other of these Englishes will
probably have been taught.
iii) Basic scientific reading texts must be used as a core for the discipline
and text attack strategies will have to be taught because the students
are also coming to grips with the more advanced scientific subject
matter (even in Portuguese). Discourse and text analysis - study of
such aspects of scientific texts as cohesion, pronoun referencing,
deixis, linking words, cause and effect, definition and classification. It
would be possible to use corresponding texts in different subject areas
with respect to these (McCarthy and Carter 1994) or even to conduct
analyses of Portuguese and English texts on the same subject (Leech
1997).
iv) Mathematical symbols, formulae and numbers in British and
American English will have to be taught, including contrasting
Portuguese and English use, as these are likely to have been omitted
on school courses but are extremely important to the understanding of
the scientific texts on the students’ bibliographies. Formulae would
also involve the revision of the alphabet and learning how equations
are put into words or form part of the text being studied. The patterns
created by the use of mathematical functions, graphs and tables are

1
Many of the Departments in the University of Aveiro promote conferences aimed at undergraduates, usually in the
final years of their courses, which may have international speakers. The Departments of Management and Tourism
(amongst others) also often have visiting scholars some of whom were American and lecture undergraduates in
English.
Field Code Changed
297
 considered by Lemke (1998:102) to be “important in the value-scheme
of natural science” which the students must be able to perceive in
order to understand the relationship between the patterns and
assumptions made in scientific theory. Weights and measures in both
the metric and imperial systems must be examined including British
and American differences. Consciousness raising and alerting the
students to areas that are different will include looking at some
cultural differences between the two languages as mentioned earlier.
v) Note-taking and summarising should be included in the course as
these are skills that will be necessary throughout the students
courses. The summaries and notes produced by students might well be
in Portuguese if they are for personal use for understanding and
storing information for use in other disciplines. The study of
comparative texts mentioned earlier might help here to highlight
signalling devices in discourse that the students must be conscious of
to organise their notes.
vi) Reference skills such as dictionary work will have to be included to
provide knowledge of sources of information for the students to further
their studies independently but increasingly these reference skills will
have to be extended to the use of CD-ROM material and the Internet.
Dictionary work focusing on abbreviations, countable and uncountable
nouns, spelling and pronunciation is one simple means of providing
the students with a means of discovering more about the language
when they need to. An attempt to equip the students with the means to
proceed further in their studies on their own initiative can be
approached through more detailed study of types of definitions; both
those used in reference materials and those to be found in the
textbooks on the bibliography of both an overt and covert type (Darian
1981) with the corresponding text-type signalling.

298
vii) The students will have to be taught to interpret graphs, tables and
diagrams in English and to recognise the referencing to these in the
text and the means by which they complement or add information to
the main text. Listening activities where the students have to complete
graphs and tables and interpret them into another form must be
included. These activities could usefully be carried out in a language
laboratory. Laurillard (1993-97:112) rates the combination of audio
and visual material as one of the most productive because it gives
greater control to the student and could be used to set tasks that
“enhance and interpret students’ experience of the world”. She (ibid.)
suggests that the visual part need not necessarily be printed material
(which is however the most flexible medium), but may be an object
which the student has to observe or a situation could be created
whereby the audio material guides the student to perform some other
operation, such as work on a computer. Lemke (1998:93) argues that
the juxtaposition and combination of visuals in texts will multiply the
meanings so that “we can mean more, mean new kinds of meanings
never before meant and not otherwise mean-able.” The way in which
this is achieved has to be explored as “the user must integrate visual
and verbal realisations of objects, concepts, relations and processes in
the joint interpretation of text and figure.”(Lemke 1998:110).
viii) Appropriate specific vocabulary will have to be taught in
appropriate contexts with their usual collocations taken from the
corpora, together with the pronunciation of these and the semantic
variation that science and technology texts cause in lexis and their
associated grammatical structures. Specific grammar will have to be
taught with its usual realisations in scientific texts as necessary for the
effective realisation of the tasks undertaken. The development of more
student autonomy in using the corpora for their own difficulties with

Field Code Changed

299
 science and technology language is important given the differing
student backgrounds.

7.3 The Students’ Background Knowledge and the Syllabus

It would appear from the list given above that some of the items
would already have been taught in MT. However, although teachers can
most definitely appeal to frameworks taught in school, it is not a sound
notion that everyone thinks in the same "scientific way". An example of
this is the "choke" on a car. In Portuguese, in older models of cars, the
driver abre o ar literally translated as "opens the air " on a cold morning
in order to start the car. In English exactly the opposite is done, the
"choke " is pulled out, thereby cutting off (choking off) the air to the petrol
mixture. The basic scientific principle, of enriching the petrol mixture, is
the same but it is not expressed in the same way. In other words, on the
surface of this expression there is a different scientific explanation of
what takes place. Halliday and Martin (1993:16) argue that there are
some minor variations among different languages of how grammar
construes phenomena into a scientific theory. They (ibid.) suggest that
English and French are different not so much because the grammar of
scientific theory is different between them, but because the English
language constructs reality more along empiricist lines whilst the French
language constructs reality along rationalist lines in scientific theory. Dr
Catherine Middlecamp, Director of Chemistry at the University of
Wisconsin-Madison (whose report is included in Rosenthal’s survey of
science teaching for language minority students in the USA, 1996) argues
that Western scientists are more inclined to use categories than others.
She explains that although the categories into which chemistry is usually
broken down such as organic, inorganic, analytic biophysical etc. appear
to be culture-free, in reality they are not. Furthermore, she argues that

300
even if two cultures are similar in their tendencies to categorise the world
“there is no guarantee that the lines will be drawn in the same places”.
The question needs to be raised as to whether there is a Portuguese
‘scientific way’ and if so, what this is. Kaplan (1966:15) describes
different scientific discourse patterns in paragraph development
employed by different linguistic systems. He finds that there is a
difference between English and the Romance languages (which would
include Portuguese) because Romance languages include digressions and
include extraneous information. The differences that exist between
different linguistic systems has significance for Portuguese students who
may well have recourse to many of these (scientific) systems. The
students may be unaware that there are distinctions between scientific
approaches when they are consulting books in different languages. The
naming of processes and theories are also often different between the
languages of science the French, for example, have not always adhered to
the International Scientific (SI) system, preferring to coin their own
terms.
Equally well, the fact that reading skills are transferable should be
used to help to get the students to read effectively in English. However, a
number of the students on this course undoubtedly opted for science and
technology because they did not like, or demonstrated less aptitude for,
foreign languages. This being the case, it will be an uphill struggle to
create the conditions necessary for successful transfer of skills. Halliday
and Martin (1997:49) discuss the “ongoing apprenticeship of students
into science discourse” which implies that what has gone before has to be
taken into account to decide what will follow. An attempt to motivate
students, who may well have an active dislike of English, must be made
to encourage them to engage with and enjoy the study of science and
technology through English. As Hutchinson and Waters (1987:141) say,

Field Code Changed

301
 Enjoyment isn't just an added extra, an unnecessary frill. It is the simplest
of all ways of engaging the learner's mind. The most relevant materials,
the most academically respectable theories are as nothing compared to the
rich learning environment of an enjoyable experience. This is an aspect of
pedagogy that is taken for granted with children, but is too often forgotten
with adults. It doesn't matter how relevant a lesson may appear to be; if it
is bores the learners, it is a bad lesson.

One of the factors that could add to enjoyment, and therefore to

learning, is variety and materials must also be chosen to provide this.
However, one of the problems to be overcome in this situation is the fact
that the students have different backgrounds in English and different
interests in scientific or technological subjects, therefore a range of
materials would be necessary to provide practice and to meet the
different levels of English which the students bring to the task. The
results presented in Chapter 6 point to the different subject areas in
academic prose that it would be appropriate to use to provide variety
without sacrificing any of the appropriate discourse features that the
students have to learn to cope with. Furthermore, it is likely that some of
the students stopped studying English more than a year before coming to
university and are therefore somewhat 'rusty'.
Halliday and Martin (1993:71) identify the following as the
difficulties that students have with scientific English:

1. interlocking definitions
2. technical taxonomies
3. special expressions
4. lexical density
5. syntactic ambiguity
6. grammatical metaphor
7. semantic discontinuity

The approach that they suggest to deal with these difficulties is to

analyse any text and to relate it to the context in the discourse, however
302
they point out that this must be done in a principled way. Scientific texts
should be analysed and interpreted in contrast with other texts so that
the similarities and, more importantly, their differences can be studied. If
the tasks are varied, it would be possible to use 'authentic' texts taken
from the corpora and at the same time avoid difficulty for those students
who have a lower level of English whilst still engaging the students whose
English is more advanced. The gradual build-up of both linguistic
knowledge and scientific knowledge could then develop with the use of
schemata theory and other text attack strategies. Steffansen and Joag-
Dev (1984:54) say that no text is explicit and that students need to use
“guessing” strategies to understand the message, however, if the student
does not share the same background information as that of the author
they will “re-interpret vague aspects to conform with their own schemata
and will be unaware of other possible interpretations which in fact
conform to the author’s schemata.” This danger can be avoided, they
suggest, by “establishing a correspondence between what is known …and
the givens in a message” whereby the students themselves can “monitor
their comprehension and know whether they have understood a text.”
This is along the lines of Krashen’s (1982:16) theories of ‘monitoring’
where students can apply what they have learnt about a language to
their own production of language. However, he believes this is only
possible provided that the students have plenty of time available,
correctness of language use is considered important and they can
remember the ‘rules’.
The knowledge the students bring to a task could be pooled to
benefit the whole group and active student-to-student collaboration
encouraged in order to overcome some of the difficulties of teaching such
heterogeneous groups. The benefits of students collaborating with each
other is that they will be more aware of where the difficulties lie, both in
terms of scientific knowledge and language problems and students may
Field Code Changed
303
find it easier to expose their difficulties to a colleague rather than to the
teacher in a large class. Laurillard (1993-97:187) suggests that it is
“impossible for teaching to succeed if it does not address the current
forms of students’ understanding of a subject.” She argues that as
university education becomes less elitist teachers will recognise that
students cannot succeed unaided and there is a need to try to improve
the situation by investigating students needs and using the data obtained
to design (better) future teaching. The point, as Laurillard (1993:93) says,
“is not just to change what is taught, but also how it is learned.”
Bazerman (1998:21) suggests the “Difficulties that students and
others have with scientific language are in the recognition and
appropriate manipulation of the verbal objects which correspond to
conceptual objects” and furthermore, “once the object is given a stable
name, its details, problems and material peculiarities and relations to
other objects in its network vanish in a higher level abstraction which
becomes difficult to unpack once made.” If this is so, the development of
new conceptual frameworks which the students are undergoing in both
the science and technology they are studying and the interpretation of
these same concepts and their unpacking in English must be borne in
mind and allowed for in the syllabus through discussion and
examination of what students understand about the different concepts
contained in the topics covered.
It is clear from the study carried out here that the language of
science and technology is different from that of general English and that
these differences need to be addressed by the course (see Chapter 6). The
means this thesis proposes for doing this is through data-driven learning.

7.4 Data-Driven Learning

As syllabus design should always be responsive to learning theory,

it is important to consider what has been discovered about language
304
acquisition in recent years. Over the last two decades there has been
considerable research done on language acquisition which has shown
that children acquire language in unanalysed “chunks” or as some
researchers term it they use “prefabricated language” in certain,
predictable, social contexts. That is to say, children use a kind of formula
of undifferentiated morphemes for many situations they encounter and
only later gradually refine their perception and use of these “chunks” or
“prefabricated pieces” of language. An example of this might be “what is
that?” which is rendered as /hw
sdæt/ or /hw
sdæ/ (See Hakuta 1974;
Huang 1971; Brown 1973; Clark 1974; Cruttenden 1981; Wong-Fillmore
1976; Newmark 1979, Peters 1983)2. The next stage in language
acquisition is the analysis of these phrases and the recognition that they
are made up of a number of words. Many people can remember an
example in second language learning when something was
misunderstood because the words used were undifferentiated or
differentiated incorrectly. An example of this from Portuguese would be
confusion between nouns and verbs with such close items as a baixa -
abaixa and a travessa - atravessa. Other examples may occur even in the
mother tongue when what is heard and how something is written is
confused. An example of this would be the word ‘misled’3, although this is
a slightly different aspect of the same phenomenon of undifferentiation.
Some of the phrases used by children are found to be more
productive because they allow substitution to a greater or lesser extent.
Different items can be inserted in ever greater complexity. An example of
this kind of substitution would be ‘Modal + you + VP’ as in could you
pass the salt?, could you hand me that pencil?, would you lend me a

2
Even in language acquisition there has been a move towards the use of corpora for recent studies. Biber,
Conrad and Reppen (1998:172-202) report on analyses of 8-12 year olds using the CHILDES corpora as
they found that previous research was often limited in scope, used only one or two subjects and focused
on a small number of linguistic features and often a single register.

Field Code Changed

305
dollar? (Nattinger and DeCarrico 1992:18). Aston (1997:56) argues that
“rather than generating and interpreting utterances by combining or
analysing morphemes on the basis of generalised grammatical and
pragmatic rules, users appear to make use of larger memorised chunks
associated with particular types of problem in particular types of
contexts, instantiating these as necessary with relatively simple
modifications.” He claims that work in corpus linguistics by Sinclair
(1991) has provided evidence for this view, showing the extent of
recurrent co-selection of lexicogrammatical forms with relatively simple
variation.
Furthermore, researchers such as Wilson and Sperber (1986 and
1988) suggest that high frequency items require less processing effort on
the part of the hearer and that there is a tendency for those phrases
which are used very often to become fixed. In English this finding can
certainly be confirmed quite simply by looking at the shifts that take
place in individual words. For example, some years ago ‘to-day’ was
hyphenated in this way yet now ‘today’ is written as one word with no
hyphenation. Originally it was two separate words, ‘to’ and ‘day’. This
trend can be seen over and over again in such words as ‘albeit’,
‘nevertheless’ and more up to date still in such words as ‘alright’ which
was written as two separate words until quite recently. Some of these
latest changes can often give rise to arguments about ‘correct’ language
use. There are always those who object to any change occurring in
language use or spelling and, of course, in other cultures there are even
institutions designed to try to stop the rot. The French Academy, for
example, attempts to discourage or even prevent the adoption of English
words and expressions like le weekend. White (1998:285-6) reports that a
similar process of shift takes place in the language of technology where

3
The word is read in combination *mi+sle+ d /mizld/ rather than two parts miss + led /mis'led/. The
confusion arises because of the overgeneralisation of the -ed ending being seen as a past tense suffix
added to the verb which would then be *misle.
306
the “existence of a lexically minimal term - a single word form - to
reference a given category is generally seen as evidence that the category
is stable and salient within its ideational domain.”
Nattinger and DeCarrico (1992) suggest that, although much of the
research done was concerned with language acquisition in children, there
is no reason to believe that adults would go about the language-learning
task any differently, and indeed misunderstandings like those mentioned
above confirm this. Nattinger and DeCarrico (1992) go even further and
suggest that “It is our ability to use lexical phrases, in other words, that
helps us to speak with fluency.”
On the other hand, there is increasing evidence from computer
corpus-based research that language itself occurs in a largely predictable
way. That is, the commonest forms of language occur in overwhelmingly
high frequencies and collocations. Collocation here means the co-
occurrence of certain words within a short space of each other in a text. A
certain word or ‘node’ is the focus of attention and the words to either the
left or the right of it are studied, these are called the collocates. The use
of a concordance which focuses on the node can reveal important
language patterns in texts. Often the position in the sentence can also be
revealed by this type of concordance study of language patterns which is
an important piece of evidence for students to observe for their own use
in writing in English. One of the most important aspects of this
computer-based research is that it is reflecting natural language use,
that is, it is descriptive and not prescriptive and examples of use are not
invented ones which can be, as Sinclair (1991) points out, “extremely
unlikely to occur in speech or writing”. Researchers in these areas report
that the commonest forms are in the majority in the frequency and
collocation studies they have done, no matter how large the corpus they
are using. Sinclair says that if those words that occur only once in a
corpus were removed the corpus would be reduced by half. He also
Field Code Changed
307
suggests that “grammatical and lexical distinctions may be closer
together than is normally allowed”.
What are these common forms and what does this research imply
in terms of language learning and teaching? This research does indeed
suggest that language is a much more finite system than has hitherto
been believed. The commonest language is used most of the time in
predictable or “prefabricated” chunks and it should be this language that
students should be provided with, in order to give them a rapid, fairly
comprehensive grasp of naturally-occurring language in the shortest
possible time scale. The idea of teaching what is most frequent has been
around for a long time (at least since West in the 1920’s); the only danger
is that what is most frequent today will not be the same as what is most
frequent tomorrow and decisions about what needs to be taught should
be based on the most up-to-date data from corpora that are made up
from that specific language that is the students’ target.
Some of the findings from computer corpus-based research run
counter to what intuition about language would suggest and, more
importantly, run counter to what coursebook writers believe to be the
case. This is true of both meaning, form and usage, as the following
examples demonstrate. A list of the commonest meanings of the verb ‘see’
would include ‘using the eyes’, ‘looking at’, ‘meeting’, ‘grasping with the
mind or imagination’, ‘discovering or checking’, ‘experiencing or
witnessing’, ‘other meanings e.g. accompany or escort’ and phrasal verbs
( taken from the Oxford Advanced Learners Dictionary in that order).
The actual findings in percentages, however, show that the most
common (53% of the Birmingham corpus), examples are in the sense of ‘I
see’ and ‘you see’4. When coursebooks were examined these were found

4
Similarly, Brown (1994:61-79) examines the inter-relationships between the sense of a verb and the
various syntactic patterns in which it can be found and which are often absent in the Oxford Advanced
Learner’s Dictionary. Nevertheless, he (1994:77) regards understanding “the kinds of mechanisms that
can be employed in texts to convey more than is explicitly asserted” as essential for advanced students.

308
to account for only 10% of occurrences. Biber, Conrad and Reppen
(1998:80-82) describe a similar misrepresentation in ESL textbooks in
their representation of subject position that-clauses. Biber et al (ibid.77)
find that that-clauses in subject position are rare in all genres (only 5-10
occurrences per million words) but that these are virtually non-existent
in the spoken corpus they examined. One of the ESL textbooks examined,
however, had two exercises for the students to use subject position that-
clauses orally. Biber et al. (ibid. 81) conclude that the results from corpus
analyses could improve textbooks in two ways;

“First, books could emphasize those constructions most commonly found in

the target register. Students typically study English for particular purposes -
for example, conversational English for fluency in everyday interactions, or
academic English for proficiency in educational reading and writing tasks.
Textbooks can build on these goals to teach the grammatical constructions
that students are most likely to encounter, given their communicative goals.”

Added to this phenomenon is the fact that many coursebooks

encourage students to view different forms of a word in groups suggesting
that there is some kind of affinity between them but research suggests
that there is much more reason to see these as different. Sinclair and
Renouf (1987) say:

“From a lexical point of view, it is not always desirable to imply that there is
an identity between the forms of a word. ...But often, particularly with the
commoner words of the language, the individual word forms are so different
from each other in their primary meanings and central patterns of behaviour
(including the pragmatic and stylistic dimensions), that they are essentially
different ‘words’, and really warrant separate treatment on a language course.”

Field Code Changed

309
An example of this that Sinclair gives is the lemma move. He says
(1994:20), “The forms ‘moving’ and ‘moved’ share some meanings with
‘move’, but each form has a very distinctive pattern of meaning. Some of
the meanings found elsewhere in the lemma will be realised, and some
will not. In the word ‘moving’, for example, there is the meaning of
emotional affection which is quite prominent.” Similarly, Sinclair
(1997:32-38), drawing on the more recent Bank of English, gives
examples of not only the most usual meanings and uses of language but
also the most usual collocations. He shows that ‘nice’ although a very
neutral adjective has very strong patterns of language associated with it.
Sinclair finds that nice

“selects the indefinite article a and most emphatically rejects the definite
article the. When in predicative position, it attracts strongly a modifier such
as very, pretty, extremely. When attributive, it is commonly found with
another adjective with which it combines in meaning, so that a nice relaxing
time is nice because it is relaxing. Where nice immediately precedes a noun,
and has no modifier itself, the nouns it goes with seem to be frequently
selected from a few short lists – day, evening, etc., boys, girls, etc., and
surprise. Often there are set phrases.

Sinclair (ibid.) also argues that teachers should concentrate on

teaching meaning and suggest that if this is done, then it is obvious that
often one meaning is associated with one structure whilst another
associated with a different one. He claims that the more concrete, narrow
meaning goes with the noun and the more figurative, vaguer meaning
goes with the verb. He gives the example of combat which as a noun
means actual physical fighting but as a verb means ‘struggle against’
usually with abstractions like inflation, recession etc. Sinclair suggests
that these distinctions are easy to appreciate when they are pointed out,

310
but they are not always distinguishable grammatically. Hopper (1997:93)
argues that there is little terminology that the modern linguist uses that
would have been unfamiliar to Quintilian and that, because of this, some
integral parts of the language which, as Stubbs (1993:17) says, “lie
somewhere between word and group … are missed both by current
grammatical descriptions and also by conventional definitions of
collocation”. Hopper (ibid.) suggests that this situation is also the case
with the English verbal expression. He (1997:94) uses Firth’s sentence
“She kept on popping in and out of the office all the afternoon” as an
example of the difficulty of identifying the verb in such sentences. He
(1997:99) concludes that corpus linguistics is showing that the “category
of Verb itself might be more in the nature of a cluster or family-
resemblance category rather than a simple word class” or “folk category”.
He (1997:101) recommends the use of discourse as a data source so that
this can be made evident.
According to Sinclair (1997:37) rather than making the language
limited, the fact that regular linking of grammar or form and meaning will
not only cut down on the load the learner has to cope with but it will
make the curriculum more interesting and will allow the learner to
‘develop unique and personal utterances which are almost guaranteed to
be acceptable’. The example that he gives here is the structure
‘a(n) X of Y’
where X can be measures such as pint, yard, ounce, etc.; informal
portions blob, dash, lump, shred, etc.; shapes shaft, stick, tuft, etc.; flows
of liquid dribble, jet, spurt, etc.; containers bag, bucket, tank, tub, etc.;
formal collectives herd, flock, team, etc.; and informal collectives bunch,
clump, group, etc. If -ful is added to some things which are not normally
seen as containers such as bag to become bagful then almost anything
can become a container - a skirtful, a houseful, a shipful, etc. Sinclair
argues that this is what language is like and therefore, this is what
Field Code Changed
311
should be taught. He provides the following checklist for the language
teacher:

Present real examples only

Know your intuition
Inspect contexts
Teach by meaning
Highlight productivity

Intuition, as Sinclair uses it here, means for teachers (or learners) to be

able to give the meaning of words in isolation and to pronounce upon the
well-formedness of sentences in isolation.
He claims (1997:38) that corpora will “clarify, give priorities, reduce
exceptions and liberate the creative spirit.” Biber, Conrad and Reppen
(1998) have also reached the same conclusion in their studies of
synonymous words and structures. They find that words have preferred
collocates and senses and that structures are also used in certain ways
in certain registers. They conclude that students should be made aware
of these differences and taught accordingly rather than that there are
synonyms and synonymous structures that can be used indiscriminately.
McCarthy and Carter (1994:38) argue for a discourse view of
language and suggest that this involves “examining how bits of language
contribute to the making of complete texts….” exploring “the relationship
between the linguistic patterns of complete texts and the social contexts
in which they function ….”, considering “the higher-order operations of
language at the interface of cultural and ideological meanings and
returning to the lower-order forms of language which are often crucial to
the patterning of such meanings.”
Kjellmer (1991) suggests that

312
 “lexical items should not be taught and learnt in isolation but only in their
proper contexts. This means shifting the emphasis from individual words to
the collocations in which they normally occur.
It is only when the student has acquired a good command of a very
considerable number of collocations that the creative element can be relied
on to produce phrases that are acceptable and natural to the native
speakers.”

Mindt (1997:40-50) describes the experiences teachers in Germany

have had with using corpora for teaching in Germany. He (1997:41)
questions the content of the grammatical syllabus and suggests that
tradition is not sufficient reason for including grammatical knowledge. He
believes that research is the only reliable source of information for what
should constitute the grammatical syllabus as there is no comprehensive
grammar for the teaching of English as a foreign language. He argues
that first a corpus must be compiled and then a didactic grammar should
be constructed from the corpus; finally a pedagogical grammar should be
produced from the didactic grammar. He and Tesch have been
conducting research to try to produce a grammar for their students in
Germany. One of the first areas studied was the teaching of any. Far from
being concerned with the rules as given in the books they use in schools
where ‘any’ is contrasted with ‘some’ and rules like “Some is generally
used in affirmative sentences, ‘any’ in questions and negations” are given.
They find three types of any as in:

Any 1: I thought any fool would know

Any 2: I shan’t get any scripts from the assistants before then
Any 3: But is there any truth in it?

and suggest that these occur in the following situations and

frequencies:

Any 1 generally occurs in affirmative and declarative sentences and applies

to a referent whose existence is presupposed. Type 1 makes up more than
50% of all cases of any.
Field Code Changed
313
 Any 2 occurs in negative and declarative sentences and applies to the
referent whose existence is not presupposed. This type covers between 30
and 40 per cent of all instances in authentic texts.
Any 3 occurs in affirmative and interrogative sentences and applies to a
referent whose existence is not presupposed. This type makes up about 10
per cent of all cases of any.

Mindt (1997:44) points out that although ‘any 1’ as defined above is the
most frequent form of ‘any’ this is rarely mentioned in teaching materials
and is rarely mentioned in grammars of contemporary English. However,
he (ibid.) notes that in the English textbooks he examined this usage was
present in the same frequency but it was never explicitly taught in any of
the exercises on ‘any’ which restricted the teaching to types 2 and 3.
Based on these findings Tesch (1990:345f) proposes a new approach
in the teaching of some and any. The grading she suggests is not
assumed to take place within one lesson but would normally spread over
several teaching units but would include the use of the ‘missing’ meaning
of any which is the most frequent and where:

The traditional opposition of some and any, which is normally introduced

as the first distinction and at the very beginning, only occurs in step 5 (for
fast learners it would be possible to combine steps 4 and 5). The new
grading emphasises the main uses of some and any contrasting them step
by step with their appropriate counterparts in a new and unconventional
way which is consistent with the use of some and any by native speakers.

Similarly, in a large number of grammars will and would are

treated within the same framework of reference. Would is generally
considered as the past tense form of will. Mindt (1997:47) makes a
distinction between temporal meaning and modal meaning. The modal
meanings of will and would can be divided into five principal meanings,
three of which make up 97% of all cases of will and four of them 95% of

314
all cases of would: certainty/prediction, volition/intention, possibility/
high probability, hypothetical event or result, and habit.
The results were as follows:

will would
certainty/prediction: 71% 31%
volition/intention: 16%
possibility/ high probability: 10% 33%
hypothetical event or result: 18%
habit: 13%

Mindt argues from these results that because of their different semantic
profiles will and would should be treated separately in teaching materials.
Similar work has been carried out in Portugal by Prof. Casanova
from the University of Lisbon who argues (1995:100) that most English
grammars (and therefore language teachers) give inaccurate explanations
of English grammar which makes them inadequate or unusable. In the
case of the present perfect, the emphasis that is normally found in
grammars is on an incomplete action which was started in the past
represented by the verb tense but as Prof. Casanova shows this is simply
not correct and causes many exceptions to need to be cited. One of the
examples Prof. Casanova uses to demonstrate the inadequacy of this
explanation is the difference between John has lived in Paris and John
has lived in Paris for ten years. In these cases it is the adverb of time that
indicates that the action is incomplete rather than the verb tense. In the
former case he no longer lives in Paris yet in the latter he does which is
expressed by for ten years rather than the verb tense.
Mindt (1997:46) suggests that his research work emphasises the
importance of distributional data in grammars for teaching purposes.
Without distributional data there can be no informed grading of the
functions of a grammatical form in a language course. The absence of
distributional data in almost all preceding grammars results in a grading
Field Code Changed
315
that is based on intuition rather than on empirical evidence and very
often does not reflect the actual use of English. Halliday (1993:1) argues
that it is only with the development of the modern corpus that “serious
quantitative work in the field of grammar” can take place, the results of
which can show the probabilities of one grammatical pattern occurring
rather than another. The results that are obtained from such quantitative
research, Halliday (1993:6) suggests, are important for “learning and
teaching languages”. Through his work on the COBUILD corpus, Halliday
(1993:20-21) argues that positive and negative occur in English on a ratio
of 9:1 and that the 25 most frequent forms mostly occur only as verbs
whereas in the next 25 a large number of the forms function as both
noun and as verb. Francis (1991:145), working on the same University of
Birmingham COBUILD corpora, finds that “different senses of a noun
display different grammatical behaviour”. Todaka (1996:13) working from
the UCLA Oral Corpus and the Brown University Corpus finds that the
difference in usage of between and among can better be explained by
regarding their difference as a “distinction between ‘individual’ and
‘collective’”, that is, if the items in the NP objects are seen individually,
between is used, if not among is used. Added to this, the sentence
construction most often used with between is between A+B+(C...) whereas
that with among is most often among plural noun. He notes however that
when either of these could be used the preference for one or the other
depended upon the discourse register (formality) and the prescriptive
rules. He (1996:13) suggests that learners of English can apply his
findings to “everyday uses of these prepositions”. Despite all these
studies, there is, as yet, no work that is available for either teachers or
learners that describes English language usage comprehensively.
Minugh (1997) argues that, whenever school grammars use the
words ‘usually’ or ‘often’, students should be encouraged to go to a
corpus and examine a series of instances. In this way, he says they could

316
gain insight into the fact that the rules in school grammars are
‘necessarily overly simplistic and categorical’.
Johns (1997:102) says that working with data leads to not only “a
radical revision of preconceived ideas about what one should be teaching”
but also “how one might teach it.”
a) The simple principle ‘It is probably not worth teaching anything that
does not occur at least x times in a corpus of y million words’ (x and y
being redefinable taking into consideration the level of the learners) makes
it possible to exclude immediately much that is traditionally enshrined in
classroom tradition.
b) Pari passu the work suggests ways of dealing with areas of language
which have traditionally been poorly taught or regarded as unteachable
(e.g. article usage) and reveals areas of language structure (e.g. the
contextual patterning of nouns) that have been neglected both descriptively
and pedagogically.
c) The data controls not only which features of the language are taught, but
which exponents are presented and which meanings are taken as primary
(e.g. in Academic English, may, showing an estimate of probability based
on ‘experience’).
d) More fundamentally, the traditional division between independent
‘levels’ of language (e.g. lexis-syntax-discourse) appears increasingly
untenable once one starts to place at the centre of one’s concern the ways in
which words behave in context. As a result, although the materials have for
the most part a syntactic/functional starting point they could (as the
students themselves have observed) as well be labelled ‘Remedial
Vocabulary’ as ‘Remedial Grammar’.
Work done by Phillips (1985) and later by Hoey (1991) suggest that
discourse can also be explained better by means of lexical phrases.
Nattinger and DeCarrico (1992) say “Lexical phrases are parts of
language that have clearly defined roles in guiding the overall discourse.
In particular, they are the primary markers which signal the direction of
Field Code Changed
317
discourse, whether spoken or written.” Although corpus-based research
can aid teachers to see what is natural language use, they must be
careful to bear in mind not only the date of the corpus but also to make a
clear distinction between spoken and written language. Much of the work
done has shown a contrast between the two and has even gone as far as
noting differences between different age groups. There are dictionaries
based on computer corpora which clearly demonstrate the most frequent
meanings and collocations of words together with an explanation of
differences between spoken and (general) written language use
(Longmans, Collins, Cambridge etc.). The Longman’s Dictionary of
Contemporary English (LDOCE) claims to have 25,000 fixed phrases and
collocations. The editors say:

“The English Language is made up of building blocks or chunks of words.

When we produce a phrase, for example “if you don’t mind me saying so”,
we don’t think about each individual word and then link them together: we
automatically think of the phrase as one block of words working together,
one chunk of language. Through extensive corpus analysis, we have
identified these chunks of English, these fixed phrases. So now students
too can have access to these phrases...”

There are a number of books published giving collocations and

examples of use of such items as prepositions, phrasal verbs and other
structures through concordance samples which can be used to generate
exercises. Kennedy (1989) shows that two generalised frames for
prepositional phrases with at ‘at + (the) + Proper N denoting place’ and ‘at
+ Personal Pronoun’ account for 63% of the occurrences of this
preposition. Similarly, the corpora used here could form the basis of
exercises for inclusion in the materials for undergraduates. Goethals,
Engels and Leenders (1990:231-268) have even developed an automatic
exercise generator based on electronic texts, as has Wilson (1997). Wilson
318
produced exercises on pronouns, participles and appropriate words in
context which are somewhat simplistic as yet because of the difficulty of
getting the computer to perform more complex tasks but which,
nevertheless, demonstrate a trend that would considerably simplify the
work of the teacher wishing to develop data-driven learning exercises in
the future.
Most personal computers can be used to examine text and a simple
concordance can be bought or indeed written for use on an average
computer. An example of a KWIC concordance teachers can write
themselves is given in Tribble and Jones (1990:84-89) Concordances in
the Classroom. Others are available on-line (from ICAME and even OUP)
so that they can be downloaded by teachers for their own use. They also
give some very useful guidance on the use of concordances including
using them to analyse the students’ own work in order to bring out the
contrasts between natural language use and student use. Work of this
kind could not only personalise student correction but would foster more
learner autonomy. McCarthy and Carter (1994) give examples of
classroom activities contrasting discourse features in texts including
students’ own work in order to raise the students’ awareness and help
student improvement. In terms of correction, just using a spelling
checker with a word processor helps to highlight the student’s individual
difficulties.
A number of CD-ROM encyclopaedias have search devices which
will allow research into particular uses of words or alternatively texts can
be saved and then a simple concordance can be used. This has the
advantage of identifying very specific language use in specific subject
areas for more advanced language study. That is to say, these
encyclopaedia help not only to identify the frequency of a word but also
the range of the word, that is, how widespread the use of the word is in a

Field Code Changed

319
more general domain. The most commonly used language can be found
and also that language which is used in different registers.

7.5 Methodological Implications

What are the implications of the research in terms of methodology,

once the forms and meanings that are to be presented to the students
have been identified? First of all the course materials to be used must be
designed to ensure that they do not mislead the students about words
and structures and their meanings and uses but do reflect natural
language use, as in the case of the most common meanings and usage as
discussed above.
The work done on language acquisition suggests that the pattern
practice exercise is necessary and useful at an early stage in learning and
that later substitution exercises should be used to reflect the developing
awareness and adaptability of certain language patterns and the use of
more natural chunks of language to replace certain words or expressions
(Sinclair 1997). Frameworks for paragraphs have shown themselves to
work and students can indeed improve if they are given exercises of the
type which include cohesive devices which the students then use as a
basis for their own work. An example of this is where students are given
information which they have to reorder into a framework such as; There
are several problems such as (1).........and (2)........ However, (3)......
Therefore, (4)........ In addition, (5)....... Finally, (6)... . An example is given
initially and then a similar paragraph has to be produced based on new
data. The research mentioned earlier rather than suggesting that these
techniques are ‘old hat’, encourages their use as reflecting normal
language acquisition. Model opening and closing paragraphs for different
types of text could also usefully be presented and analysed at a more
advanced level. Indeed these features of text must be analysed to ensure

320
that students are aware of the discourse features that are associated with
them, which may include a transition of the tense used as Biber et al.
(1998:128) have shown in their study on research articles. McCarthy and
Carter (1994:58) suggest that students might be encouraged to produce
text frames which map the article or text being studied.
Hoey’s work (1991) demonstrates that students should be taught to
recognise cohesive devices in order to understand texts and that in order
to write more natural texts in English they should be aware of and use
different forms of repetition. Nattinger and DeCarrico (1992:60) say that
lexical phrases “signal the direction of discourse” whether the
information to follow is in contrast to, is in addition to, or is an example
of information that has preceded and, therefore, students should
recognise and practise this. An obvious way of doing this is through the
use of Cloze exercises which highlight the fact that only certain words are
possible and reflect the limited nature of most of the language that native
speakers use (with the exception of poetry and other forms of imaginative
creative writing which deliberately extends or breaks the rules). Pronoun
referencing and deictic features of text are very specific to academic prose
as Biber et al. (1998) and McCarthy and Carter (1994) show and their
specific use should be studied once again by taking actual examples from
the corpora and getting students to work on them in a number of ways.
Sinclair and Renouf (1987) suggest that “the main focus of study
should be on:

a) the commonest word forms in the language

b) their central patterns of usage
c) the combinations which they typically form”

All these are available through computer corpora but even Sinclair
and Renouf allow that the use of a grammatical table “may improve the
Field Code Changed
321
learning process” by shedding “light from a different angle” and support
an ‘eclectic’ position.
Three final principles that can be deduced from the research
described are:

1. Building on what students already know

2. Extending that knowledge further and
3. Remembering to ensure that there is adequate reinforcement of
language through recurrence of words, phrases and frameworks.

This first idea has also been backed up by psycholinguistic research into
the manner in which learners remember vocabulary. It is suggested that
schema are used and that these schema or word groupings are referred
to in order to enlarge upon and refine understanding. The second
principle is that recycling of items can lead the student to extend the
range of the word and gain insight into its use and facets, thereby
refining the meaning of the word in specific contexts. The third principle,
that is, repetition of items, is something which coursebook writers often
fail to do or fail to do consistently and which teachers must make an
effort to remedy. The better a teacher knows the materials that are being
used on the course, the easier this is. Students often take the stance that
work done in an earlier part of a course is no longer relevant later in the
course. This may be a response which has been produced from school
activities which divide up the material to be taught into convenient
sections which are then tested (and forgotten?) and not referred to
directly again later in the course. The detrimental effect that testing can
have on teaching leading to ‘teaching for the test’ has to be avoided
especially at this tertiary level where the students are learning ever more
detailed information in fewer subject areas and so cannot afford to ‘forget’
the earlier concepts on which the more specialised work is based. This

322
position cannot be applied to language learning either because the
process is clearly cumulative.
Finally, Portuguese corpora are being produced and when these are
available there should be an even more valuable tool to help teaching.
Bahns (1993:56) describes work carried out on lexical collocations
between German and English. Through contrastive analysis of “tens of
thousands” of lexical collocations the students are helped to identify
equivalent phrases and observe where differences occur so that they can
avoid errors in English. There is a need to reduce the learning load for
students through analysing and isolating the differences and similarities
between the two languages so that the students can be helped to produce
natural language and to avoid specific types of errors. It might also be
possible to find parallel texts (from European sources) in both English
and Portuguese which would be useful for examining differences and
similarities in scientific and technological discourse. Leech (1997:21)
describes such texts being developed through the C.R.A.T.E.R. (McEnery
et al 1994) and Multext (Ide and Vérons 1994) research projects. He
suggests that the fact that these texts are often highly specialised and
technical is a drawback but this may be a positive aspect for our
undergraduate students.
Using parallel texts would also accept the fact that often students
would be producing some kind of summary or translation from English
into Portuguese for their own use. Halliday (1993:125) suggests that
when texts are translated the translator does not normally alter the
discourse structure of the text that is being translated so this can help
the students to analyse scientific and technological discourses. Carter
(1993:146) argues that this kind of contrastive analysis can help to
produce awareness of socio-cultural meaning which is an extremely
important need if the textbooks continue to be based on the American
models as studied here. He (1993:147) goes further than this however,
Field Code Changed
323
and suggests that greater language awareness of this kind increases
learner autonomy and gives learners greater control over their learning,
which, for university undergraduates, is an essential part of the
educational process. Aitcheson (1994:95) suggests that understanding
words is “not just a case of sorting out the meanings of individual lexical
items” but that, to understand something fully “involves understanding
the mental models of a culture.” Adams, Heaton and Howarth (1991:11)
suggest that understanding “how cross cultural problems arise can help
the course designer, the teacher and the student to make reasoned
choices at the rhetorical and stylistic levels.” It has been argued that
recognising the different meanings of technical words in scientific
discourse is one of the basic skills that the reader needs in order to
understand that discourse but it is obvious from the research mentioned
above this is a somewhat simplistic view and the respective culture that
underlies the text must also be taken into account. Brumfit (1994:32)
suggests that the emphasis must be on knowledge as a process rather
than as static information and that it is essential for teachers to be
sensitive to the different understandings developed by particular cultural
and linguistic groups in order to be able to help students with their
individual needs.
Computer corpora can be used in at least three ways in teaching.
Fligelstone (1993:98) identifies these as:

• Teaching about (i.e. the principles and theories of corpus linguistics)

• Teaching to exploit (i.e. the practical aspects of corpus study)
• Exploiting to teach (i.e. deriving language-teaching materials from corpora)

There is a fourth activity however which results from teaching itself

Renouf (1997:256):

324
 • Teaching to establish resources (i.e. designing and creating the corpus)

The first of these principles will take some time to develop and might
more easily be used on mainstream language course with students who
have more time to focus on language itself. As with the debate about
teaching students the phonetic alphabet time is the principal constraint
in teaching about language. The second and third of these principles
would require the expertise of the lecturer working with the students on
the corpora and is feasible now that the corpora have been produced. The
fourth principle is something that could be applied if students had
specific areas of their studies which they felt needed addressing so that
the corpora could be built up or driven by what the students perceived
they needed to work on.
In conclusion, the data gathered through this study can provide
the examples of natural science and technology language of explanation
and exposition of the science textbook together with the means to present
actual language use of this medium to students on the English discipline
in the first year of university. Initially the corpora would be exploited for
teaching (and testing) resources but at a later stage, with adequate
resources available, could lead on to being exploited by the students
themselves on an individual basis to solve their individual language
problems or difficulties.

7.6 Modern Technology and the Syllabus

CALL - Computer Assisted Language Learning lends itself

particularly well to developing reading skills in English, and the use of
computer corpora, CD-ROM material, the Internet and e-mail should be
included in the English discipline for undergraduates. Johansson
(1991:305-6) believes that in the future teachers will be able to draw on
Field Code Changed
325
vast data sources to select material from, but that there is still a need for
smaller corpora because these can be “analysed exhaustively in a variety
of ways”. Computers have had and are having more and more
influence over life today and one of the features that more and more
students are coming across and enjoying is electronic mail. In Britain e-
mail is included in the National Curriculum for I.T. but it is an area that
would lend itself well to the EFL classroom. The students could use e-
mail as part of a tutoring system to help with individual needs and more
advanced students could be involved in projects on the corpora
themselves, analysing particular areas that they wish to find out more
about. Correspondence of this kind contains certain features which will
have to be taken into account in teaching . McCarthy and Carter (1994:5-
10) discuss how far spoken and written modes can be distinguished by
informants and go on to consider how ‘writerly’ a text is or the degree of
‘spokenness’ of a text together with the degree to which it is monologic or
dialogic with the reader. In other words these more modern genres are
often a mixture of modes, including features that would normally be
associated more formally with one or other of the spoken or written
modes. E-mail has a number of distinguishing features such as the fact
that it is a written dialogue including many of the features of speech and
the fact that the same messages are often repeated over and over again if
the dialogue continues for any length of time. One earlier taboo in
teaching was to present written dialogue as representing true
conversation but now it can be seen that a situation that goes back to the
very beginnings of foreign language teaching must once again be
returned to in order to help students to cope with the future.
These types of activities would also be a means of developing and
encouraging strategies for the students to carry over into the future.
Carter (1993:139) defends the idea that students should learn about the
language as it “is valuable in its own right and has ‘educative’ potential in

326
the broadest sense … it can enhance learning ‘through the language’
about the cultures and ideologies which inform the target language and
its uses.” There are some difficulties attached to activities like the use of
e-mail however, as Johansson (1991:307) points out; it is a medium
somewhere between speaking and writing and for this reason it is more
prone to error than more studied, revised writing. He (ibid.) suggests that
it is also more “playful and creative, less bound by conventions” so that it
would be a means for students to feel less inhibited in their use of written
language but it would not be a suitable means for encouraging accuracy
in language use.
The repetitive aspect of e-mail “conversations” through computers
also requires some reading skill techniques like scanning and skimming.
However these would be conducted in an even more interesting, on-
screen situation where the text scrolls up and down. The replies given
and further discussion of points raised have to be picked out from the
repeated material and signings on and off in the electronic conversations
that take place on chat pages.
Computers change the roles that normally pertain, in that the
reader may become the writer or editor of the text and can control the
amount and type of information that is displayed on the monitor (Gill and
Whedbee 1997:160). The Internet is another source of material which
students can access and which they could then edit for their own
purposes. Substantial editing is necessary especially if information
gained through e-mail or the Internet is to be incorporated in the
students’ own documents, and practice in doing this would be required.
With more of these activities taking place there would be a need to revise
the strategies used and to feed the insights gained through such
activities back into teaching materials, thus keeping flexibility and
openness to change a basic requirement of the syllabus.

Field Code Changed

327
 Electronic dictionaries are already available, including in
Portuguese, which would be useful tools for both teachers and learners to
test their hypotheses about general language use and perhaps through
special glossaries of specific terms for their subject requirements.
However, for specific language use the corpora produced for this analysis
should be made available so that the actual, immediate, authentic
situation can be examined and studied by the students themselves.
There are administrative considerations that have to be taken into
account for corpus-based studies to be carried out with students.
Appropriate computational infrastructure has to be provided first but
another consideration is the level of confidence that teachers feel with
computers so that they can be exploited fully and appropriately (Hughes
1997:292-307). Renouf (1997:255-266) describes her experiences with
teaching corpus linguistics to teachers of English (at post-graduate level).
She suggests that things have improved considerably over the last decade
in terms of the technology available for teaching classes and discusses
the need for and possibility of providing continued support through
distance contact. Her students were post graduates who after the course
returned to their own teaching environments abroad but through
computer communications they could keep in touch and obtain support
for their work. The Council of Europe attaches particular importance to
distance learning and to the use of new technologies for modern language
learning in Europe (Trim 1998:208). In the Appendix to Recommendation
Nº R (98) 6, adopted by the Committee of Ministers on 17 March 1998,
article 21 in Section E Adult Education agrees to:

21. Support the provision of national and international structures so as

to ensure the widest availability of facilities for distance education
(including the use of communication and information technologies), in
order to promote the development of diversified advanced

328
 communication skills, where possible linking autonomous learning to
institutionalised learning.

This agreement could be met through the methodology described

above for teaching undergraduates in Portuguese universities. Distance
education is certainly a possibility in the University of Aveiro which
already has considerable expertise in this area through the Department
of Didactics and Information Technology. The first year undergraduates
can also since 1998 available themselves of first year courses on-line. It
is entirely feasible that lecturers in all departments begin to ‘converse’
with their students on an individual basis through e-mail as the
university’s computer network already covers all departments. Increasing
expertise in this area would suggest that certain issues would present
themselves for remedial teaching. The FAQs (Frequently Asked
Questions) that those working in distance learning report (Laurillard
1993, Rowntree 1992) serve to alert the lecturer to common problems
which can then be dealt with generally rather than on an individual
basis. The only foreseeable disadvantage to this system is the number of
messages that could be received, at any time and that require time to
answer. Speed typing skills may become a necessity for tutors swamped
by student’s queries. Computer communication systems using e-mail
connecting students with individual mentors (scientists) have been tried
in some education systems (New York 1993) to aid success in education.
The building up of such relationships appears to be fruitful for both the
scientists involved and for the students. The scientists gain insights into
the student’s problems and the students feel comfortable enough to ask
for guidance on any number of problems, academic and personal. As with
the results from computer corpora which will change our view of how
language operates in fact and will cause new theoretical positions to be
developed, the whole nature of relationships between students and their Field Code Changed
329
lecturers will undergo profound changes because of communications by
computer. We must be prepared to meet these new situations in the near
future.

330
Chapter 8 Conclusion
 Chapter 8

Conclusion

My results show that although there is general agreement with

some of the dimensions studied by Biber (1988), the accepted attributes
of academic prose and the undergraduate textbooks studied here are by
no means congruent. I believe that these mismatches occur because of
the pedagogic nature of the texts and intention of the authors to interact
with and ‘teach’ or instruct the anticipated, native speaker,
undergraduate reader. The analogies with real-world applications used in
both the physics and chemistry textbooks is where this mismatch is most
often apparent. These real-world analogies appear both in essays and in
the problems provided at the end of chapters for the students to use to
practice the scientific topic presented in each chapter. American English
speech is also found to be replete with sports analogies from basketball,
baseball and American football like those found in the problems for
students to solve at the end of the chapters, so these textbooks are in
this sense merely reflecting the culture from which they come (Rosenthal
1996:105).
I would argue that many of the variations identified could be
attributable to the mixture of the scientific and the technological found in
the textbooks with their specific and different language usage. I call into
question some of the previous studies and corpora for their lack of rigour
which meant that they leant more towards a popular or journalistic
representation of science and technology rather than an academic
science and technology setting. Furthermore, the problems and real-
world analogies in the textbooks involve much more culturally-specific Field Code Changed
331
information which is more difficult for undergraduates, who are
simultaneously foreign language learners, than general scientific writing.
The level of English shown by the undergraduates studied here
varied, from very competent to quite weak, although there was more
evidence that recently undergraduates had generally studied English for
more years than in the first years of the foundation year course. The
stronger students showed that their reading comprehension skills had
also developed sufficiently well to allow them to apply strategies to what
for them were new English language situations like that of a (popular)
scientific or technological text which they had not studied in language
classes in school. The weaker students showed that their reading
comprehension skills were very poor or non-existent. Further research is
necessary however to determine if the students who did well on the tests
could cope equally well with test material drawn specifically from the
corpora developed for this study, that is, from the materials on the
bibliographies. It is now practical for this to take place using the corpora
built up for this study. This test material should also include a
component focusing on the different combinations of typographics,
footnotes, formulae, numbers, equations, tables, diagrams and drawings
found in these kinds of textbooks to check comprehension and
interpretation of these features by students.
Further research into the lack of success of students would involve
two distinct parameters, the difficulties with English per se and the
difficulties with the language of science and scientific concepts, so that it
is possible to evaluate which of these is proving to cause most difficulty
for students so that a strategy could be developed for coping with it in the
discipline. Knowledge about the students’ competence in science and
technology on entering the university might be determined through
collaboration with other departments in the university. Looking at the
entry marks in specific science subjects would not necessarily be

332
sufficient to assess the students’ future success in their first year as
undergraduates. The number of different combinations of circumstances
that the students present on entering the university is vast and
knowledge about those competencies, seen as core competencies in
science, but which many of the students lack would also aid in targeting
the syllabus for these students.
The wide variety in levels between the strongest and the weakest
students suggests that new strategies must be found to cope with large
groups of such a heterogeneous nature. The suggestion that is put
forward here is that these new strategies should be based on materials
and evidence obtained from the frequency counts and variation studies
carried out on the undergraduate textbooks. The teachers would then
have appropriate and relevant materials, and good information, so that
the focus would be on the items that would be most useful to students.
The use of computers and corpus analysis in the discipline would allow
the students themselves to approach their individual problems with the
language of science and technology and would eventually allow self-
access and distance and continuous learning to take place by means of
the university computer network.
In addition, the opportunity to work with colleagues from the
departments which teach the first year undergraduates in an
interdisciplinary manner would help to reinforce the teaching at this level
and provide a coherent framework for students to appreciate the
relevance of the work being done in language classes. From this co-
operation it would be possible to develop projects where the English
language needed by students for their project work could be analysed
and formulated from the language classes. This interdisciplinarity would
also serve to motivate those students who find it difficult to perceive the
relevance of their language studies to their courses. In other words,
content-based EFL will provide a focus and goal for the students.
Field Code Changed
333
 The testing of the students would then have to change. The present
system does not take into account the target language of science and
technology and so the corpora produced here should be exploited for
testing of students, both at the preliminary stage and for the normal
university evaluation tests throughout the year. In this way, it would be
possible to see if the students who were released from the discipline did
indeed cope with actual language from the corpora of undergraduate
textbooks rather than that perceived by their teachers to be relevant.
Added to this, through the use of tests available through computers, it
would be perfectly possible to design a novel testing procedure which
could in itself be more flexible, allowing students to attempt certain tests
when they felt that they were ready. Incidentally, it should not prove too
difficult to improve the speed of marking and feedback to students by
having a computer-marking system, releasing the teacher for other
valuable activities.
The possibility of developing further teaching materials through
analyses developed by students themselves from their own interests and
needs is feasible provided that the necessary resources are available.
These would comprise not only up-to-date computers with network
connections but also teaching staff who are confident with both the
technology and corpus-based techniques. This latter knowledge would to
a great extent avoid the complaint that language lecturers do not feel
confident with the subject matter of the materials they are trying to use
with students of science and technology as their focus would be entirely
on the evidence presented from the corpora in a linguistic analysis. In
other words teachers would be focusing on the language and not on
science and technology per se.
The role of the lecturer would also undergo a change in the type of
contact and interaction with students. The use of e-mail would allow a
much closer one-to-one interaction between student and tutor and might

334
develop a different relationship from that enjoyed in a large group of
students meeting for a limited amount of time. The use of e-mail itself
would be a means of moving forwards into the modern world of
communications and language use itself, although the emphasis would
be on individual support from the teacher for students. Experience from
other universities (Motteram, University of Manchester 1998) who have a
highly developed system of tutoring through e-mail would suggest that
tutors would eventually develop a series of frequently asked questions
(FAQs) which could be made available for students to consult and
thereby save some of the tutors’ time answering the same questions over
and over again. Similarly, support material could be provided on-line for
students to work on their own.
More and more corpora are becoming available on-line and on CD-
ROM and with a small investment of time and money many other
resources could be exploited in the language class. As was mentioned at
the beginning of this thesis (see 1.5 The Situation in Portugal), the
undergraduates in the first year are on numerous, different engineering
or degree courses and the use of different corpora in this way would allow
for a diversity of interests which might only become apparent at a later
stage in the students’ courses. Use of the European Union terminological
database EURODICAUTOM on-line would be one means of addressing
the diverse engineering needs in the students who should be encouraged
to focus on precision in language for the communication of scientific and
technical data and perhaps even to make or perfect comparable
terminology in Portuguese where this is lacking. Allowing for this kind of
subject flexibility would also lay the groundwork for the students to take
up a means of continuing their language studies beyond the end of the
first year and adapting the materials they use to their actual needs.
What is missing from this work is a comprehensive analysis of the
use of lectures and other spoken communication for students in the first
Field Code Changed
335
year (and subsequent years) of their courses. The spoken corpus would
also vary widely after the first year and would require other multimedia
resources. Video in particular should be exploited more to present and
practice listening comprehension and note-taking. The actual kinds of
lectures (or spoken communication such as papers at conferences) that
the students could be expected to come into contact with should also be
gathered into a database of materials for both self-access and class use.
Interdisciplinary work with the other departments could allow videoing of
actual lectures or parts of lectures delivered in the university in English.
These lectures, or excerpts from lectures, could then form the basis for
language study materials. There are examples of university lectures in
science and technology available through the Internet from some
American universities. The reason that these lectures are available on-
line is for the students on those courses to study from and then to
contact the tutor via e-mail with any queries and to deliver their
assignments. A similar system could be experimented with in the way
described above.
In conclusion, the corpora produced here could be exploited for use
with students in the first year in collaboration with other disciplines to
focus more closely on those areas identified by colleagues in other
disciplines to be central to the first year students’ needs. Further corpora
are needed to include spoken language from science and technology. The
testing of the students also needs to take into account the competencies
the first year students require to cope with English science and
technology texts. The use of information technology needs to be
reinforced to provide the students with more resources, support and
individual contact with their tutors, as well as to prepare the students for
their future professional lives and as life-long learners.
Many of the recommendations made here can be realised in the
short or medium term in this university with its sophisticated resources

336
and in other Portuguese universities which want to adopt common-core
courses and modern technology. What would need to be introduced to
continue the relevance and utility of the language taught/learned by
these undergraduates is to extend the English discipline into other years
of the courses. Language classes might be provided in parallel with
courses to be taken on an ad hoc basis as students saw fit. This would
necessitate a reappraisal of the language needs of the students at later
stages in their courses and the development of an appropriate syllabus,
methodology and materials. The suggestion (see 7.5 Methodological
Implications) that students themselves could be encouraged to provide the
materials that they need to work on which could then be turned into an
electronic corpus of materials which would form the basis for the
language studies carried out by the students would be relevant in this
case.
There is evidence (see Chapter 4) that in the first year,
approximately 10% of the new students (those with fewer than five years
of English studies at school) would benefit from more hours of study to
bring them up to the level of the other students and to make their science
studies through English a feasible proposition. Increasing the number of
hours devoted to English only for these students and including provision
for them to work extensively through self-access material would equip
them better for their future studies. Nevertheless, the results of the
research carried out in this thesis can form the basis for specific
materials for different language competencies in students by drawing on
parallel texts which, nevertheless demonstrate the relevant discourse
features displayed in the main corpora. In this way these students could
be brought closer to understanding the texts that they are encouraged to
consult through understanding of the characteristics of those text-types.
Steps are already being taken to exploit computer resources with
students and to provide on-line English texts for students to work on in a
Field Code Changed
337
variety of ways (including pronunciation of new vocabulary). Using
suitable materials in the language laboratory which accurately reflect the
students’ needs is also being attempted rather than continuing the
tradition of decontextualised drills and pronunciation work. The results
of this analysis has alerted lecturers to making their materials reflect the
target material for these undergraduates and to place emphasis on
interpretation of visual materials together with texts. All of these different
facets are being brought together into a syllabus which recognises that
much of the work has to be carried out outside the classroom by the
students on their own and gives weight to oral classroom interaction in
order to make the most of the contact time available. Constant
reappraisal of the syllabus has always been a feature of the discipline.
New insights into both the students’ competence and needs and new
research findings and the materials used, taking into account materials
that have worked successfully with students are fed back into the
syllabus for the first year students. The corpora will go further than this
however, as they will serve as a guide and object of study for the lecturers
themselves to use to inform their ideas and judgements of what scientific
English is and more importantly is not.

338
Bibliography
Bibliography

Abbot, G. (1980) Towards a more rigorous analysis of Foreign Language

Learners Errors IRAL Offprint Vol. XVIII/2 1980 Julius Groos Verlag,
Heidleberg

Adams, P., Heaton, B., Howarth, P. (eds. 1991) Socio Cultural Issues in
English for Academic Purposes, Macmillan, London.

Ahmad, K. et al. (1985) Computers, language learning and language

teaching. Cambridge: Cambridge University Press.

Aijmer, K. & Altenberg, B. (eds. 1991) English Corpus Linguistics, London,

Longman.

Aitcheson, J. (1994) “Understanding words” in Brown, G., Malmkjaer, K.,

Pollit, A., Williams, J. (eds.1994) Language and Understanding, Oxford:
Oxford University Press. Pp.83-95.

Alderson, J. C. (1988) “Testing and its Administration in ESP” in

Chamberlain, R. and Baumgardner, R. J. (eds. 1988) ESP in the
Classroom: Practice and Evaluation, ELT Documents 128 London: Modern
English Publications. Pp.87-97.

Alderson, C. and Urquhart, A. (eds. 1984) Reading in a Foreign Language,

London: Longman.

Allen, J.P.B. & Widdowson, H.G. (eds. 1974) English in Focus, Oxford,
Oxford University Press

Allen, J.P.B. & Widdowson, H.G. (1978) Teaching the Communicative Use of
English, in Mackay & Mountford (eds. 1978) English for Specific Purposes,
London, Longman.

Allwright, J. & Allwright, R. (1977) “An approach to the teaching of

Medical English” in Holden, S. (ed. 1977) English for Specific Purposes.
Oxford: Modern English Publications. Pp 58-62.

Armstrong, S. (ed. 1994) Using Large Corpora, London: MIT Press.

Arroteia, Jorge Carvalho; Martins, António Maria (1997) Inserção

Profissional do Diplomados pela Universidade de Aveiro: Trajectórias
Academicas e Profissionais, Aveiro: Universidade de Aveiro.

339
Astor, C., (ed. 1997) “Voices in Education” in Education in the United
States: Continuity and Change, U.S. Society & Values Electronic Journals of
the U.S. Information Agency, Vol. 2 Nº 5 December 1997 pp.37-39

Bahns, J. (1993) “Lexical collocations: a constrastive view”, ELT Journal,

Volume 47/1 January 1993: Pp 56-63.

Bakhtin, M. M. (1986) Speech Genres and Other Late Essays. Trans. Vern
McGee. Eds. C. Emerson and M. Holquist. Austin: University of Texas
Press.

Barber, C. L. (1962) Some measurable characteristics of modern scientific

prose in Swales, J. (1985), Episodes in ESP, Oxford: Pergamon Press Ltd.

Barros, A. M. (1998) Communication and Management Skills, European

Chemistry Thematic Network Report.

Bates, M. and Dudley-Evans, A. (1974) “Notes on the Introductory English

courses for Students of Science and Technology at the University of
Tabriz, Iran, ELT Documents (74/4), The British Council. Pp. 12-18.

Bazerman, C. (1998) “Emerging perspectives on the many dimensions of

scientific discourse” in Martin, J. R. and Veel, R. (1998) Reading Science,
London: Routledge. Pp15-28.

Beaugrande, R. de (1997) New Foundations for a Science of Text and

Discourse: Cognition, Communication, and the Freedom of Access to
Knowledge and Society, New Jersey: Ablex Publishing Corporation.

Beaugrande, R. de (1997) The Story of Discourse Analysis in Van Dijk, T.

A. (ed. 1997) Discourse as Structure and Process London: SAGE
Publications.

Biber, D. (1986) Spoken and written textual dimensions in English: resolving

the contradictory findings. Language 62:384-414

Biber, D. (1988) Variation across speech and writing Cambridge,

Cambridge University Press.

Biber D., Conrad S. and Reppen R. (1994) Corpus-based Approaches to

Issues in Applied Linguistics, Applied Linguistics, Vol. 15, No.2 Oxford
University Press

Biber D., Conrad S. and Reppen R. (1998) Corpus Linguistics. Investigating

Language Structure and Use, Cambridge: Cambridge University Press.

340
Birnbaum, I. (1987) “IT for Better Teachers” Educational Computing, Sept.
1987 Vol. 8, Issue 6. Pp. 19-21.

Bloor, M. & Bloor, T. (1991) Cultural Expectations and Socio-pragmatic

Failure in Academic Writing in Adams, P., Heaton, B., Howarth, P. (eds.
1991) Socio Cultural Issues in English for Academic Purposes, Macmillan,
London.

Bowen, J. D., Madsen, H., Hilferty, A. (1985) TESOL Techniques and

Procedures, Massachusetts: Newbury House Publishers.

Bracy, G. W. (1997) in Astor, C., (ed. 1997) “Voices in Education” in

Education in the United States: Continuity and Change, U.S. Society &
Values Electronic Journals of the U.S. Information Agency, Vol. 2 Nº 5
December 1997 pp.37-39

Bright, J. A. & McGregor, G. P. (1970), Teaching English as a Second

Language, London, Longman.

Brown, G., Malmkjaer, K., Pollit, A., Williams, J. (eds.1994) Language and
Understanding, Oxford: Oxford University Press.

Brown, K. (1994) “Syntactic clues to understanding” in Brown, G.,

Malmkjaer, K., Pollit, A., Williams, J. (eds.1994) Language and
Understanding, Oxford: Oxford University Press. Pp. 61-79.

Brown, R. 1973. A First Language: the Early Years. Cambridge, Mass.:

Harvard University Press.

Brumfit, C. (1994) “Understanding, language, and educational processes”

in Brown, G., Malmkjaer, K., Pollit, A., Williams, J. (eds.1994) Language
and Understanding, Oxford: Oxford University Press. Pp. 22-33.

Brumfit, C. J. and Johnson, K. (1979-81) The Communicative Approach to

Language Teaching, Oxford: Oxford University Press.

Bucy, J. F. (1985) “Computer Sector Profile” in Keatley, A. G. (ed. 1985)

Technological Frontiers and Foreign Relations, Pp. 46-78, Washington:
National Academy Press.

Burnard, L. (1992) “Tools and Techniques for Computer-assisted Text

Processing” in

Butler, C. S. (ed. 1992) Computers and Written Texts, Oxford: Basil

Blackwell.

341
Butler, C. (1985) Statistics in Linguistics, Oxford: Blackwell.

Butler, C. S. (ed.1992) Computers and Written Texts, Oxford: Basil

Blackwell.

Byram, M and Riagáin, P. O. (1999) “Towards a Framework for Language

Education Policies in Europe” Strasbourg: Council of Europe
DECS/EDU/LANG(99) 6rev.

Candlin, C. N., Kirkwood, J. M. and Moore, H.M. (1978) “Study Skills in

English: Theoretical Issues and Practical Problems” in Mackay, R. &
Mountford, A.J. (eds.1978), English for Specific Purposes, London,
Longman.

Carter, R. (1990) “Towards Discourse-Sensitive Cloze Procedures: the Role

of Lexis” in Halliday, M.A.K., Gibbons, J., Nicholas, H. (eds. 1990)
Learning Keeping and Using Language Volume II Selected Papers from the
8th World Congress of Applied Linguistics, Sydney, 16-21 August 1987.
Amsterdam/Philadelphia: John Benjamin’s Publishing. Pp. 445-453.

Carter, R. (1993) “Language Awareness and Language Learning” in Hoey,

M. (ed.1993), Data, Description, Discourse, London, Harper Collins. Pp.
139-150.

Carter, R., Goddard, A, Reah, D., Sanger, K. and Bowring, M. (1997)

Working with Texts : A core book for language analysis, London: Routledge.

Casanova, I. (1995) “Gramática Inglesa: Procura-se” in the Actas do XI

Encontro da Associação Portuguesa de Linguística Vol III Gramática e Varia,
Depósito Legal 102905/96, Lisboa, Setembro de 1996 Pp.97-103.

Chamberlain, R. and Baumgardner, R. J. (eds. 1988) ESP in the

Classroom: Practice and Evaluation, ELT Documents 128 London: Modern
English Publications.

Chang, R. (1991) Chemistry International Edition, Fourth Edition, McGraw

Hill Inc., USA.

Chattel, R. K. (1999) “ESP - the case of English for Students of Sociology

in the University of Coimbra” Actas do 4º Encontro Nacional do Ensino
das Línguas Vivas no Ensino Superior, Faculdade de Letras da
Universidade de Porto. Pp. 245- 254.

Clark, R. (1974) ‘Performing without competence’. Journal of Child

Language 1:1-10.

342
Clinton, W.J (1997) in Astor, C., (ed. 1997) “Voices in Education” in
Education in the United States: Continuity and Change, U.S. Society &
Values Electronic Journals of the U.S. Information Agency, Vol. 2 Nº 5
December 1997 pp.37-39

Clinton, W.J (1998) “Excerpt from President William J. Clinton’s State of

the Union Address”, in American Studies Journal, Nº 41, Summer 1998 Pp
59-60.

Close, R. A. (1965) The English We Use for Science, London: Longman

Connor, U. and Kaplan, R.B. (eds. 1987) Writing across Languages:

Analysis of L2 Text, Reading, MA.: Addison-Wesley.

Cooper, M. Aspects of the Structure of Written Academic Discourse and

Implications for the Design of Reading Programmes, in Hoedt, J. &
Lundquist, L., Picht, H. & Qvistgaard, J. (eds. 1982)) Proceedings of the
3rd European Symposium on LSP, Copenhagen August 1981, The
Copenhagen School of Economics. Pp.403-433.

Corder, S. P. (1973), Introducing Applied Linguistics, London: Penguin

Corder, S. P. & Roulet, E. (eds. 1974) Linguistic Insights in Applied

Linguistics, Second Neuchatel Colloquium in Applied Linguistics, Brussels:
AIMAV; Paris: Didier.

Coulthard, M. (1977) An Introduction to Discourse Analysis, London:

Longman.

Coulthard, M. (1978) Discourse Analysis in English - a short review in

Kinsella, V. (ed. 1974) Language Teaching and Linguistics: Surveys,
Cambridge: Cambridge University Press.

Coulthard, M. (ed. 1994 - 1998) Advances in Written Text Analysis,

London: Routledge.

Cruttenden, A. (1981) ‘Item-learning and system-learning’. Journal of

Psycholinguistic Research 10:79-88.

Crystal, D. (1997) English as a global language, Cambridge: Cambridge

University Press.

Crystal, D (1998) “To surf or not to surf: that is the question” in Network A
Journal for English Language Teacher Education Vol.1 Number 1 December
1998 The British Council.

343
Danesi, M. and Di Pietro, R. (1990) Contrastive Analysis for the
Contemporary Second Language Classroom, Ontario: Ontario Institute for
Studies in Education.

Darian, S. The Role of Definitions in Scientific and Technical Writing: Forms,

Functions, and Properties, in Hoedt, J. & Lundquist, L., Picht, H. &
Qvistgaard, J. (eds. 1982)) Proceedings of the 3rd European Symposium
on LSP, Copenhagen August 1981, The Copenhagen School of Economics.

Davies, A. (1988) Procedures in Language Test Validation in Testing

English for University Study, ELT Documents: 127, MEP 1988

Di Pietro, R.J. (1971) Language Structures in Contrast. Rowley, Mass.:

Newbury House.

Dunn, S. and Morgan, V. (1987) The Impact of the Computer on Education,

London: Prentice Hall International.

Dudley-Evans, T. (1994) “Genre analysis: an approach to text analysis for

ESP” in Coulthard, M. (ed. 1994- 1998) Advances in Written Text Analysis,
London: Routledge. Pp.219-228.

Eastment, D. (1987) Computer-Assisted Language Learning in the Bell

Educational Trust. Cambridge: Bell Educational Trust.

“Education and the wealth of nations” in The Economist March 29th 1997
Pp.15-16.

Eggins and Martin (1997) “Genres and Registers of Discourse” in van Dijk,
T. A., (ed. 1997) Discourse as Structure and Process, London: SAGE
Publications.

Ellis, G. & Sinclair, B. (1989), Learning to Learn English, Cambridge, CUP.

ELTDU (1970) English for Business: research and preliminary planning

report. Colchester: English Language Teaching Development Unit.

Evelyn Ng, K.L. and Olivier, W.P (1987) “Computer Assisted Language
Learning: An Investigation on some Design and Implementation Issues” in
System, Vol. 15, No. 1. Pp. 1-17.

Ewer, J. R. (1975) “Teaching English for Science and Technology: the

specialised training of teachers and programme organisers.” In English for
Academic Purposes: Information Guide No. 2. London: British Council.

344
Ewer, J. and Hughes-Davies, E. (1971-72) Further notes on developing an
English language programme for students of science and technology,
English Language Teaching Journal XXVI/1 and 3

Ewer, J. R. and Hughes Davies, E. (1974) “Instructional English”, ELT

Documents (74/4), The British Council. Pp. 19-24.

Ewer, J.R. & Latorre, G. (1969) A Course in basic scientific English,

London, Longman.

Fairclough, N. (ed. 1992) Critical Language Awareness, London: Longman.

Ferreira, A. M., Ramos, D., Braga da Silva, F. (1999) “Evaluation des

curricula de FLE au Portugal” Actas do 4º Encontro Nacional do Ensino
das Línguas Vivas no Ensino Superior, Faculadade de Letras da
Universidade de Porto. Pp.333- 337.

Fligelstone, S. (1993) “Some reflections on the question of teaching, from a

corpus linguistics perspective.” ICAME Journal 17:97-109.

Flowerdew, J. (1993) Concordancing as a tool in course design. System 21

(2): P. 213-29

Fordham, S. (1997) Portuglish: Mistakes made by Portuguese-speaking

learners of English, Lisbon: Plátano Editora.

Fox, J. Matthews, A., Matthews, C., Rope, A., (1990) Educational

Technology in Modern Language Learning in the secondary, tertiary and
vocational sectors, for the British Government Employment Department
Group Training Agency, Learning Technology Unit by the University of
East Anglia and the Bell Educational Trust, March 1990.
Sheffield: Crown Copyright.

Francis G. (1991) “Nominal group heads and clause structure” WORD

Journal of the International Linguistic Association, Volume 42, Number 2
August 1991 Pp.145-156.

Francis G. and Sinclair J. (1994)‘I Bet He Drinks Carling Black Label’: A

Riposte to Owen on Corpus Grammar, Applied Linguistics, Vol. 15, No.2
Oxford University Press Pp190-200.

Francis, W.N. and Kucera, H. (1982) Frequency analysis of English usage:

lexicon and grammar. Boston: Houghton Mifflin.

345
Friel, M. (1978) A verb frequency count in legal English, ESPEMA, 10,
Spring 1978

Gavioli, L (1997) “Exploring Texts through the Concordancer: Guiding the

Learner” in Wichmann, A., Fligelstone, S., McEnery, T., Knowles, G. (eds.
1997) Teaching and Language Corpora, London: Longman. Pp. 83-99.

Gill, A. M. and Whedbee, K. (1997) “Rhetoric” in van Dijk, T. A., (ed. 1997)
Discourse as Structure and Process, London: SAGE Publications.

Glaser, R. (1982) The Problem of Style Classification in LSP (ESP) in Hoedt,

J. & Lundquist, L., Picht, H. & Qvistgaard, J. (eds. 1982) Proceedings of
the 3rd European Symposium on LSP, Copenhagen August 1981, The
Copenhagen School of Economics. Pp. 69-82.

Goethals, M., Engels, L. K., Leenders, T. (1990) “Automated Analysis of the

Vocabulary of English Texts and Generation of Practice Materials” in
Halliday, M. A. K., Gibbons, J. & Nicholas, H. (eds. 1990) Learning Keeping
and Using Language, Vol. II P.231-268 Amsterdam: John Benjamins
Publishing Company. Pp. 231-268.

Goodale, M., Collins Cobuild Phrasal Verbs, Harper Collins 1993

Grabe, W. P. (1984) “Written Discourse Analysis.” Annual Review of

Applied Linguistics 5: 101-23.

Grabe, W. P. (1987) in Connor, U. and Kaplan, R.B. (eds. 1987) Writing

across Languages: Analysis of L2 Text, Reading, MA.: Addison-Wesley.
Pp.115-137

Granger, S. The Learner Corpus: a revolution in applied linguistics, in

English Today, 39 Vol.10 No.3 July 1994, Cambridge University Press. Pp.
25-32.

Green, G. M. (1996) Pragmatics and Natural Language Understanding, 2nd

Edition, New Jersey: Lawrence Erlbaum

Guillot, M-N. & Kenning M-M. (1995) Exploiting the Potential of CD-ROM
Databases: Staff Induction at the University of East Anglia in Computer
Assisted Language Learning 1995 Vol.8 No.4 pp. 365-381.

Hakuta, K. (1974). ‘Prefabricated patterns and the emergence of structure

in second language acquisition’. Language Learning 24:287-97.

346
Hakuta, K. (1976) ‘Becoming bilingual: a case study of a Japanese child
learning English’. Language Learning 26: 321-51.

Halliday, M.A.K. (1985) An introduction to functional grammar, London:

Edward Arnold.

Halliday, M.A.K. (1993) “Quantitative Studies and Probabilities in

Grammar” in Hoey, M. (ed.1993), Data, Description, Discourse, London,
Harper Collins. Pp.1-25.

Halliday, M.A.K. (1994) “The construction of knowledge and value in the

grammar of scientific discourse, with reference to Charles Darwin’s The
Origin of Species” in Coulthard, M. (ed. 1994 - 1998) Advances in Written
Text Analysis, London: Routledge. Pp. 136-156.

Halliday, M.A.K. (1998) “Things and relations: Regrammaticising

experience as technical knowledge” in Martin, J. R. and Veel, R. (1998)
Reading Science, London: Routledge. Pp185-235.

Halliday, M., McIntosh, A. & Strevens, P. (1964) The linguistic sciences and
language teaching. London: Longman.

Halliday, M. A. K. and Hasan, R. (1976) Cohesion in English, London:

Longman.

Halliday, M.A.K. and Martin, J.R. (1993) Writing Science. Literacy and
Discursive Power London: The Falmer Press.

Halliday, M.A.K., Gibbons, J., Nicholas, H. (eds. 1990) Learning Keeping

and Using Language Volume II Selected Papers from the 8th World
Congress of Applied Linguistics, Sydney, 16-21 August 1987.
Amsterdam/Philadelphia: John Benjamin’s Publishing.

Harris, R. and Talbot, J. T. (1989) Landmarks in Linguistic Thought: The

Western Tradition from Socrates to Saussure, London: Routledge.

Hatch, E. & Brown, C. (1995) Vocabulary, Semantics, and Language

Education, Cambridge: Cambridge University Press.

Heppel, S. (1987) “Trainers Reach Crisis Point” in Educational Computing,

Sept. 1987, Vol. 8, Issue 6.

Herbert, A.J. (1965) The Structure of Technical English, London, Longman

347
Heslot, J. (1982) Tense and other Indexical Markers in the Typology of
Scientific Texts in English, in Hoedt, J., Lundquist, L., Picht, H. &
Qvistgaard, J. (eds. 1982)) Proceedings of the 3rd European Symposium
on LSP, Copenhagen August 1981, The Copenhagen School of Economics.

Higgins, J. (1985) “Computers in English Language Teaching” ELT

Document 122, London: Pergamon Press.

Higgins, J. (1988) Language Learners and Computers, London: Longman.

Hill, A. A. (ed. 1969-78) Linguistics, Washington: Forum.

Hockey, S. and Ide, N. (eds. 1994) Research in Humanities Computing

Oxford: Clarenden Press.

Hoedt, J. & Turner, R. (eds. 1981) New Bearings in LSP, The Copenhagen
School of Economics.

Hoedt, J. & Lundquist, L., Picht, H. & Qvistgaard, J. (eds. 1982))

Proceedings of the 3rd European Symposium on LSP, Copenhagen August
1981, The Copenhagen School of Economics.

Hoey, M. (1991) Patterns of Lexis in Text, Oxford, OUP.

Hoey, M. (ed.1993), Data, Description, Discourse, London, Harper Collins.

Hoffman, L. (1981) The Linguistic Analysis and Teaching of LSP in the

German Democratic Republic, in Hoedt, J. & Turner, R. (eds. 1981) New
Bearings in LSP, The Copenhagen School of Economics. Pp. 107-130.

Hofland, K. & Johansson, S. (1982) Word Frequencies in British and

American English. Bergen: Norwegian Computing Centre for the
Humanities/London: Longman.

Holden, S. (ed. 1977) English for Specific Purposes. Oxford: Modern

English Publications.

Holliday, A. (1984) Research into classroom culture as necessary input

into syllabus design. In Swales, J. & Mustafa, H. (eds.1984) English for
Specific Purposes in the Arab World, 29-51. Birmingham: University of
Aston.

Holliday, A. & Cooke, T. (1982). An ecological approach to ESP. Lancaster

Practical Papers in English Language Education, 5 (Issues in ESP), 123-43.

348
Hopper, P. J. (1997) “Discourse and the category ‘Verb’ in English” in
Language and Communication, Vol. 17, Nº 2: Pergamon. Pp.93-102.

Howcroft, S.J. (1989) “CALL in Coimbra” MUESLI News Micro Users in

English as a Second Language Institutions, The I.A.T.E.F.L. Special
Interest Group for Computer Assisted Language Learning December 1989
Pp.6-7.

Howcroft, S.J. (1986) “English for Special Groups: CTT Computer

Personnel Course”, Report on the British Council Conference on Computer
Assisted Language Learning, Lisbon 1986, Lisbon: The British Council.

Huang, J. (1971) A Chinese Child’s Acquisition of English Syntax.

Unpublished Master’s Thesis. University of California at Los Angeles
quoted in Nattinger and De Carrico (1992) Lexical Phrases and Language
Teaching Oxford: Oxford University Press.

Huddlestone, R. D. (1971) The Sentence in Written English, Cambridge:

Cambridge University Press.

Hughes, A. Achievement and Proficiency: The Missing Link? in Testing

English for University Study, ELT Documents: 127, MEP 1988

Hughes, G. (1997) “Developing a Computing Infrastructure for Corpus-

based Teaching” in Wichmann, A., Fligelstone, S., McEnery, T., Knowles,
G. (eds. 1997) Teaching and Language Corpora, London: Longman. Pp.
292-307.

Hunstan, S. (1994) “Evaluation and organization in a sample of written

academic discourse” in Coulthard, M. (ed.1994 - 1998) Advances in
Written Text Analysis” Pp.191-218 London:Routledge

Hutchinson, T. (1988) “Making Materials Work in the ESP Classroom” in

Chamberlain, R. and Baumgardner, R. J. (eds. 1988) ESP in the
Classroom: Practice and Evaluation, ELT Documents 128 London: Modern
English Publications. Pp. 71-75.

Hutchinson, T. & Waters, A. (1980). ESP at the crossroads. ESP

Newsletter, 36 (Reprinted in Swales (ed. 1985), Pp 177-85.)

Hutchinson, T. & Waters, A. (1984)”How communicative is ESP?” in the

ELT Journal Volume 38/2 April 1984 Pp.108-113.

Hutchinson, T. & Waters, A. (1987), English for Specific Purposes,

Cambridge: Cambridge University Press.

349
Hymes, D. (1971) On Communicative Competence. Philadelphia: University
of Pennsylvania Press.

Johansson, S. (ed. 1982) Computer Corpora in English Language Research.

Bergen: Norwegian Computing Centre for the Humanities.

Johansson, S. (1991) “Times change and so do corpora” in Aijmer, K. &

Altenberg, B. (eds. 1991) English Corpus Linguistics, London, Longman.

Johansson, S., Leech, G. N. and Goodluck, H. (1978) Manual of

information to accompany the Lancaster-Oslo/Bergen Corpus of British
English, for use with digital computers. Department of English, University
of Oslo.

Johns, A.M. (1986) Coherence and Academic Writing: some Definitions and
Suggestions for Teaching, in TESOL Quarterly Vol. 20 Nº2 1986 Pp247-
265).

Johns, T (1994-98) “The text and its message” in Coulthard, M. (ed. 1994 -
1998) Advances in Written Text Analysis, London: Routledge. Pp. 102-116.

Johns, T. (1997) “Contexts: the Background, Development and Trialling of

a Concordance-based CALL Program” in Wichmann, A., Fligelstone, S.,
McEnery, T., Knowles, G. (eds. 1997) Teaching and Language Corpora,
London: Longman. Pp.100-115.

Johns, T. (1989) “Whence and whither classroom concordancing?” in

Bongaerts, T. et al. (eds. 1989) Computer Applications in Language
Learning. The Netherlands: Foris Publications. Pp. 9-27.

Jones, C. (1991) An integrated model for ESP syllabus design. English for
specific Purposes,10, 3, Pp.155-72.

Jones, C. and Fortescue, S. (1987) Using Computers in the Language

Classroom, London: Longman.

Jordan, R. R. (1997) English for Academic Purposes: A guide and resource

book for teachers, Cambridge: Cambridge University Press.

Jordan, R. & Mackay, R. (1973) A survey of the spoken English problems

of overseas postgraduate students at the universities of Manchester and
Newcastle-upon-Tyne. Journal of the Institute of Education of the
Universities of Newcastle-upon-Tyne and Durham, 125.

350
Kachru, B.B. (1990) “World Englishes and Applied Linguistics” in Halliday,
M.A.K., Gibbons, J., Nicholas, H. (eds. 1990) Learning Keeping and Using
Language Volume II Selected Papers from the 8th World Congress of
Applied Linguistics, Sydney, 16-21 August 1987. Amsterdam/
Philadelphia: John Benjamin’s Publishing. Pp.203-229.

Kaplan, R. B. (1993) “Conquest of Paradise - Language Planning in New

Zealand" in Hoey, M. (ed. 1993) Data, Description, Discourse, London:
Harper Collins.

Kaplan, R. B. (1966) “Cultural thought patterns in inter-cultural

education. Language Learning 16, Pp 1-20.

Keatley, A. G. (ed. 1985) Technological Frontiers and Foreign Relations,

Washington: National Academy Press.

Kelly, L. G. (1969) 25 Centuries of Language Teaching, Rowley, Mass.:

Newbury House

Kennedy, C. (1983) “An ESP Approach to EFL/ESL Teacher Training” in

The ESP Journal Volume 2 1983 Pergamon Press Ltd. Pp.73-85.

Kennedy, C. & Bolitho, R. (1984) English for Specific Purposes, London,

Macmillan.

Kennedy, G. (1989) Collocations: where grammar and vocabulary teaching

meet. Paper presented at the RELC Seminar, Singapore.

Kenning, M. and Kenning, M.M. (1983) Introduction to computer-assisted

language teaching. Oxford: Oxford University Press.

Kenworthy, J. (1991) Language in Action, London: Longman

Kerridge, R. & Sammells, N. (eds. 1998) Writing the Environment:

Ecocriticism and Literature, London: Zed Books.

Kinsella, V. (ed. 1978) Language Teaching and Linguistics: Surveys,

Cambridge: Cambridge University Press.

Krashen, S.D. (1982) Principles and Practice in Second Language

Acquisition, Oxford: Pergamon.

Kress, G., Leite García, R. and van Leeuwen, T (1997) “Discourse

Semiotics” in van Dijk, T. A., (ed. 1997) Discourse as Structure and
Process, London: SAGE Publications.

351
Kroch, A.S. and Hindle, D.M. (1982) A Quantitative Study of the Syntax of
Speech and Writing, Final Report to the National Institute of Education.

Kubanek, A. (1998) “Primary foreign language teaching in Europe - trends

and issues” Language Teaching 31 October 1998 Pp.193-205, Cambridge
University Press.

Labov, W. (1972) Rules for Ritual Insults, in Sudnow, D. (ed. 1972) Studies
in Social Interaction, New York: The Free Press.

Lackstrom, J.E., Selinker, L. and Trimble, P. (1972) “Grammar and

Technical English” in English Teaching Forum, September-October 1972.
Pp. 3- 14.

Langkilde, C. Typical Aspects of the Reading Comprehension Barrier of

Undergraduate Language Students at the Copenhagen School of Economics.
in Hoedt, J. & Lundquist, L., Picht, H. & Qvistgaard, J. (eds. 1982))
Proceedings of the 3rd European Symposium on LSP, Copenhagen August
1981, The Copenhagen School of Economics. Pp.515-524.

Laurillard, D. (1993-97) Rethinking University Teaching: a framework for

the effective use of educational technology, London: Routledge.

Leech, G. (1993) 100 million words of English, English Today, 33 Vol.9

No.1 January 1993, Cambridge University Press. Pp 9-15

Leech, G. and Candlin, C. (eds. 1986) Computers in English Language

Teaching and Research, London: Longman.

Leech, G. (1997) “Teaching and Language Corpora: a Convergence” in

Wichmann, A., Fligelstone, S., McEnery, T., Knowles, G. (eds. 1997)
Teaching and Language Corpora, London: Longman. Pp. 1-23.

Leech, G. & Fligelstone, S. (1992) Computers and Corpus Analysis, in

Butler, C. S. (ed.1992) Computers and Written Texts, P.115-140 Oxford:
Basil Blackwell.

Leitner, G. (1993) “Where to Begin or Start?: Aspectual Verbs in

Dictionaries,” in Hoey, M. (ed. 1993) Data, Description, Discourse, London:
Harper Collins.

352
Lemke, J. L. (1990) “Technical Discourse and Technocratic Ideology” in
Halliday, M.A.K., Gibbons, J., Nicholas, H. (eds. 1990) Learning Keeping
and Using Language Volume II Selected Papers from the 8th World
Congress of Applied Linguistics, Sydney, 16-21 August 1987.
Amsterdam/Philadelphia: John Benjamin’s Publishing. Pp.435-460.

Lemke, J. L. (1998) “Multiplying meaning: Visual and verbal semiotics in

scientific text” in Martin, J. R. and Veel, R. (1998) Reading Science,
London: Routledge. Pp.87-113.

Luke, A. (ed.1992) Critical Perspectives on Literacy and Education, writing

in the Introduction of Halliday, M.A.K. and Martin, J.R. (1993) Writing
Science. Literacy and Discursive Power London: The Falmer Press.

Mackay, R. (1978) “Identifying the nature of learners’ needs,” in Mackay &

Mountford (eds. 1978), English for Specific Purposes, London, Longman.
Pp 21-42.

Mackay, R. & Mountford, A.J. (eds.1978), English for Specific Purposes,

London, Longman.

Mackay, R. & Palmer, J. (eds. 1981) Languages for Specific Purposes,

Rowley, MA: Newbury House

Martin, J. R. (1992) English Text: System and Structure, Philadelphia and

Amsterdam: Benjamins.

Martin, J. R. and Veel, R. (1998) Reading Science, London: Routledge.

Matthews, M. R. (1994) Science Teaching, London: Routledge.

McCarthy, M. (1990), Vocabulary, Oxford, Oxford University Press.

McCarthy, M. (1991), Discourse Analysis for Language Teachers,

Cambridge: Cambridge University Press.

McCarthy, M. & Carter, R. (1994) Language as Discourse. Perspectives for

Language Teaching, London: Longman.

Meijs, W. (1992) Computers and Dictionaries in Butler, C. S. (ed. 1992)

Computers and Written Texts, Oxford: Basil Blackwell.

Meyers, G. (1994-98) “Narratives of science and nature in popularizing

molecular genetics” in Coulthard, M. (ed. 1994 - 1998) Advances in Written
Text Analysis, London: Routledge. Pp.179-190.

353
Mettinger, A. (1994) Aspects of Semantic Opposition in English, Clarendon
Press Oxford, Oxford University Press.

Mindt, D. (1997) Corpora and the Teaching of English in Germany in

Wichmann, A., Fligelstone, S., McEnery, T., Knowles, G. (eds. 1997)
Teaching and Language Corpora, London: Longman. Pp. 40-50

Minugh, D. (1997) “All the Language that’s Fit to Print: Using British and
American Newspaper CD-ROMs as Corpora,” in Wichmann, A.,
Fligelstone, S., McEnery, T., Knowles, G. (eds. 1997) Teaching and
Language Corpora, P 67-82 London: Longman.

Moon, R. “The analysis of meaning” in Sinclair, J.M. (ed. 1987) Looking Up

An account of the COBUILD Project in lexical computing, London, Harper
Collins Publishers Pp. 86-103.

Moon, R. (1994) “The analysis of fixed expressions in text” in Coulthard,

M. (ed. 1994 - 1998) Advances in Written Text Analysis, London:
Routledge. Pp. 117-135.

Moreira, A.A. (1996) Desenvolvimento da Flexibilidade Cognitiva dos

Alunos-Futuros professores: uma experiência em Didáctica do Inglês,
Unpublished PhD Thesis, University of Aveiro.

Mountford, A. (1988) “Factors Influencing ESP Materials Production and

Use” in Chamberlain, R. and Baumgardner, R. J. (eds. 1988) ESP in the
Classroom: Practice and Evaluation, ELT Documents 128 London: Modern
English Publications. Pp.76-84.

Moura Carvalho, J. (1991) “The Minerva Project” in MUESLI News October

1991. Pp 3-7.

Munby, J. (1978) Communicative Syllabus Design, Cambridge, Cambridge

University Press.

Murray, D. E. (1988) “Computer-mediated communication: Implications

for ESP”, English for Specific Purposes, vol.7,1, Pp. 3-18.

Myers, G. (1998) “Narratives of science and nature in popularizing

molecular genetics” in Coulthard, M. (ed. 1994 - 1998) Advances in Written
Text Analysis, London: Routledge. Pp179- 190.

354
National Education Goals Report: Building a Nation of Learners (1997)
Washington: U.S. Department of Education. National Education Goals
Panel. (http://www.negp.gov/).

Nattinger, J. R. and DeCarrico, J. S. (1992) Lexical Phrases and Language

Teaching Oxford: Oxford University Press.

Nelson, M. (1992) A model for course design in ESP for Business.

Unpublished M.Ed. TESOL dissertation, University of Manchester.

Newmark, L. (1979) “How not to interfere with language learning” in

Brumfit, C. J. and Johnson, K. (1979-81) The Communicative Approach to
Language Teaching, Oxford: Oxford University Press. Pp.160-166

Nunes, M. A. (1999) “Teaching English for Specific Purposes: The Guts to

do it” Actas do 4º Encontro Nacional do Ensino das Línguas Vivas no
Ensino Superior, Faculadade de Letras da Universidade de Porto. Pp.255-
261.

O’Brien, T. & Jordan, R.R. (1985) Developing Reference Skills, London:

Collins.

Opitz, K. (1983) Linguistics between Artificiality and Art: Walking the

Tightrope of LSP Research, CILA Bulletin 37, Neuchatel.

Owen, G. T. (1973) “A Reading/Comprehension Course for Students of

Science and Technology”, ELT Documents 73/4, The British Council.

Palmer, J. (1981) Register research design, in Mackay & Palmer (eds.

1981), Language for Specific Purposes, Rowley, MA: Newbury House.

Paltridge, B. (1996) “Genre, text type, and the language learning

classroom” ELT Journal Volume 50/3 July 1996: Oxford University Press
pp 237-243.

Papert, S. (1980) Mindstorms: Children, Computers and Powerful Ideas,

Brighton:Harvester Press.

Peters, A. (1983). The Units of Language Acquisition. Cambridge:

Cambridge University Press.

Peters P. (1998) “Langscape: Surveying contemporary English usage” in

English Today: The International review of the English Language ET 53
Volume 14 Nº 1 January 1998 Pp 3-6.

355
Phillips, M. (1987) Communicative Language Learning and the
Microcomputer, London: The British Council.

Pongsurapipat, S. (1975) Text and Discourse Analysis on some Aspects of

EST from Three Scientific Textbooks (Biology, General Chemistry, Physics),
SEAMEO Regional English Language Centre, Thailand.

Pope, R. (1998) The English Studies Book, London: Routledge.

Porter D. (1976) Scientific English: An Oversight in Stylistics? Studia

Anglica Posnaniensa, no.8, Poznan, Poland

Pilbeam, A. (1979) “The language audit.” Language Training, 1,2, Pp. 4-5.

Pugh, A. K. & Uljin, J. M. (eds.1984), Reading for Professional Purposes,

London, Heinemann.

Quin, A. and Porter, N. (1994) “Investigating English usage with ICECUP”

in English Today, 33 Vol.9 No.1 January 1993, Cambridge University
Press. Pp 19-24

Quirk, R. (1995), Grammatical and Lexical Variance in English, London,

Longman.

Quirk, R., Greenbaum, S., Leech, G., Svartvik, J. (1985) A Comprehensive

Grammar of the English Language. London: Longman.

Quirk, R. and Widdowson, H. G. (eds.1985) English in the World: Teaching

and Learning the Languages and Literatures, Cambridge: Cambridge
University Press and the British Council.

Renouff, A. (1987) “Corpus Development” in Sinclair, J. M. (ed. 1987)

Looking up. London: Collins. Pp1-22.

Renouff, A. (1987) “Moving On” in Sinclair, J. M. (ed. 1987) Looking up: An

account of the COBUILD Project in lexical computing. London: Collins.
Pp167-178.

Renouff, A. (1997) “Teaching Corpus Linguistics to Teachers of English,” in

Wichmann, A., Fligelstone, S., McEnery, T., Knowles, G. (eds. 1997)
Teaching and Language Corpora, P.255-266 London: Longman.

Renouff, A. and Sinclair, J. McH. (1991)“Collocational Frameworks in

English”, in Sinclair, J. (1991) Corpus, Concordance, Collocation. Oxford
University Press Pp. 128-143.

356
Richterich, R. (1971) Analytical classification of the categories of adults
needing to learn foreign languages. Reprinted in Trim et al.(1973/1980)
Strasbourg: Council of Europe/Oxford: Pergamon.

Richterich, R. (1973/1980) Definition of the language needs and types of

adults. In Trim et al. (1973/1980) Pp 29-88 Strasbourg: Council of
Europe/Oxford: Pergamon.

Riley, P. (1989) “Keeping Secrets: ESP/LSP and the sociology of

knowledge” in the European Journal of Teacher Education, Vol. 12, No. 2,
1989 Pp.69-80.

Riley, R. W. (1997-8) “Foundation of a Nation - Strong and Effective

Schools in Education in the United States: Continuity and Change, U.S.
Society & Values Electronic Journals of the U.S. Information Agency, Vol. 2
Nº 5 December 1997 5-8 and American Studies Journal Number 41
Summer 1998 Pp 55-58

Robinson, M. (1994) “Using Email and the Internet in Science Teaching”,

Journal of Information Technology for Teacher Education Vol. 3, Nº2 1994

Robinson, P. (1980), ESP (English for Specific Purposes), Oxford, Pergamon

Press Ltd.

Robinson, P.C. (1991) ESP TODAY: A Practitioner’s Guide, London: Prentice

Hall International

Rosenthal, J. W. (1996) Teaching Science to Language Minority Students,

Clevedon: Multilingual Matters.

Rowntree, D. (1992) Exploring Open and Distance Learning, London:

Koogan Page.

Scollon R. and Scollon, S. W. (1995) Intercultural Communication: a

discourse approach, Oxford:Blackwell

Sedelow, S., Yeates and Sedelow, W.A. Jr (1994) “A Topologic Model of the
English Semantic Code and its Role in Automatic Disambiguation for
Discourse Analysis” in Hockey, S. and Ide, N. (eds 1994) Research in
Humanities Computing Oxford: Clarenden Press.

Selinker, L. (ed. 1981) English for Academic and Technical Purposes,

London: Newbury House.

Selinker, L. 81991) Rediscovering Interlanguage, London: Longman.

357
Selinker, L. & Trimble, L. (1976) Scientific and technological writing: the
choice of the tense, Forum (English Teaching Forum), 14 (4), 22-26.

Serway, R. A. (1992) Physics for Scientists and Engineers with Modern

Physics Third edition, Updated Version, Saunders Golden Sunburst
Series, USA

Sinclair, J.M. (ed. 1987) Looking Up An account of the COBUILD Project in

lexical computing, London, Harper Collins Publishers

Sinclair, J. (1991) Corpus, Concordance, Collocation. Oxford University

Press

Sinclair, J. McH. (1991) “Shared Knowledge.” In Georgetown University

Round Table on Languages and Linguistics 1991. Washington, DC:
Georgetown University Press, Pp. 489-500.

Sinclair , J. McH. (1992) “Priorities in discourse analysis.” In R.M.

Coulthard (ed.1994 - 1998), Advances in Spoken Discourse Analysis.
London: Routledge, Pp.79-88.

Sinclair, J. McH. (1994-98) “Trust the text” in Coulthard, M. (ed. 1994 -

1998) Advances in Written Text Analysis, London: Routledge. Pp.12-25.

Sinclair, J. McH. (1997) “ Corpus Evidence in Language Description” in

Wichmann, A., Fligelstone, S., McEnery, T., Knowles, G. (eds. 1997)
Teaching and Language Corpora, London: Longman. Pp. 27-39.

Sinclair, J. McH. and Brazil, D. (1982) Teacher Talk, Oxford: Oxford

University Press

Sinclair, J. McH. and Coulthard, R. M. (1975) Towards an Analysis of

Discourse. London: Oxford University Press.

Steffensen, M. and Joag-Dev, C. (1984) “Cultural Knowledge and Reading”

in Alderson, C. and Urquhart, A. (eds. 1984) Reading in a Foreign
Language, London: Longman. Pp48-64.

Stern, H. H., (1983), Fundamental Concepts of Language Teaching, Oxford,

OUP.

Stern, H.H. (1992) Issues and Options in Language Teaching, Oxford:

Oxford University Press.

358
Strevens, P. (1973) Technical, technological and scientific English, English
Language Teaching Journal XXVII/3

Strevens, P. (1978) “Special -purpose Language learning: a perspective,” in

Kinsella, V. (ed. 1978) Language Teaching and Linguistics: Surveys,
Cambridge: Cambridge University Press.

Strevens, P. (1988) “The learner and teacher of ESP” in Chamberlain, R.

and Baumgardner, R. J. (eds. 1988) ESP in the Classroom: Practice and
Evaluation, ELT Documents 128 London: Modern English Publications.
Pp.39-44.

Stuart, W. & Lee, E. (1972/1985) The non-specialist user of foreign

languages in industry and commerce. Sidcup London Chamber of
Commerce & Industry Examination Board.

Stubbs, M. (1992) “English Teaching, Information Technology and Critical

Language Awareness in Fairclough, N. (ed. 1992) Critical Language
Awareness, London: Longman. Pp 203-222.

Stubbs M. (1994) Grammar, Text, and Ideology: Computer-assisted Methods

in the Linguistics of Representation Applied Linguistics, Vol. 15, No.2
Oxford University Press

Stubbs, M. (1996) Text and Corpus Analysis, Oxford: Blackwell

Stubbs, M. & Gerbig, A. (1993) Human and inhuman geography: on the

computer-assisted analysis of long texts, in Hoey, M. (ed. 1993) Data,
Description, Discourse, London: Harper Collins. Pp.64-85

Sudhikam, S. (1975) Lexis, in Guidelines Sample Materials for the Teaching

of English to First Year Tertiary Level Scientific and Technical Students in
Universities in Thailand, SEAMEO/RELC (Mimeo)

Sudnow, D. (ed. 1972) Studies in Social Interaction, New York: The Free
Press.

Svartvik, J. and Quirk, R. (eds.) A corpus of English conversation. Lund:

CWK Gleerup.

Swales, J. (1971) Writing Scientific English, London: Nelson

Swales, J. (1973) “Introducing Teachers to English for Science and

Technology”, ELT Documents 73/6 The British Council.

359
Swales, J. (1978) Writing “Writing Scientific English”, in Mackay &
Mountford (eds 1978) English for Specific Purposes, London: Longman. Pp.
43-55.
Swales, J. (1985), Episodes in ESP, Oxford: Pergamon Press Ltd.

Swales, J. M. (1990) Genre Analysis. Cambridge: Cambridge University

Press.

Swales, J. & Mustafa, H. (eds.1984) English for Specific Purposes in the

Arab World, Birmingham: University of Aston.

Swales, J.M. and Najjar, H. (1987) “The writing of research article

introductions.” Written Communication, 9(2), Pp.175-192.

Swift, J.(1976) Gulliver’s Travels, Oxford University Press.

Tarone, E. & Yule, G. (1989) Focus on the language learner. Oxford: Oxford
University Press.

Tavares, C. F., Valente, M. T. and Roldão, M. C. (1996) “Dimensões

Formativas de Disciplinas do Ensino Básico Língua Estrangeira.” Lisboa:
Instituto de Inovação Educacional.

Tavares, J., Santiago, R.A, Lencestre, L., Soares, I. (1996) Niveis de

Sucesso dos Alunos do 1º Ano dos Cursos de Ciências e Engenharia da
Universidade de Aveiro, Universidade de Aveiro.

Tesch, F. (1990) Die Indefinitpronomina some und any in authentischen

englischen Sprachgebrauch und in Lehrwerken:eine empirische
Untersuchung. Narr, Tubingen.

Tickoo, M. (1988) “Michael West in India: a centenary salute.” English

Language Teaching Journal, 42,4, Pp.294-300.

Todaka, Y. (1996) “Between and among: a data-based analysis” in Journal

of the International Linguistic Association WORD, Volume47, Number 1
(April 1996). Pp. 13-39.

Tribble, C. & Jones, G. (1990) Concordances in the Classroom, London,

Longman.

Trim, J. L. M., (1998) “European perspectives on modern language

learning” Language Teaching 31, October 1998,. Cambridge University
Press. Pp.206-217.

360
Trim, J. L. M, Richterich, R., Van Ek, J. & Wilkins, D. (1973/80). Systems
development in adult language learning. Strasbourg: Council of
Europe/Oxford: Pergamon.

Trimble, L. (1985), English for Science and Technology. A discourse

approach, Cambridge, CUP.

Todd Trimble, M & Trimble, L Rhetorical-Grammatical Features of Scientific

and Technical Texts as a major Factor in Written Scientific Communication,
in Hoedt, J., Lundquist, L., Picht, H. & Qvistgaard, J. (eds. 1982))
Proceedings of the 3rd European Symposium on LSP, Copenhagen August
1981, The Copenhagen School of Economics. Pp 199-218

Turner, , R. (1981) “A note on “special languages” and “specific purposes””,

in Hoedt, J. & Turner, R. (eds. 1981) New Bearings in LSP, The
Copenhagen School of Economics.

van Dijk, T. A., (ed. 1997) Discourse as Structure and Process, London:
SAGE Publications.

van Dijk, T. A., (ed. 1997) Discourse as Social Interaction, London: SAGE
Publications.

Van Ek, J. A., The Threshold Level for Modern Language Learning in
Schools Council of Europe, Strasbourg 1976, Longman 1977

Van Ek, J. A., Alexander, L. G. & Fitzpatrick, M. A. (1980) Waystage

English, Oxford: Oxford University Press.

Van Ek, J. A. and Trim, J. (1998) Threshold 1990, Cambridge: Cambridge

University Press.

Veel, R. (1998) “The greening of school science: Ecogenesis in secondary

classrooms” in Martin, J. R. and Veel, R. (1998) Reading Science, London:
Routledge. Pp. 114-151.

Warschauer, M. (1999) “CALL vs. Electronic Literacy: Reconceiving

Technology in the Language Classroom” London: CILT Research Forum.

Waters, M. and Waters, A. (1992) “Study skills and study competence:

getting the priorities right” in ELT Journal Volume 46/3 July 1992 Oxford:
Oxford University Press.

361
Weber, H. Language for specific Purposes, Text Typology, and Text
Analysis: Aspects of a Pragmatic-Functional Approach in Hoedt, J. &
Lundquist, L., Picht, H. & Qvistgaard, J. (eds. 1982)) Proceedings of the
3rd European Symposium on LSP, Copenhagen August 1981, The
Copenhagen School of Economics. Pp. 219- 234.

Weir, C. (1988) The Specification, Realization and Validation of an English

Language Proficiency Test in Testing English for University Study, ELT
Documents: 127, MEP 1988

West, M. (1953) Supplementary Scientific and Technical Vocabulary, in A

General Service List of English Words, London: Longman

West, R. (1994) “Needs analysis in language teaching: State of the art

article,” Language Teaching, January 1994, Cambridge University Press,
Pp 1-19.

White, R. (1988). The ELT Curriculum: Design, Innovation and Management,

Oxford: Blackwell.

White, P. R. R. (1998) “Extended reality, proto-nouns and the vernacular:

Distinguishing the technological from the scientific” in Martin, J. R. and
Veel, R. (1998) Reading Science, London: Routledge. Pp.266- 296.

Wichmann, A., Fligelstone, S., McEnery, T., Knowles, G. (eds. 1997)

Teaching and Language Corpora, London: Longman.

Widdowson, H. G. (1974) The deep structure of discourse and the use of

translation, in Corder, S. P. & Roulet, E. (eds. 1974) Linguistic Insights in
Applied Linguistics, Second Neuchatel Colloquium in Applied Linguistics,
Brussels: AIMAV; Paris: Didier and in Brumfit, C. J. and Johnson, K.
(1979-81) The Communicative Approach to Language Teaching, Oxford:
Oxford University Press. Pp.61-71.

Widdowson, H. G. (1974) An approach to the teaching of scientific English

discourse, RELC Journal, 5, Pp 27-40.

Widdowson, H. G. (ed. 1979), Reading and Thinking in English: Discovering

Discourse, Oxford, Oxford University Press.

Wilkins, D. A (1976) Notional Syllabuses, Oxford: Oxford University Press.

Wilks, Y.A., Slator, B.M.,& Guthrie, L.M. (1996) Electric Words: Dictionaries
Computers and Meanings, Mass: MIT Press.

362
Wilson, E. (1997) “The Automatic Generation of CALL Exercises from
General Corpora” in Wichmann, A., Fligelstone, S., McEnery, T., Knowles,
G. (eds. 1997) Teaching and Language Corpora, London: Longman. Pp.116-
130

Wingard, P. (1981) Some verb forms and functions in six medical texts, in
Selinker (ed. 1981) London, Newbury House.

Wong-Fillmore, L. (1976) The Second Time Around: Cognitive and Social

Strategies in Second Language Acquisition. Unpublished doctoral
dissertation, Stanford University, quoted in Nattinger, J.R. and DeCarrico,
J. S. (1992) Lexical Phrases and Language Teaching Oxford: Oxford
University Press.

“World Education League: Who’s top?” The Economist March 29th 1997
Pp.21-25.

363
Appendices
 Appendix A

Biber’s Algorithms and Functions

Appendix A Biber’s Algorithms and Functions
67 linguistic features were counted. These features include all features that: (1)
have been assigned distinctive functions by previous research, and (2) can be
automatically identified in spoken and written texts. Each of these features is
discussed in turn here.
The following notation is used in the descriptions of the algorithms:

+: used to separate constituents

0: marks optional constituents
1: marks disjunctive options
xxx: stands for any word
#: marks a word boundary
T#: marks a 'tone unit' boundary, as defined in Quirk et al. (1972: 937-8)
for use in the London-Lund corpus.'
DO: do, does, did, don't, doesn't, didn't, doing, done
HAVE: have, has, had, having, -'ve#, -'d#, haven't, hasn't, hadn't
BE: am, is, are, was, were, being, been -'m#, -'re#, isn't, aren't, wasn't, weren't
MODAL: can, may, shall, will, -'ll#, could, might, should, would, must,
can't, won't, couldn't, mightn't, shouldn't, wouldn't, mustn't
AUX: MODAL/ DO/ HAVE/ BE/ -'s
SUBJPRO: I, we, he, she, they (plus contracted forms)
OBJPRO: me, us, him, them (plus contracted forms)
POSSPRO: my, our, your, his, their, its (plus contracted forms)
REFLEXPRO: myself, ourselves, himself, themselves, herself, yourself,
yourselves, itself
PRO:SUBJPRO/OBJPRO/POSSPRO/REFLEXPRO/you/her/it
PREP: prepositions (e.g. at, among - see no. 39)
CONJ: conjuncts (e.g. furthermore, therefore - see no. 45)
ADV: adverbs (see no. 42)
ADJ: adjectives (see nos. 40, 41)
N: nouns (see nos. 14, 15, 16)
VBN: any past tense or irregular past participial verb
VBG: -ing form of verb
VB: base form of verb
VBZ: third person, present tense form of verb
PUB: 'public' verbs (see no. 55)
PRV: 'private' verbs (see no. 56)
SUA: 'suasive' verbs (see no. 57)
V: any verb
WHP: WH pronouns - who, whom, whose, which
WHO: other WH words - what, where, when, how, whether, why,
whoever, whomever, whichever, wherever, whenever, whatever, however
ART: articles - a, an, the, (dhi)
DEM: demonstratives - this, that, these, those
QUAN: quantifiers - each, all, every, many, much, few, several, some,
any
NUM: numerals - one… twenty, hundred, thousand
DET: ART/DEM/QUAN/NUM
369
 ORD: ordinal numerals -first ... tenth
QUANPRO: quantifier pronouns - everybody, somebody, anybody,
everyone, someone, anyone, everything, something, anything
TITLE: address titles
CL-P: clause punctuation (‘.’, ‘!’, ‘?’, ‘:’, ‘;’,’-‘)
ALL-P: all punctuation (CL-P plus',')

In the following discussion, the 67 linguistic features have been grouped

into sixteen major categories: (A) tense and aspect markers, (B) place and time
adverbials, (C) pronouns and pro-verbs, (D) questions, (E) nominal forms, (F)
passives, (G) stative forms, (H) subordination features, (I) adjectives and
adverbs, (j) lexical specificity, (K) specialized lexical classes, (L) modals, (M)
specialized verb classes, (N) reduced or dispreferred forms, (0) coordination, and
(P) negation.

(A) TENSE AND ASPECT MARKERS (nos. 1-3)

1. past tense
Any past tense form that occurs in the dictionary, or any word not
otherwise identified that is longer than six letters and ends in ed#. Past tense
forms are usually taken as the primary surface marker of narrative.
2. perfect aspect
(a) HAVE + (ADV) + (ADV) + VBN
(b) HAVE+NIPRO+VBN (questions)
(includes contracted forms of HAVE)
Perfect aspect forms mark actions in past time with ‘current relevance’
(Quirk et al. 1985:189ff).
3. present tense
All VB (base form) or VBZ (third person singular present) verb forms in the
dictionary, excluding infinitives.
Present tense verbs deal with topics and actions of immediate relevance. They
can also be used in academic styles to focus on the information being presented
and remove focus from any temporal sequencing.

(B) PLACE AND TIME ADVERBIALS (nos. +-5)

mark direct reference to the physical and temporal context of the text, or in the
case of fiction, to the text-internal physical and temporal world.
4. place adverbials aboard, above, abroad, across, ahead, alongside, around,
ashore, astern, away, behind, below, beneath, beside, downhill, downstairs,
downstream, east, far, hereabouts, indoors, inland, inshore, inside, locally, near,
nearby, north, nowhere, outdoors, outside, overboard, overland, overseas, south,
underfoot, underground, underneath, uphill, upstairs, upstream, west
This list is taken from Quirk et al. (1985:514ff). Items with other major
functions, for example, in, on, which often mark logical relations in a text, have
been excluded from the list.
5. time adverbials
afterwards, again, earlier, early, eventually, formerly, immediately, initially,
instantly, late, lately, later, momentarily, now, nowadays, once, originally,
370
Appendix A Biber’s Algorithms and Functions
presently, previously, recently, shortly, simultaneously, soon, subsequently,
today, tomorrow, tonight, yesterday
This list is taken from Quirk et al. (1985:526ff). Items with other major
functions, for example, last, next, which often mark logical relations within a
text, have been excluded from the list.

(C) PRONOUNS AND PRO-VERBS (nos. 6-12)

Some studies have grouped all pronominal forms together as a single category
which is interpreted as marking a relatively low informational load, a lesser
precision in referential identification, or a less formal style (e.g., Kroch and
Hindle 1982; Brainerd 1972). Other studies have grouped all personal pronouns
into a single category, and interpret that category as marking interpersonal focus
(Carroll 1960; Poole 1973; Poole and Field 1976). In the present analysis, I
separate personal and impersonal pronominal forms, as well as each of the
persons within the personal pronouns.

1. (C1) PERSONAL PRONOUNS

6. first person pronouns
I, me, we, us, my, our, myself, ourselves (plus contracted forms)
First person pronouns have been treated as markers of ego-involvement in a
text.
7. second person pronouns
you, your, yourself, yourselves (plus contracted forms)
Second person pronouns require a specific addressee and indicate a high
degree of involvement with that addressee (Chafe 1985). They have been used as
a marker of register differences by Hu (1984), Finegan (1982), and Biber
(1986a).
8. third person personal pronouns
she, he, they, her, him, them, his, their, himself, herself, themselves (plus
contracted forms)
Third person personal pronouns mark relatively inexact reference to persons
outside of the immediate interaction.

(C2) IMPERSONAL PRONOUNS

9. pronoun it
It is the most generalized pronoun, since it can stand for referents ranging from
animate beings to abstract concepts. This pronoun can be substituted for nouns,
phrases, or whole clauses. Chafe and Danielewicz (1986) and Biber (1986a) treat
a frequent use of this pronoun as marking a relatively inexplicit lexical content
due to strict time constraints and a non-informational focus. Kroch and Hindle
(1982) associate greater generalized pronoun use with the limited amounts of
information that can be produced and comprehended in typical spoken situations.
10.demonstrative pronouns (e.g., this is ridiculous)
(a) that/ this/ these/ those + V/AUX/ CL-P/T#/WHP/and
(where that is not a relative pronoun)
(b) that's
(c) T# + that
371
 (that in this last context was edited by hand to distinguish among
demonstrative pronouns, relative pronouns, complementizers, etc.)
Demonstrative pronouns can refer to an entity outside the text, an
exophoric referent, or to a previous referent in the text itself. In the latter case, it
can refer to a specific nominal entity or to an inexplicit, often abstract, concept
(e.g., this shows ...). Chafe (1985; Chafe and Danielewicz 1986) characterizes
those demonstrative pronouns that are used without nominal referents as errors
typically found in speech due to faster production and the lack of editing.
11.indefinite pronouns
anybody, anyone, anything, everybody, everyone, everything, nobody, none,
nothing, nowhere, somebody, someone, something (Quirk et al. 1985:376ff)
These forms have not been used frequently for register comparison. They are
included here as markers of generalized pronominal reference, in a similar way to
it and the demonstrative pronouns.

(C3) PRO-VERBS
12.pro-verb do (e.g., the cat did it)
DO when NOT in the following constructions:
DO + (ADV) + V (DO as auxiliary)
ALL-P/T#/WHP+DO (DO as question)
This feature was included in Biber (1986a) as a marker of register
differences. Do as pro-verb substitutes for an entire clause, reducing the
informational density of a text and indicating a lesser informational focus, due to
processing constraints or a higher concern with interpersonal matters.

(D) QUESTIONS (no. 13)

Questions, like second person pronouns, indicate a concern with interpersonal
functions and involvement with the addressee (Marckworth and Baker 1974;
Biber 1986a). Yes/no questions were excluded from the present analysis because
they could not be accurately identified by automatic analysis in formal spoken
genres, where every phrase tends to be a separate intonation unit; that is, many
intonation units begin with an auxiliary and therefore are identical in form to
direct questions.
13. direct WH-questions
CL-P/T#+WHO+AUX
(where AUX is not part of a contracted form)

(E) NOMINAL FORMS (nos. 14-16)

The overall nominal characterization of a text and the distinction between
nominal and verbal styles is identified as one of the most fundamental
distinctions among registers by Wells (1960) and Brown and Fraser (1979). A
high nominal content in a text indicates a high (abstract) informational focus, as
opposed to primarily interpersonal or narrative foci. Nominalizations, including
gerunds, have particularly been taken as markers of conceptual abstractness.
14. nominalizations
All words ending in -tion#, -ment#, -ness#, or -ity# (plus plural forms).

372
Appendix A Biber’s Algorithms and Functions
Nominalizations have been used in many register studies. Chafe (1982, 1985,
and Danielewicz 1986) focuses on their use to expand idea units and integrate
information into fewer words. Biber (1986a) finds that they tend to co-occur with
passive constructions and prepositions and thus interprets their function as
conveying highly abstract (as opposed to situated) information. Janda (1985)
shows that nominalizations are used during note-taking to reduce full sentences
to more compact and efficient series of noun phrases.
15. gerunds
All participle forms serving nominal functions - these are edited by hand.
Gerunds (or verbal nouns) are verbal forms serving nominal functions. As
such, they are closely related to nominalizations in their functions. Some
researchers (e.g., Chafe 1982) do not distinguish among the different participial
functions, treating gerunds, participial adjectives (nos. 40-1), and participial
clauses (nos. 25-8) as a single feature. In the present study, these functions are
treated separately.
16. total other nouns
All nouns included in the dictionary, excluding those forms counted as
nominalizations or gerunds.
This count provides an overall nominal assessment of a text. Nominalizations
and gerunds are excluded from the total noun count so that the three features will
be statistically independent.

(F) PASSIVES (nos. 17-18)

Passives have been taken as one of the most important surface markers of the
decontextualized or detached style that stereotypically characterizes writing. In
passive constructions, the agent is demoted or dropped altogether, resulting in a
static, more abstract presentation of information. Passives are also used for
thematic purposes (Thompson 1982; Finegan 1982; Weiner and Labov 1983;
Janda 1985). From this perspective, agentless passives are used when the agent
does not have a salient role in the discourse; by-passives are used when the
patient is more closely related to the discourse theme than the patient.
17. agentless passives 18. by-passives**
(a) BE + (ADV) + (ADV) + VBN + (by)**
(b) BE + N/PRO + VBN + (by)** (question form)
(** no. 18 with the by-phrase)

(G) STATIVE FORMS (nos. 19-20)

Only a few studies have used stative forms for register comparisons. These forms
might be considered as markers of the static, informational style common in
writing, since they preclude the presence of an active verb. Conversely, they can
be considered as non-complex constructions with a reduced informational load,
and therefore might be expected to be more characteristic of spoken styles. Kroch
and Hindle (1982) analyze existential there as being used to introduce a new
entity while adding a minimum of other information. Janda (1985) notes that
stative or predicative constructions (X be Y) are used frequently in note-taking,
although the be itself is often dropped. Predicative constructions with be-ellipsis
are also common in sports announcer talk (Ferguson 1983). These predicative
373
 constructions might be characterized as fragmented, because they are typically
alternatives to more integrated attributive constructions (e.g., the house is big
versus the big house). Be as main verb is used for register comparisons by Carroll
(1960) and Marckworth and Baker (1974).
19.be as main verb
BE + DET/POSSPRO/TITLE/PREP/ADJ
20.existential there (e.g., there are several explanations ...)
(a) there + (xxx) + BE
(b) there's

(H) SUBO.RDINATION (nos. 21-38)

Subordination has perhaps been the most discussed linguistic feature used for
register comparisons. It has generally been taken as an index of structural
complexity and therefore supposed to be more commonly used in typical writing
than typical speech. Some researchers, though, have found higher use of
subordination in speech than writing (e.g., Poole and Field 1976). Halliday
(1979) claims that even conversational speech has more subordination than
written styles, because the two modes have different types of complexities:
spoken language, because it is created and perceived as an on-going process, is
characterized by 'an intricacy of movement [and by] complex sentence structures
with low lexical density (more clauses, but fewer high-content words per
clause)'; written language, in which the text is created and perceived as an object,
is characterized by 'a denseness of matter [and by] simple sentence structures
with high lexical density (more high-content words per clause, but fewer
clauses)'.
Thompson (1983,1984,1985; Thompson and Longacre 1985; Ford and
Thompson to appear) has carried out some of the most careful research into the
discourse functions of subordination. She distinguishes between dependent
clauses (complementation and relative clauses) and other types of subordination
(e.g., adverbial clauses) that function to frame discourse information in different
ways. Her studies have focused on the discourse functions of detached participial
clauses, adverbial clauses in general, purpose clauses, and conditional clauses
(see below). In all of these studies, Thompson emphasizes that subordination is
not a unified construct, that different types of structural dependency have
different discourse functions, and that particular subordination features are
therefore used to different degrees in different types of discourse.
Beaman (1984) and Biber (1986a) also find that different subordination
forms are distributed differently. Based on an analysis of spoken and written
narratives, Beaman observes that there are more finite nominal clauses
(that-clauses and WH-clauses) in speech and more non-finite nominal clauses
(infinitives and participial clauses) in writing. She also discusses the distribution
of relative and adverbial clauses in these texts (see below). In my own earlier
studies, I find that that-clauses, WH-clauses, and adverbial subordinators co-
occur frequently with interpersonal and reduced-content features such as first and
second person pronouns, questions, contractions, hedges, and emphatics. These
types of subordination occur frequently in spoken genres, both interactional
(conversation) and informational (speeches), but they occur relatively
374
Appendix A Biber’s Algorithms and Functions
infrequently in informational written genres. Relative clauses and infinitive were
found to have a separate distribution from the other types of subordination, but
they did not form a strong enough co-occurrence pattern for interpretation. These
same features are discussed from the perspective of discourse complexity in
Finegan and Biber (1986b).
These studies by Thompson and Beaman, and my own earlier studies, all
show that different types of subordination function in different ways. Based on
these analyses, I have divided the subordination features used in the present study
into four sub-classes: complementation (H1), participial forms (H2), relative
clauses (H3), and adverbial clauses (H4). Each of these is now discussed in turn.

(H1) COMPLEMENTATION (nos. 21-4)

21.that verb complements (e.g., I said that he went)
(a) and/nor/but/or/also/ ALL-P + that +
DET/PRO/there/plural noun/proper noun/TITLE
(these are that-clauses in clause-initial positions)
(b)PUB/PRV/SUA/SEEM/APPEAR + that + xxx
(where xxx is NOT: V/AUX/CL-P/T#/and)
(that-clauses as complements to verbs which are not included in the above verb
classes are not counted - see Quirk et al. 1985:1179ff.)
(c) PUB/PRV/SUA + PREP + xxx + N + that
(where xxx is any number of words, but NOT = N)
(This algorithm allows an intervening prepositional phrase between a verb and its
complement.)
(d)T# + that
(This algorithm applies only to spoken texts. Forms in this context are checked
by hand, to distinguish among that complements, relatives, demonstrative
pronouns and subordinators.)
Chafe (1982, 1985) identifies that-complements as one of the indices of
integration, used for idea-unit expansion in typical writing. Ochs (1979)
describes complementation as a relatively complex construction used to a greater
extent in planned than unplanned discourse. In contrast, Beaman (1984) finds
more that complementation in her spoken than written narratives. Biber (1986a)
finds that that-complements co-occur frequently With interactive features such as
first and second person pronouns and questions, and that all of these features are
more common in spoken than written genres. In that paper and in Finegan and
Biber (1986b), this is interpreted in a similar way to Halliday's characterization:
that this type of structural complexity is used in typical speech, where there is
little opportunity for careful production or revision, while other types of
linguistic complexity, notably lexical variety and density, are used in typical
academic writing, which provides considerable opportunity for production and
revision.
Other studies that have used that-complements for register comparisons
include Carroll (1960), O'Donnell (1974), Frawley (1982), and Weber (1985).
Winter (1982) notes that both verb and adjective that-complements provide a
way to talk about the information in the dependent clause, with the speaker's

375
evaluation (commitment, etc.) being given in the main clause and the
propositional information in the that-clause.
Some verb complements do not have an overt complementizer (e.g., I
think he went); these are counted as a separate feature (no. 60).
22.that adjective complements (e.g., I'm glad that you like it)
ADJ + (T#) + that
(complements across intonation boundaries were edited by hand)
Most studies of that-clauses consider only verb complements. Winter
(1982) points out, however, that verb and adjective complements seem to have
similar discourse functions, and so both should be important for register
comparisons. Because there is no a priori way to know if that verb and adjective
complements are distributed in the same way among genres, they are included as
separate features here. Householder (1964) has compiled a list of adjectives that
occur before that-clauses; Quirk et al. (1985:1222-5) give a grammatical and
discourse description of these constructions.

23.WH-clauses (e.g., I believed what he told me)

PUB/PRVISUA + WHP/WHO + xxx
(where xxx is NOT = AUX - this excludes WH questions)
This algorithm captures only those WH clauses that occur as object
complements to the restricted verb classes described below in nos. 5 5-7; see
Quirk et al. 1985:1184-5. Other WH clauses could not he identified reliably by
automatic analysis and so were not counted.
Similar to that-clauses, WH-clauses are complements to verbs. Chafe
(1985) analyzes them as being used for idea unit expansion, and thus they should
be more frequent in typical writing. Bearnan (1984) did not find WH-clauses in
her written narratives; she writes that they resemble questions and serve
interpersonal functions in discourse, accounting for their use in spoken but not
written narratives. Winter (1982) notes that WH complements provide a way to
talk about questions in the same way that that-complements provide a way to talk
about statements, that is, with the speaker's evaluation, commitment, etc.
provided in the main clause. Biber (1986a) finds WH-clauses to be distributed in
a similar pattern to that-clauses, both of which co-occur frequently with
interpersonal features such as first and second person pronouns and questions.
24. infinitives
to + (ADV) + VB
Infinitives are the final form of complementation to be included in the
present study. The algorithm above groups together all infinitival forms:
complements to nouns, adjectives, and verbs, as well as ‘purpose’ adverbial
clauses (see below). The distribution and discourse functions of infinitives seem
to be less marked than that of other types of subordination. Chafe (1982, 1985)
includes infinitives as one of the devices used to achieve integration and idea-
unit expansion in typical writing. Beaman (1984) finds that infinitives co-occur
with other non-finite nominal clauses (especially participial clauses), and that
they are more common in written than spoken narratives. Biber (1986a) finds a
weak co-occurrence relationship between infinitives and relative clauses. Finally,
Thompson (1985) carefully distinguishes between those infinitives functioning as
376
Appendix A Biber’s Algorithms and Functions
complements and those functioning as adverbial purpose clauses, and she
analyzes the thematic discourse functions of the latter in some detail. Although
this is an important functional distinction, it is not made here because of the
limitations of the automatic analysis.

(H2) PARTICIPIAL FORMS (nos. 25-8)

Participles are among the most difficult forms to analyze. They can function as
nouns, adjectives, or verbs, and within their use as verbs, they can function as
main verbs (present progressive, perfect, or passive), complement clauses,
adjectival clauses, or adverbial clauses. Some studies do not distinguish among
these functions, counting all participial forms (excluding main verbs) as a single
feature (e.g., Chafe 1982; Beaman 1984). Many studies also do not distinguish
between present and past participial clauses, or they count only present participle
forms. In the present analysis, each of the different grammatical functions of
participles is treated as a separate linguistic feature, since these grammatical
functions are likely to be associated with different discourse functions.
Studies that consider participles typically find that they occur more
frequently in writing than in speech; the usual interpretation associated with this
distribution is that participles are used for integration or structural elaboration.
Thompson (1983) distinguishes syntactically detached participial clauses
(e.g., Stuffing his mouth with cookies, Joe ran out the door) from other participial
functions. She shows how these clauses are used for depictive functions, that is,
for discourse that describes by creating an image. No. 25 and no. 26 below are
algorithms for detached participial clauses (present and past). These forms were
edited by hand to exclude participial forms not having an adverbial function.
Participial clauses functioning as reduced relatives, also known as WHIZ
deletions, are treated separately (nos. 27 and 28). Janda (1985) notes the use of
these forms in note-taking to replace full relative clauses, apparently because
they are more compact and integrated and therefore well-suited to the production
of highly informational discourse under severe time constraints. In the present
analysis, these forms were also edited by hand to distinguish between subordinate
clause functions and other functions; in particular, past participles following a
noun can represent either a simple past tense form or the head of a reduced
relative clause, and these forms thus needed to be checked by hand. Finally,
participles functioning as nouns and adjectives were distinguished (nos. 15 and
40-1 respectively); these forms were also edited by hand to verify their
grammatical function.
25.present participial clauses
(e.g., Stuffing his mouth with cookies, Joe ran out the door)
T#/ALL-P + VBG + PREP/DET/WHP/WHO/PRO/ADV
(these forms were edited by hand)

26.past participial clauses

(e.g., Built in a single week, the house would stand for fifty years.)
T#/ALL-P + VBN + PREP/ADV
(these forms were edited by hand)
27.past participial WHIZ deletion relatives
377
(e.g., the solution produced by this process)
N/QUANPRO + VBN + PREP/BE/ADV
(these forms were edited by hand)

28.present participial WHIZ deletion relatives

(e.g., the event causing this decline is . . . )
N + VBG
(these forms were edited by hand)

(H3) RELATIVES (nos. 29-34)

Relative clauses have been used frequently as a marker of register differences.
Relatives provide a way to talk about nouns, either for identification or simply to
provide additional information (Winter 1982; Beaman 1984). Ochs (1979) notes
that referents are marked differently in planned and unplanned discourse: simple
determiners are preferred in unplanned discourse while relative clauses are used
for more exact and explicit reference in planned discourse. Chafe (1982, 1985)
states that relative clauses are also used as a device for integration and idea unit
expansion.
In general, these studies find that relative clauses occur more frequently
in writing than in speech. Some studies, however, do not treat all relative clauses
as a single feature and consequently do not find a uniform distribution. Kroch
and Hindle (1982) and Beaman (1984) provide two of the fullest discussions.
Beaman analyzes that relatives separately from WH relatives and finds more that
relatives in her spoken narratives but more WH relatives in her written
narratives; further, she finds more relativization on subject position in her spoken
narratives versus more relativization on object positions in her written narratives.
In contrast, Kroch and Hindle find more relativization on subject position in their
written texts and more relativization on object position in their spoken texts.
They attribute this to a greater use of pronouns in subject position in speech,
making this position unavailable for relativization. Both of these studies also
analyze pied-piping constructions separately, finding more in written than in
spoken texts. In the present analysis, I separate that from WH relatives, and
relativization on subject position from relativization on object position. Pied-
piping constructions are also treated separately.
29.that relative clauses on subject position
(e.g., the dog that bit me)
N + (T#) + that + (ADV) + AUX/V
(that relatives across intonation boundaries are identified by hand.)
30.that relative clauses on object position
(e.g., the dog that I saw)
N + (T#) + that + DET/ SUBJPRO / POSSPRO / it/ ADJ / plural noun/ proper
noun / possessive noun / TITLE
(This algorithm does not distinguish between that complements to nouns and true
relative clauses.)
(In spoken texts, that relatives sometimes span two intonation units; these are
identified by hand.)
31.WH relative clauses on subject position
378
Appendix A Biber’s Algorithms and Functions
(e.g., the man who likes popcorn)
xxx + yyy + N + WHP + (ADV) + AUX/V
(where xxx is NOT any form of the verbs ASK or TELL; to exclude indirect WH
questions like Tom asked the man who went to the store)
32.WH relative clauses on object positions
(e.g., the man who Sally likes)
xxx + yyy + N + WHP + zzz
(where xxx is NOT any form of the verbs ASK or TELL, to exclude indirect WH
questions, and zzz is not ADV, AUX or V, to exclude relativization on subject
position)
33.pied-piping relative clauses
(e.g., the manner in which he was told)
PREP + WHP
34.sentence relatives
(e.g., Bob likes fried mangoes, which is the most disgusting thing I've ever heard
of)
T#/, +which
(These forms are edited by hand to exclude non-restrictive relative clauses.)
Sentence relatives do not have a nominal antecedent, referring instead to
the entire predication of a clause (Quirk et al. 1985:1118-20). They function as a
type of comment clause, and they are not used for identificatory functions in the
way that other relative clauses are. A preliminary analysis of texts suggested that
these constructions were considerably more frequent in certain spoken genres
than in typical writing, and they are therefore included here as a separate feature.

(H4) ADVERBIAL CLAUSES (nos. 35-8)

Adverbial clauses appear to be an important device for indicating informational
relations in a text. Overall, Thompson (1984) and Biber (1986a) find more
adverbial clauses in speech than in writing. Several studies, though, separate
preposed from postposed adverbial clauses, and find that these two types have
different scopes, functioning to mark global versus local topics, and that they
have different distri-butions (Winter 1982; Chafe 1984a; Thompson 1985;
Thompson and Longacre 1985; Ford and Thompson 1986).
There are several subclasses of adverbial clauses, including condition,
reason/cause, purpose, comparison, and concession (Quirk et al. 1985:1077-18;
Tottie 1986; Smith and Frawley 1983). The most common types, causative,
concessive, and conditional adverbials, can be identified unambiguously by
machine (nos. 35-7); the other subordinators are grouped together as a general
category (no. 38).
35.causative adverbial subordinators: because
Because is the only subordinator to function unambiguously as a causative
adverbial. Other forms, such as as, for, and since, can have a range of functions,
including causative. Most researchers find more causative adverbials in speech
(Beaman 1984; Tottie 1986), although the functional reasons for this distribution
are not clear. Tottie (1986) and Altenberg (1984) both provide detailed analyses
of these subordination constructions. For example, Tottie notes that while there is

379
more causative subordination overall in speech, the form as is used as a causative
subordinator more in writing.
36.concessive adverbial subordinators: although, though
Following a general pattern for adverbial clauses, concessive adverbials can also
be used for framing purposes or to introduce background information, and they
have different functions in pre- and post-posed positions (McClure and Geva
1983; Altenberg 1986). Both Altenberg and Tottie (1986) find more concessive
subordination overall in writing.
37.conditional adverbial subordinators: if, unless
Conditional clauses are also used for discourse framing and have differing
functions when they are in pre- or post-posed position (Ford and Thompson
1986). Finegan (1982) finds a very frequent use of conditional clauses in legal
wills, due to the focus on the possible conditions existing when the will is
executed. Several researchers have found more conditional clauses in speech than
in writing (Beaman 1984; Tottie 1986; Biber 1986a; Ford and Thompson 1986),
but the functional reasons for this distribution are not clear.
38.other adverbial subordinators: (having multiple functions)
since, while, whilst, whereupon, whereas, whereby, such that, so that xxx, such
that xxx, inasmuch as, forasmuch as, insofar as, insomuch as, as long as, as soon
as
(where xxx is NOT: N/ADJ)

(I1) PREPOSITIONAL PHRASES (no. 39)

39.total prepositional phrases
against, amid, amidst, among, amongst, at, besides, between, by, despite, during,
except, for, from, in, into, minus, notwithstanding, of, off, on, onto, opposite, out,
per, plus, pro, re, than, through, throughout, thru, to, toward, towards, upon,
versus, via, with, within, without
This list of prepositions is taken from Quirk et al. (1985:665-7),
excluding those lexical items that have some other primary function, such as
place or time adverbial, conjunct, or subordinator (e.g., down, after, as)
Prepositions are an important device for packing high amounts of
information into academic nominal discourse. Chafe (1982, 1985; and
Danielewiez 1986) describes prepositions as a device for integrating information
into idea units and expanding the amount of information contained within an idea
unit. Biber (1986a) finds that prepositions tend to co-occur frequently with
nominalizations and passives in academic prose, official documents, professional
letters, and other informational types of written discourse.

(I2) ADJECTIVES AND ADVERBS (nos. 40-2)

Adjectives and adverbs also seem to expand and elaborate the information
presented in a text. Chafe (1982, 1985; and Danielewicz 1986) groups adjectives
together with prepositional phrases and subordination constructions as devices
used for idea unit integration and expansion. However, the descriptive kinds of
information presented by adjectives and adverbs do not seem equivalent to the
logical, nominal kinds of information often presented in prepositional phrases. In
my earlier work (e.g., Biber 1986a), I find that prepositions, subordination
380
Appendix A Biber’s Algorithms and Functions
features, adjectives, and adverbs are all distributed differently; for example,
prepositional phrases occur frequently in formal, abstract styles, while many
types of subordination occur frequently in highly interactive, unplanned
discourse; adjectives and adverbs are distributed in yet other ways. All of these
features elaborate information in one way or another, but the type of information
being elaborated is apparently different in each case.
Some studies distinguish between attributive and predicative adjectives
(e.g., Drieman, O'Donnell, and Chafe). Attributive adjectives are highly
integrative in their function, while predicative adjectives might be considered
more fragmented. In addition, predicative adjectives are frequently used for
marking stance (as heads of that or to complements; see Winter 1982). The
present analysis distinguishes between attributive and predicative adjectives,
including both participial and non-participial forms.
40.attributive adjectives (e.g., the big horse)
ADJ + ADJ/N
(+ any ADJ not identified as predicative - no. 41)
41.predicative adjectives (e.g., the horse is big)
(a) BE + ADJ + xxx
(where xxx is NOT ADJ, ADV, or N)
(b)BE + ADJ + ADV + xxx
(where xxx is NOT ADJ or N)
42.total adverbs
Any adverb form occurring in the dictionary, or any form that is longer than five
letters and ends in -ly. The count for total adverbs excludes those adverbs
counted as instances of hedges, amplifiers, downtoners, amplifiers, place
adverbials, and time adverbials.

(J) LEXICAL SPECIFICITY (nos. 43-4)

Two measures of lexical specificity or diversity are commonly used: type/token
ratio and word length. Unlike structural elaboration, differences in lexical
specificity seem to truly correlate with the production differences between
speaking and writing; the high levels of lexical diversity and specificity that are
found in formal academic writing are apparently not possible in spoken texts due
to the restrictions of on-line production (Chafe and Danielewicz 1986; Biber
1986a). Type/token ratio (the number of different words per text) was a favorite
measure of psychologists and researchers in communication studying linguistic
differences between speech and writing (Osgood 1960; Drieman 1962; Horowitz
and Newman 1964; Gibson et al. 1966; Preston and Gardner 1967; Blankenship
1974). Longer words also convey more specific, specialized meanings than
shorter ones; Zipf (1949) has shown that words become shorter as they are more
frequently used and more general in meaning. Osgood, Drieman and Blankenship
include measures of word length in their studies. These two features are found to
co-occur frequently in planned written genres by Biber (1986a), and this
distributional pattern is interpreted as marking a highly exact presentation of
information, conveying maximum content in the fewest words.
43.type/token ratio
the number of different lexical items in a text, as a percentage
381
 This feature is computed by counting the number of different lexical
items that occur in the first 400 words of each text, and then dividing by four;
texts shorter than 400 words are not included in the present analysis. In a
preliminary version of the computer programs used here, I computed this feature
by counting the number of different lexical items in a text, dividing by the total
number of words in the text, and then multiplying by 100. If the texts in the
analysis were all nearly the same length, these two methods of computing
type/token ratio would give nearly equivalent results. If text length varies widely,
however, these two methods will give quite different results, because the relation
between the number of 'types' (different lexical items) and the total number of
words in a text is not linear. That is, a large number of the different words used
in the first 100 words of a text will be repeated in each successive 100-word
chunk of text. The result is that each additional 100 words of text adds fewer and
fewer additional types. In a comparison of very short texts and very long texts,
the type/token ratio computed over the entire text will thus appear to be much
higher in the short texts than in the long texts. To avoid this skewing, the present
study computes the number of types in the first 400 words of each text,
regardless of the total text length.
44.word length
mean length of the words in a text, in orthographic letters

(K)LEXICAL CLASSES (nos. 45-51)

45.conjuncts
alternatively, altogether, consequently, conversely, eg, e.g., else, furthermore,
hence, however, i.e., instead, likewise, moreover, namely, nevertheless,
nonetheless, notwithstanding, otherwise, rather, similarly, therefore, thus, viz.
in+comparison/contrast/particular/addition/conclusion/consequence/sum
/summary/any event/any case/other words
for + example/instance
by + contrast/comparison
as a + result/consequence
on the + contrary/other hand
ALL-P/T# + that is/else/altogether + T#/,
ALL-P/T# + rather + T#/,/xxx
(where xxx is NOT: ADJ/ADV)
Conjuncts explicitly mark logical relations between clauses, and as such
they are important in discourse with a highly informational focus. Quirk et al.
(1985:634-6) list the following functional classes of conjuncts: listing,
summative, appositive, resultive, inferential, contrastive, and transitional.
Despite their importance in marking logical relations, few register comparisons
have analysed the distribution of conjuncts. Ochs (1979) notes that they are more
formal and therefore more common in planned discourse than unplanned. Biber
(1986a) finds that they co-occur frequently with prepositions, passives, and
nominalizations in highly informational genres such as academic prose, official
documents, and professional letters. Altenberg (1986) looks at concessive and
antithetic conjuncts and finds that they are generally more common in writing
than speech.
382
Appendix A Biber’s Algorithms and Functions
46.downtoners
almost, barely, hardly, merely, mildly, nearly, only, partially, partly, practically,
scarcely, slightly, somewhat
Downtoners 'have a general lowering effect on the force of the verb'
(Quirk et al., 1985:597-602). Chafe and Danielewicz (1986) characterize these
forms as 'academic hedges', since they are commonly used in academic writing to
indicate probability. Chafe (1985) notes that downtoners are among those
evidentials used to indicate reliability. Holmes (1984) notes that these forms can
mark politeness or deference towards the addressee in addition to marking
uncertainty towards a proposition.
47.hedges
at about/something like/more or less/almost/maybe/xxx sort of/xxx kind of
(where xxx is NOT: DET/ADJ/POSSPRO/WHO - excludes sort and kind as true
nouns)
Hedges are informal, less specific markers of probability or uncertainty.
Downtoners give some indication of the degree of uncertainty; hedges simply
mark a proposition as uncertain. Chafe (1982) discusses the use of these forms to
mark fuzziness in involved discourse, and Chafe and Danielewicz (1986) state
that the use of hedges in conversational discourse indicates an awareness of the
limited word choice that is possible under the production restrictions of speech.
Biber (1986a) finds hedges co-occurring with interactive features (e.g., first and
second person pronouns and questions) and with other features marking reduced
or generalized lexical content (e.g., general emphatics, pronoun it, contractions).
48.amplifiers
absolutely, altogether, completely, enormously, entirely, extremely, fully, greatly,
highly, intensely, perfectly, strongly, thoroughly, totally, utterly, very
Amplifiers have the opposite effect of downtoners, boosting the force of
the verb (Quirk et al. 1985:590-7). They are used to indicate, in positive terms,
the reliability of propositions (Chafe 1985). Holmes (1984) notes that, similar to
downtoners, amplifiers can be used for non-propositional functions; in particular,
they can signal solidarity with the listener in addition to marking certainty or
conviction towards the proposition.
49.emphatics
for sure/a lot/such a/real + ADJ/so + ADJ/DO + V/just/really/most/more
The relation between emphatics and amplifiers is similar to that between hedges
and downtoners: emphatics simply mark the presence (versus absence) of
certainty while amplifiers indicate the degree of certainty towards a proposition.
Emphatics are characteristic of informal, colloquial discourse, marking
involvement with the topic (Chafe 1982,1985). As noted above, Biber(1986a)
finds emphatics and hedges co-occurring frequently in the conversational genres.
Labov (1984) discusses forms of this type under the label of 'intensity': the
‘emotional expression of social orientation toward the linguistic proposition'.
Other studies of emphatics include Stenstrom's (1986) analysis of really and
Aijmer's (1985) analysis of just.
50.discourse particles
CL-P/T# + well/now/anyway/anyhow/anyways

383
 Discourse particles are used to maintain conversational coherence
(Schiffrin 1982, 1985a). Chafe (1982, 1985) describes their role as 'monitoring
the information flow' in involved discourse. They are very generalized in their
functions and rare outside of the conversational genres.
51.demonstratives that/this/these/those
(This count excludes demonstrative pronouns (no. 10) and that as relative,
complementizer, or subordinator.)
Demonstratives are used for both text-internal deixis (Kurzon 1985) and
for exophoric, text-external, reference. They are an important device for marking
referential cohesion in a text (Halliday and Hasan 1976). Ochs (1979) notes that
demonstratives are preferred to articles in unplanned discourse.

(L)MODALS (nos. 52-4)

It is possible to distinguish three functional classes of modals: (1) those marking
permission, possibility, or ability; (2) those marking obligation or necessity; and
(3) those marking volition or prediction (Quirk et al. 1985:219-3 6; Coates 1983;
Hermeren 1986). Tottie (1985; Tottie and Overgaard 1984) discusses particular
aspects of modal usage, including the negation of necessity modals and the use of
would. Chafe (1985) includes possibility modals among the evidentials that mark
reliability, and necessity modals among those evidentials that mark some aspect
of the reasoning process.
52.possibility modals
can/may/might/could (+ contractions)
53.necessity modals
ought/should/must (+ contractions)
54.predictive modals
will/would/shall (+ contractions)

(M) SPECIALIZED VERB CLASSES (nos. 55-8)

Certain restricted classes of verbs can be identified as having specific functions.
Several researchers refer to 'verbs of cognition', those verbs that refer to mental
activities (Carroll 1960; Weber 1985). Chafe (1985) discusses the use of 'sensory'
verbs (e.g., see, hear, feel) to mark knowledge from a particular kind of evidence.
In the present analysis, I distinguish four specialized classes of verbs: public,
private, suasive, and seem/appear. Public verbs involve actions that can be
observed publicly; they are primarily speech act verbs, such as say and explain,
and they are commonly used to introduce indirect statements. Private verbs
express intellectual states (e.g., believe) or nonobservable intellectual acts (e.g.,
discover); this class corresponds to the 'verbs of cognition' used in other studies.
Suasive verbs imply intentions to bring about some change in the future (e.g.,
command, stipulate). All present and past tense forms of these verbs are included
in the counts.
55.public verbs
(e.g., acknowledge, admit, agree, assert, claim, complain, declare, deny, explain,
hint, insist, mention, proclaim, promise, protest, remark, reply, report, say,
suggest, swear, write)
This class of verbs is taken from Quirk et al. (1985:1180-1).
384
Appendix A Biber’s Algorithms and Functions

56.private verbs
(e.g., anticipate, assume, believe, conclude, decide, demonstrate, determine,
discover, doubt, estimate, fear, feel, find, forget, guess, hear, hope, imagine,
imply, indicate, infer, know, learn, mean, notice, prove, realize, recognize,
remember, reveal, see, show, suppose, think, understand)
This class of verbs is taken from Quirk et al. (1985:1181-2).
57.suasive verbs
(e.g., agree, arrange, ask, beg, command, decide, demand, grant, insist, instruct,
ordain, pledge, pronounce, propose, recommend, request, stipulate, suggest,
urge)
This class of verbs is taken from Quirk et al. (1985:1182-3).
58.seem/appear
These are 'perception' verbs (Quirk et al. 1985:1033, 1183). They can be used to
mark evidentiality with respect to the reasoning process (Chafe 1985), and they
represent another strategy used for academic hedging (see the discussion of
downtoners - no. 46).

(N) REDUCED FORMS AND DISPREFERRED STRUCTURES (nos.59-63)

Several linguistic constructions, such as contractions, stranded prepositions, and
split infinitives, are dispreferred in edited writing. Linguists typically disregard
the prescriptions against these constructions as arbitrary. Finegan (1980, 1987;
Finegan and Biber 1986a), however, shows that grammatical prescriptions tend
to be systematic if considered from a strictly linguistic point of view: they tend to
disprefer those constructions that involve a mismatch between surface form and
underlying representation, resulting in either a reduced surface form (due to
contraction or deletion) or a weakened isomorphism between form and meaning
(e.g., split infinitives). Biber (1986a) finds that these features tend to co-occur
frequently with interactive features (e.g., first and second person pronouns) and
with certain types of subordination. Chafe (1984b) discusses the linguistic form
of grammatical prescriptions and analyzes the historical evolution of certain
prescriptions in speech and writing. Features 59-63 are all dispreferred in edited
writing; nos. 59-60 involve surface reduction of form and nos. 61-3 involve a
weakened isomorphism between form and meaning.
59.contractions
(1) all contractions on pronouns
(2) all contractions on auxiliary forms (negation)
(3) 's suffixed on nouns is analyzed separately (to exclude possessive forms):
N's + V/AUX/ADV+V/ADV+AUX/DET/POSSPRO/PREP/ADJ+CL-P/ADJ+T#
Contractions are the most frequently cited example of reduced surface
form. Except for certain types of fiction, they are dispreferred in formal, edited
writing; linguists have traditionally explained their frequent use in conversation
as being a consequence of fast and easy production. Finegan and Biber (1986a),
however, find that contractions are distributed as a cline: used most frequently in
conversation; least frequently in academic prose; and with intermediate
frequencies in broadcast, public speeches, and press reportage. Biber (1987) finds
that contractions are more frequent in American writing than in British writing,
385
apparently because of greater attention to grammatical prescriptions by British
writers. Chafe and Danielewicz (1986) also find that there is no absolute
difference between speech and writing in the use of contractions. Thus, the use of
contractions seems to be tied to appropriateness considerations as much as to the
differing production circumstances of speech and writing.
60.subordinator-that deletion
(e.g., I think [that] he went to ... )
(1) PUB/PRV/SUA+ (T#) +demonstrative pro/SUBJPRO
(2) PUB/PRV/SUA+PRO/N+AUX/V
(3) PUB/PRV/SUA + ADJ/ADV/DET/POSSPRO + (ADJ) + N +AUX/V
While contractions are a form of phonological (or orthographic)
reduction, subordinator- that deletion is a form of syntactic reduction. There are
very few of these deletions in edited writing, even though few explicit
prescriptions prohibit this form. Apparently the concern for elaborated and
explicit expression in typical edited writing is the driving force preventing this
reduction.
61.stranded prepositions
(e.g. the candidate that I was thinking of)
PREP + ALL-P/T#
Stranded prepositions represent a mismatch between surface and
underlying representations, since the relative pronoun and the preposition belong
to the same phrase in underlying structure. Chafe (1985) cites these forms as an
example of spoken 'errors' due to the production constraints of speech.
62.split infinitives
(e.g., he wants to convincingly prove that ...)
to + ADV + (ADV) + VB
Split infinitives are the most widely cited prescription against surface-
underlying mismatches. This notoriety suggests that writers would use split
infinitives if it were not for the prescriptions against them, but these forms in fact
seem to be equally uncommon in spoken and written genres (Biber 1986a; Chafe
1984b). This feature did not co-occur meaningfully with the other features
included in the present study, and it was therefore dropped from the factor
analysis(Chapter 5).
63.split auxiliaries
(e.g., they are objectively shown to ...)
AUX + ADV + (ADV) + VB
Split auxiliaries are analogous to split infinitives, but they have not
received much attention from prescriptive grammarians. They are actually more
common in certain written genres than in typical conversation; Biber (1986a)
finds that they frequently co-occur with passives, prepositions, and
nominalizations.

(O) COORDINATION (nos. 64-5)

Phrase and clause coordination have complementary functions, so that any
overall count of coordinators would be hopelessly confounded. And as a clause
coordinator is a general purpose connective that can mark many different logical
relations between two clauses. Chafe (1982, 1985) relates the fragmented style
386
Appendix A Biber’s Algorithms and Functions
resulting from this simple chaining of ideas to the production constraints of
speech. And as a phrase coordinator, on the other hand, has an integrative
function and is used for idea unit expansion (Chafe 1982, 1985; Chafe and
Danielewicz 1986). Other studies that analyse the distribution and uses of and
include Marckworth and Baker (1974), Schiffrin (1982), and Young (1985). The
algorithms used in the present study identify only those uses of and that are
clearly phrasal or clausal connectives.
64.phrasal coordination
xxxxl + and + xxxx2
(where xxxl and xxx2 are both: ADV/ADJ/V/N)
65.independent clause coordination
(a) T#/, + and + it/so/then/you/there + BE/demonstrative pronoun/SUBJPRO
(b) CL-P+ and
(c) and + WHP/WHO /adverbial subordinator (nos. 35-8)/discourse particle (no.
50)/conjunct (no. 45)

(P) NEGATION (nos. 66-7)

There is twice as much negation overall in speech as in writing, a distribution
that Tottie (1981, 1982, 1983b) attributes to the greater frequency of repetitions,
denials, rejections, questions, and mental verbs in speech. Tottie (1983a)
distinguishes between synthetic and analytic negation. Synthetic negation is more
literary, and seemingly more integrated; analytic negation is more colloquial and
seems to be more fragmented.
66.synthetic negation
(a) no + QUANT/ADJ/N
(b) neither, nor
(excludes no as a response)
67. analytic negation: not
(also contracted forms)

387
 Appendix B

Alphabetical Frequency List

b 12320/15281 belief 408/505 C bodied 108/124
bachelor 110/145 beliefs 270/340 bodies 543/764 CP
back 1369/1910 P believe 346/409 CP body 1807/3647 CP
backed 205/241 believed 009/1222 C boh 122/123
background 347/441 bell 313/491 P boiling 145/195 C
backward 108/122 P belong 295/321 bold 194/228
bad 145/159 P belonged 111/112 bomb 109/252 C
bah 132/135 belonging 305/317 bond 164/321
balance 406/582 CP belongs 147/163 bonds 157/367
balanced 141/161 CP below 764/997 CP bone 286/613 P
ban 101/115 belt 183/305 bones 239/516 P
banned 167/187 beneath 257/331 P book 1558/2287 CP
bark 199/348 benefit 143/178 P books 1088/1517 P
barley 130/174 benefits 172/256 boom 106/140
basin 343/668 C bent 100/111 border 563/819
basins 132/285 berries 109/165 bordered 202/209
basis 968/1216 CP besides 159/176 borders 222/270
bass 139/321 best 2786/3536 CP bore 174/209
BC 1692/4389 better 570/687 C born 1391/1622
beach 176/289 beyond 644/785 borne 148/165
beaches 116/161 big 358/517 P borrowed 125/149
beam 173/393 bill 500/825 both 4356/7133 CP
beams 101/194 billion 534/1319 CP bottom 414/749 CP
beans 101/194 bills 154/224 bought 134/147
bear 428/589 binding bound 231/273
bearing 312/405 C 106/145 boundaries 235/300
bears 341/536 CP bird 417/729 boundary 232/316
beat 101/140 birds 605/1225 bounded 187/199
beautiful 305/358 birth 547/838 bow 120/173
becomes 566/733 CP births 217/256 box 189/262 CP
becoming 655/703 bitter 234/268 C boy 193/212 P
bed 141/195 black 1921/3783 C brain 347/975 C
beds 301/363 blacks 273/600 branch 499/662 C
bee 175/269 P blend 122/135 branches 418/611 P
beer 106/156 blind 137/206 C brass 108/170 P
began 3077/4682 CP block 257/341 C break 388/440 CP
begin 340/426 blocks 177/222 P breakdown 121/141 C
beginning 1225/1582 P blood 732/2125 C breaking 208/233 P
beginnings 191/206 bloody 112/125 breathing 131/196 CP
begins 319/383 CP bloom 105/155 bred 136/235
behalf 118/127 blow 115/136 breed 245/522
behavior 640/1441 CP blue 668/1101 CP breeding 227/380
behavioral 102/167 blues 100/243 breeds 124/235
behind 540/639 P board 422/585 P brick 143/222
beings 237/336 C boat 129/248 P bridge 305/555 P

391
Appendix B Alphabetical Frequency List

bridges 148/231 P between C census 500/633

brief 386/400 bibliography centuries 1404/2018 CP
bright 352/447 but C century 5988/14669 CP
brightly 107/130 C by CP certain 1512/2184 CP
bring 428/472 CP certainly 101/110 P
bringing 208/222 chain 330/551
broad 651/802 C chains 137/235
broadcasting 127/292 c 6214/11873 chairman 272/331
broader 104/112 C cabinet 335/473 challenge 224/261
broadly 121/130 cable 124/307 P challenged 203/218
broke 392/469 C call 304/354 CP chamber 421/631 CP
broken 338/395 CP calling 150/161 chambers 167/236
brother 700/852 calls 152/171 chancellor 159/263
brothers 318/453 came 1589/2164 P change 1067/1624 CP
brought 1462/1939 CP camp 172/224 P changed 552/628 CP
brown 1062/1544 C campaign 421/690 changes 938/1508
budget 149/253 campaigns 209/278 CP
buffalo 191/275 camps 127/180 changing 427/497 CP
build 385/471 CP canal 382/788 channel 288/503
builders 112/150 C canals 162/242 channels 158/256
building 1099/1782 P cannot 629/814 CP chapel 184/288
buildings 760/1219 P capable 470/598 CP charge 404/733 C
built 1675/2645 CP capital 1798/3099 P charged 402/551 P
bulk 171/208 C car 143/290 P charges 252/369 C
bull 171/306 care 402/828 charter 161/237
bur 113/116 careful 156/164 P chartered 107/118
bureau 154/239 P carefully 272/300 C check 112/140 CP
burial 160/270 carried 873/1071 CP chemistry 547/1054 CP
buried 255/287 carries 237/288 C chest 130/181
burn 116/144 C carry 499/654 CP chief 1384/1937
burned 227/261 C carrying 321/375 CP chiefly 334/360 C
burning 229/313 C cars 106/237 P chiefs 126/164
burns 122/190 C carved 222/364 child 576/936 P
bush 193/339 cash 156/223 C childhood 237/307
business 769/1230 cast 331/440 C children 1095/2092
businesses 118/157 castle 290/392 chloride 101/208 C
buy 122/149 CP cat 143/373 choice 255/295 CP
catch 150/189 P choose 130/148 CP
unlisted (too common) cattle 467/818 P choral 138/211
be caught 181/206 P chose 193/209 C
became cave 149/336 P chosen 379/442 CP
because caves 100/189 Christmas 119/161
become ceased 129/137 chronic
been ceded 143/155 178/262
before C cell 399/1372 P church 1690/3823
begun cells 591/2105 P churches 505/901
being cement 193/240 circle 384/562 CP

392
Appendix B Alphabetical Frequency List

circles 181/217 CP code 262/567 conceived 196/209

circular 271/373 P codes 111/154 concept 670/934 CP
cities 1012/1927 coffee 193/289 conception 145/179
citizen 197/248 P coined 131/141 C concepts 305/394 P
citizens cold 580/915 CP concern 489/582 C
319/496 collected 500/556 C concerned 505/653 CP
citrus 106/193 C collecting 109/142 C concerning 317/373 P
city 4147/10255 P collection 908/1281 concerns 365/408 C
civic 114/131 collections 392/486 concrete 229/419 P
civilian collective 164/224 condemned 203/227
205/293 collectively 137/142 conduct 274/341 C
claim 338/417 P color 1153/2019 C conducted 425/489 P
claimed 437/508 P colored 537/686 C conducting 156/197 C
claims 307/406 P colorful 219/262 conductor 211/417 C
clarity 119/127 colorless 107/125 C confidence 115/135
classes 393/522 C colors 514/770 C confined 270/307
clay 346/583 column 185/310 CP connected 439/506 P
clean 119/142 C columns 165/265 CP connecting 135/152 P
clear 553/659 P combat 189/263 connection 206/230
clearly 352/409 P come 727/854 CP connections 127/146
clergy 155/194 comes 381/433 CP connects 103/114
climate 778/1307 coming conquered 411/535
climates 212/342 339/399 P conquest 403/571
climatic 152/254 commitment 154/175 conscious 134/178
clinical 127/216 committed 292/326 C consciousness 201/318 C
close 975/1162 CP committee 375/573 P contain 804/1119 CP
closed 386/477 C commodities 104/130 C contained 494/594 CP
closely 763/384 CP common 2536/4170 CP containing 697/898
closer 193/211 CP commonly 1050/1298 CP
closest 110/137 CP contains 1076/1413 CP
cloth 149/212 commonwealth 171/265 content 437/611 C
clothing 267/344 CP compact 212/271 P convicted 121/139
cloud 171/317 CP companies 400/640 conviction 119/144
clouds 143/309 C companion 162/179 convinced 165/178
cluster 102/165 company 1047/1913 cook 150/237
clusters 275/483 P cooking 103/176 C
cm 1467/2372 P comparable 182/200 CP cool 315/396
co 211/258 compared 311/358 CP cooled 127/173 C
coal 584/1341 comparison 129/148 CP cooler 112/139
coalition 275/417 compound 270/446 CP cooling
coarse 153/170 compounds 433/1089 C 151/278
coast 1183/2260 P comprise 231/262 copper 465/962 CP
coastal 522/936 comprised 106/113 core 279/457
coastline 105/123 comprises 330/381 corn 285/488
coasts 176/242 comprising 183/196 corner 131/150 P
coat 292/467 C compromise 191/263 costly 166/205 C
coated 126/176 compulsory 142/175 cotton 483/784

393
Appendix B Alphabetical Frequency List

could 1680/3136 CP called P declining 106/118

council 761/1364 can P decrease 186/247 CP
councils 137/186 caused decreased 132/157 CP
counterpart 108/111 decreases 49/175 CP
counties 154/297 deep 829/1177 C
countries 1182/2417 deeper 127/141 P
country 1685/3428 P deeply 298/326
county 878/1293 defeat 476/617
coupled 123/143 C defeated 720/967
course 610/802 CP defeating 102/112
courses 179/257 P d 10966/14425 defense 721/1049
court 1367/2955 dah 108/110 defensive 116/169
courts 336/596 daily 444/591 CP define 164/184 CP
cousin 127/140 dairy 161/231 defined 608/772 CP
cover 434/578 P dam 108/287 definition 240/315 CP
coverage 102/138 damage 456/628 C deg 1039/4147
covered 732/947 CP damaged 234/258 C degree 714/967 C
covering 333/368 dame 126/199 degrees 392/625 C
covers 432/487 P dams 113/217 delayed 114/122 P
craft 234/406 dancer 157/266 delivered 121/133
crafts 103/157 dancers 112/195 delta 180/322 P
creek 144/215 danger 147/187 C demand 372/546
crew 141/310 CP dark 630/794 C demands 284/346
crisis 427/634 darkness 105/132 demonstrate 147/151 CP
crop 304/539 C data 422/859 CP demonstrated 429/509 C
crops 507/822 C date 510/620 denied 191/218
cross 627/891 CP dated 140/165 dense 348/410 C
crossed 204/245 P dates 389/455 densely
crosses 101/119 P dating 379/510 145/167
crossing 154/174 P daughter 533/621 density 535/808 CP
crown 493/641 daughters 106/125 depict 111/128
crowned 164/206 day 2019/2908 CP depicted 252/315
crude 152/191 P days 932/1297 P depicting 151/164
crushed 143/158 P dead 429/621 CP depicts 119/132
crust 165/338 deal 468/613 CP deposed 162/188
cup 128/206 C dealing 223/246 CP deposit 133/190 C
currency 273/362 deals 207/248 CP deposited 190/252
custom 116/133 dealt 189/215 deposition 106/151
customs 246/319 death 2634/3643 deposits 602/1141 C
cut 606/862 deaths 154/197 depression 364/537 C
cuts 107/145 P debt 208/303 depth 350/496 CP
cutting 225/354 Dec 1653/1798 depths 178/249 C
cylinder 131/309 CP decay 153/305 deputy 123/387
cylindrical 141/167 December 499/649 derive 101/108 CP
deciduous 182/246 derived 1246/1545 CP
decline 510/648 derives 142/144
unlisted (too common) declined 511/585 descendants 250/287

394
Appendix B Alphabetical Frequency List

descended 171/195 devoted 482/525 disputes 203/254

descent 287/348 diagnosis 107/160 dissolved 262/350 C
describe 382/439 CP dialect 129/164 distinguished 781/866
described 676/797 CP dialogue 173/225 CP
describes 254/273 diameter 410/592 P disturbances 111/141
CP diamond 115/189 divine 345/495
describing 146/153 CP dictator 107/152 DNA 106/479
description 240/290 CP did 1632/2368 CP do 1274/1843 CP
descriptions 106/115 P die 294/408 documentary 124/156
descriptive 106/126 died 1283/1561 does 912/1167 CP
desert 386/759 dies 109/139 dog 274/737 P
deserts 119/204 diet 242/354 dogs 195/374 P
design 859/1592 P differ 297/342 CP doh (pronun) 145/147
designated 272/309 difference 299/401 CP doing 151/168 P
designed 1201/1687 differences 462/644 CP dollar 102/148 C
designer 189/241 differing 119/126 domain
designs differs 186/204 CP 116/161 P
413/605 difficult 689/806 CP dome 133/248
desire 241/293 C difficulties 225/251 domestic 573/899 C
desired 203/260 C difficulty 198/219 dominance 135/161
despite 931/1171 C digestive 116/201 done 467/569 C
destroy 173/215 C dimensional 192/286 CP door 102/144
destroyed 670/823 C diminished 111/117 P doors 119/149
destruction 327/399 C dioxide double 487/736 CP
destructive 113/130 219/421 C doubt 117/132 P
detail 318/385 C direct 823/1094 CP down 1287/1741 CP
detailed 266/291 CP directed 570/708 downward 149/166 P
details 235/274 CP disappeared 196/219 dozen 147/161 P
detect 160/207 P disaster 116/144 Dr 180/234
detected 157/212 disastrous 119/127 draft 150/248
detection 116/156 discharge 124/234 C drainage 234/373
determination 187/233 discovered1207/1599 CP drained
determine 462/593 CP discoveries 218/270 P 177/208
determined 682/896 CP discovery 661/894 P drains 128/151 C
determines 151/172 discussed 250/168 CP draw 162/183 CP
determining 223/262 CP discussion 142/179 CP drawing 283/394 CP
devastating 130/133 disease 704/1738 drawings 287/425
develop 829/1083 CP diseases 512/1201 drawn 503/601 P
developing 579/723 C disk 121/201 P dream 209/264
developments disorder 188/347 dreams 120/188
337/418CP disorders 240/517 drew 250/287 P
develops 210/256 dispersed 130/152 dried 220/338 C
device 420/595 CP display 296/367 C drift 114/206
devices displayed 210/229 drinking 112/175 C
518/837 displays 185/220 drive 251/352 CP
devil 128/146 dispute 221/258 driven 293/368 C
devised 268/300 CP disputed 117/134 driving 145/180 C

395
Appendix B Alphabetical Frequency List

drop 187/224 CP earthquake 108/165 employed 580/702

dropped 177/214 P earthquakes 107/157 P employees 129/213
drought 122/175 ease 124/132 C employing 103/114
drove 132/152 easier 114/124 C employment 243/406
drug 337/868 easily 530/629 CP employs 141/155
drugs 342/890 C east 2106/4017 P empty 141/153 C
dry 613/945 eastern 1651/2954 enable 157/171 CP
duchy 109/144 eastward 157/194 P enabled 267/299
due 606/806 CP easy 222/254 CP enables 107/116 CP
duh (pronun) 263/276 eat 213/274 CP enabling 109/118
dur (pronun) 101/103 eating 192/243 P enacted 157/187
durable ecology 134/216 enclosed 218/244
102/112 economic 1520/3406 encompasses 119/121
duration 135/176 P economically 239/278 encountered 148/160 C
during 6721/12543 CP economics 301/440 encourage 159/172
dust 180/334 P economy 1053/2114 encouraged 470/524
duties 172/246 edge 323/420 P end 1961/3027 CP
duty 164/204 edges 137/160 ended 634/791
dwarf 102/136 edible 148/265 ending 222/238
dwellers 107/125 edited 229/255 P ends 284/333 CP
dwelling 154/196 eds 1545/1859 enduring 106/117
dying 109/141 ee (pronun) 846/861 enemies 229/282
dynamic 174/202 effort 555/673 C enemy 240/360
dynamics 121/164 P efforts 829/1086 CP energy 1011/3301 CP
egg 225/496 engage 108/119
eggs 414/795 engaged 306/330
unlisted (too common) eight 496/567 CP engine 251/823 CP
developed P either 1457/1961 CP engines 181/395 CP
development P elaborate 436/530 enhance 105/111
different P elder 205/248 enhanced 150/155 P
electric 527/1147 CP enjoyed 282/304
electricity 281/519 CP enlarged 216/249 P
elongated 186/222 enlightenment 134/207
else 120/128 P enormous 494/581 CP
elsewhere 356/414 enormously 105/107
emerge 157/182 enough 561/691 CP
emerged 471/604 enrollment 363/904
emergence 195/230 ensuing 128/134
emergency 151/210 P ensure 189/207
e 3676/5155 emerging 140/151 enter 314/355 CP
ear 127/279 emperor 839/1610 entered 770/859
earl 286/460 emperors 137/199 entering 216/230
earlier 1046/1280 CP emphasis 472/590 enterprise 174/206
earliest 886/1132 C emphasize 111/140 enterprises 114/131
earned 414/450 emphasized 288/320 P enters 192/225 P
ears 223/304 empire 1098/2330 P entertainment 105/154
earth 1209/3113 CP employ 187/209 CP entrance 201/233

396
Appendix B Alphabetical Frequency List

entry 198/237 expressed 441/526 CP farther 172/223

environment 607/945 CP eye 397/660 P fashion 251/379
environmental 448/759 eyes 385/586 fast 320/407 CP
environments 180/248 faster 165/198 CP
epic 306/469 unlisted (too common) fat 174/359
equal 621/908 CP each fate 112/145 P
equality 138/199 P early favor 455/512
equally 351/398 ed favorable 157/170
equals 103/124 CP est favored 342/384
era 786/1163 P established CP favorite 261/285
erect 117/129 even P fear 219/274
erected 173/219 every P feared 132/148
error 105/170 CP feast 260/283
erupted 113/132 feathers 128/264
essay 186/238 P feature 458/524
essayist 119/152 featured 116/127
essays 630/836 features 659/902 C
essence 115/129 Feb 1657/1782
estate 258/334 February 377/467
estimated 475/588 fed 215/266 P
evening 143/178 f 3325/5440 fee 138/160
event 326/402 fabric 129/221 feed 494/678 C
events 667/982 fabrics 104/167 feeding 223/306 C
eventual 120/130 face 518/672 P feeds 137/179
eventually 1091/1371 CP faced 294/343 P feel 118/145 CP
ever 562/679 CP faces 160/208 P feeling 190/223
evergreen 193/327 facilities 445/663 feelings 143/181
everyday 170/197 P facing 158/176 feet 324/444 CP
everything 129/152 factories 196/247 fell 429/503 CP
everywhere 112/127 factory 218/333 fellow 201/214
evidence 687/1079 C fail 116/148 P felt 303/365
evil 186/240 failed 568/698 P female 673/1254 C
excavated 149/183 failure 413/522 females 243/375
excavations 155/205 fair 257/323 C fever 219/387
excelled 111/116 fairly 257/287 C few 1910/2757 CP
exchange 406/574 faith 373/519 fewer 219/253
exerted 172/199 C fall 863/1116 P fiber 172/370
expelled 212/246 fallen 108/114 P fibers 230/528 P
expert 102/127 P falling 179/208 P field 1231/2356 CP
experts 157/169 falls 376/564 CP fields 672/972 P
explain 225/272 CP familiar 368/420 CP fifth 300/377
explained 189/235 CP far 1259/1738 CP fight 224/266
explanation 146/175 P farm 386/645 fighting 424/606
exploitation 133/149 farmer 142/182 P figure 815/1066 CP
exploration 340/484 farmers 349/550 C figures 796/1275 CP
exposure 229/383 CP farming 380/575 fill 122/138 P
express 228/293 CP farms 199/360 C filled 418/497 CP

397
Appendix B Alphabetical Frequency List

fin 112/201 focused 232/274 fuels 107/213 C

find 439/513 CP folk 374/781 full 948/1175 C
finding 180/212 P folklore 142/179 fully 458/513 CP
findings 107/119 C follow 359/393 CP fur 309/558
finds 132/159 CP followed 1346/1687 CP furniture 298/580
fine 827/1064 followers 282/329 further 1107/1424 CP
finely 110/125 following 1754/2200 CP furthermore 118/135 CP
finest 493/550 follows 275/298 CP fused 115/150
finished 204/238 food 1435/3002 CP
fins 118/209 foods 280/505
fire 619/989 foodstuffs 122/130 unlisted (too common)
fired 151/235 foot 276/427 CP for P
firm 419/535 forbidden 102/127 from CP
firmly 163/173 foreign 1169/2185
firms 146/246 foremost 292/327
first 8401/15779 CP form 3520/6359 CP
fish 705/1720 CP former 986/1291
fishes 186/467 formerly 553/624
fishing 621/953 forms 2034/3334 P
fit 130/148 CP fort 444/767
fitted 124/152 forth 276/328 CP
five 1417/1849 CP fortress 149/197 g 2864/3937 C
flag 106/230 forward 291/393 P gain 317/370 C
flat 532/692 P fostered 113/115 gained 814/978 C
flattened 161/175 fought 545/653 gaining 147/154
flavor 165/302 C found 3353/5597 P gains 120/139
fled 333/368 four 2109/3050 P game 438/1078 P
fleet 230/446 fourth 505/601 P games 316/834
flesh 150/187 P fox 226/357 gamma 118/224
fleshy 100/137 fragrant 107/173 gap 100/145
flew 102/145 frame 181/314 P garden 391/661
flight 398/930 CP framework 193/223 CP gardens 277/436
floating 135/163 P free 1493/2205 CP gasoline 100/193 CP
flood 149/240 freed 155/164 gate 110/176
floor 268/408 CP freedom 603/961 gates 125/171
flour 108/170 freely 179/192 CP gathered 167/181 C
flourished 386/434 freezing 123/177 C gathering 153/188
flow 548/1012 fresh 308/395 C gave 1039/1273 C
flower 333/629 freshwater 216/364 GDP 155/171
flowering 222/341 friend 358/379 P gene 168/465
flowers friends 283/346 genera 223/301
706/1580 friendship 134/148 general 2464/4043 CP
flowing 228/290 front 554/965 CP generate 139/164
flows 420/576 frost 104/183 generated 293/349 C
fly 164/292 P frozen 146/204 P generating 117/153
flying 217/352 CP ft 2511/5362 CP genus 800/1050
focus 381/480 C fuel 285/696 CP get 187/212 CP

398
Appendix B Alphabetical Frequency List

giant 325/447 P green 873/1399 CP heads 308/407 P

gift 134/150 grew 763/897 health 870/1694 C
gifted 109/119 C gross 159/212 heard 190/214 P
gifts 101/113 ground 916/1428 hearing 127/203
girl 161/192 grounds 208/250 heart 614/1311 P
girls 106/147 grow 614/914 heat 581/1564 CP
give 690/821 CP growing 912/1278 heated 221/307 C
given 1607/2132 CP grown 768/1156 heating 188/297 C
gives 373/427 CP grows 439/698 heaven 143/186
giving 450/517 C growth 1363/2541 CP heavier 181/224 C
gland 169/330 guh (pronun) 109/109 heavily 521/597
glass 592/1282 C guidance 117/170 heavy 899/1213 CP
go 369/449 CP gun 124/281 P height 563/745 CP
goats 112/138 guns 104/228 heights 202/247 CP
goddess 163/228 heir 143/171
gods 281/471 held 1788/2352 CP
goes 193/239 CP unlisted (too common) helium 108/258 CP
goh (pronun) 107/111 generally P help 673/807 CP
going 180/188 P got helped 1160/1378
gold 779/1502 CP helping
golden 483/648 140/150
gone 125/135 h 3881/5697 helps 126/141 C
good 961/1221 CP ha 198/290 C hence 322/372
goods 565/885 habits 120/141 herb 141/274
grade 149/224 hair 278/485 herbs 150/207
graduate 378/479 half 1369/2036 CP herds 111/154
graduated 277/305 hall 622/855 P here 583/730
graduating 141/143 hand 906/1322 CP hereditary 189/262
grain 344/577 C handbook 430/467 C heritage 318/357
grains 262/424 C handle 135/159 herself 145/160
grammar 113/193 handled 122/127 hidden 147/158
grandson 144/165 handling 189/218 C high 3145/5679
grant 410/703 hands 408/535 CP higher 1066/1890
granted427/510 harbor 336/491 highest 811/1293
grants 142/186 hard 616/809 CP highlands 188/376
grapes 122/170 hardness 160/194 highly 1494/1919
graphic 129/188 harmful 103/123 highway 171/243
grass 213/362 P harsh 193/209 C highways 126/160
grasses 116/196 harvest 111/138 hill 401/540
grasslands 109/168 harvested 116/177 hills 336/462
grave 113/139 hatch 111/134 hind 142/214
gravel 121/172 haven 108/140 hit 170/229
gray 565/737 having 1060/1283 CP hold 447/529
great 3871/6979 CP hay 151/196 holding
greater 1093/1475 CP head 1300/1864 P 289/333
greatest 1195/1575 CP headed 458/539 P holdings 122/155
greatly 825/967 C headquarters 349/386 holds 257/284 P

399
Appendix B Alphabetical Frequency List

hole 144/236 CP interrupted 113/116

holes 152/251 C intricate 120/125
hollow 167/201 P invited 114/114 C
home 1162/1671 involve
homeland 140/170 318/391 C
homes 212/277 involved 919/1160 CP
honor 340/387 involvement 194/227
hope 286/372 C involves 428/549 CP
hoped 149/179 I 3595/5925 C involving 391/444 CP
hopes 101/111 ice 426/1086 CP iron 969/2001 CP
horn 151/247 ill 391/447 irrigated 105/152
horse 339/760 P illness 184/250 irrigation 229/436
horses 245/412 improve 307/359 C island 1164/2570 P
host 210/366 improved 471/589 islands 794/2147
hot 542/806 CP improvement 177/207 C isotopes 111/207 C
hour 228/299 CP improvements 196/234 issue 372/499
hours 426/590 CP improving 125/136 P issued 229/285
house 1452/2544 incident 131/174 issues 445/596
housed 142/155 P income 395/1041 C itself 1104/1434 CP
household 171/218 C increase 748/1122 CP ivory 161/317
houses 578/843 increased 1074/1658 CP
housing 265/516 increases 456/586 CP
how 894/1310 CP increasing 772/1027 CP unlisted (too common)
however 4654/8416 CP increasingly 785/985 if P
huge 537/663 C indeed 187/213 CP in P
hundred 424/489 C indigenous 236/297 include CP
hundreds 373/430 C inexpensive 117/234 included P
hunt 232/245 inherent 106/117 includes CP
hunters 164/203 inheritance 125/182 including CP
hunting 398/616 inherited 237/276 into P
husband 348/435 injury 170/254 is CP
hy (pronun) 102/103 inland 309/411 it CP
hydroelectric 243/354 inner 466/655 P its P
hydrogen 356/918 CP input 102/226 P
inquiry 103/131
inside 448/546 CP
unlisted (too common) insight 102/113 C
had CP instance 299/382 C
has P instances 125/141
have CP instead 689/839 C
he CP insurance 256/588
her P intended 422/493
him P interest 1192/1660 CP
himself interested 289/321 C j 5518/8052
his CP interesting 103/111 January 619/744 CP
CP Jan 1963/2274
interests 421/551 jaws 108/157

400
Appendix B Alphabetical Frequency List

jay 191/252 knight 135/188 language 1335/3102 P

jee (pronun) 103/105 knighted 154/159 languages 592/1667
jet 145/286 CP knights 105/189 large 3715/6916 CP
job 193/283 P know 225/281 CP largely 1083/1357 CP
jobs 211/297 knowledge 698/1157 CP larger 1067/1460 CP
join 272/301 koh (pronun) 251/256 largest 1926/3328 CP
joined 1298/1486 kuh (pronun) 198/198 larvae 119/303
joining 188/195 P last 1791/2323 C
joint 302/454 lasted 221/239
journal 304/381 unlisted (too common) lasting 243/261
journals 120/141 known CP late 2417/3461 C
journey 229/254 latter 930/1073 CP
judge 204/291 launch 125/288 P
judges 120/188 launched 437/799 CP
judgement 185/250 law 1918/4274 CP
July 2322/2848 laws 801/1488 CP
June 2035/2473 lawyer 302/349
just 765/934 CP lay 534/610 P
layer 334/627 CP
layers 318/469 C
lb 673/1107 P
lead 789/1115 CP
leader 1371/1940
leaders 673/994
leadership 516/748
k 1780/2259 l 3450/4540 C leading 1766/2329 C
kah (pronun) 244/252 labor 897/1822 leads 258/309 C
kee (pronun) 162/166 laboratories 145/169 P leaf 202/366
keep 387/455 CP laboratory 417/613 CP league 535/1071 P
keeping 177/198 P lack 553/700 P learn 162/214 C
kept 433/502 C lacked 134/140 P learned 401/481 C
key 505/649 C lacking 153/169 learning 440/823
keyboard 106/226 lady 274/356 least 943/1236 CP
kg 658/1083 CP lah (pronun) 189/194 leather 236/312
kidney 121/252 P laid 460/523 CP leave 270/292 C
kill 165/207 lake 885/2526 CP leaves 815/1697 P
killed 613/748 C lakes 454/932 C leaving 357/389 P
killing 183/206 land 1711/4004 P lectures 128/141
kind 519/624 CP landed 162/209 left 1590/2336 CP
kinds 509/625 landing leg 109/181 P
king 2070/4107 162/313 P legend 351/433
kingdom 891/1524 P landmark 129/130 legends 138/165
kingdoms 177/281 P landmarks 213/227 legs 388/632
kings 474/697 lands 446/689 P length 1133/1573 P
kinship 101/157 landscape 416/727 lengths 118/132 P
km 2847/6227 CP landscapes 232/296 lengthy 108/117 C
knew 116/133 P lane 102/125 less 1976/2963 CP

401
Appendix B Alphabetical Frequency List

lesser 263/311 loose 181/202 makes 581/675 CP

let 113/123 CP loosely 129/135 making 1400/1753 CP
letter 270/472 P lose 158/175 C male 709/1446 C
letters 628/873 P losing 142/151 males 306/455
level 1202/1911 CP loss 558/804 CP mammals 306/607 P
levels 558/830 C losses 195/280 P man 1771/2436 P
libraries 101/168 lost 1149/1425 CP managed 215/246
library 828/1929 P love 771/1076 management 369/635
lie 304/371 low 1488/2344 CP manager 119/159
lies 795/910 CP lower 1247/1818 C manganese 136/233 C
life 4668/8121 CP lowered 107/131 Mar 1761/1929
lifelong lowest 406/522 CP marble 202/305 P
113/117 lowland 128/194 March 607/852 P
lifetime 266/309 CP lowlands 155/267 market 572/906
light 1598/3698 CP loyal 101/108 marketing 127/197
lighter 218/257 C loyalty 125/133 markets 218/332
lighting 137/195 luh (pronun) 163/165 marks 218/250 C
like 1816/2672 CP lumber 203/309 marriage 563/856
likely 308/393 CP lung 110/230 married 687/802
limbs 117/211 P lungs 146/318 C marry 112/122
limestone 268/382 C luster 114/129 marshes 105/171
lined 134/158 lying 242/268 P masks 103/158
lines 756/1282 CP massive 531/668 P
lining 106/140 unlisted (too common) master 680/885 CP
link 274/303 later CP masters 311/387
linked 432/503 led P mastery 118/133
linking 108/121 match 115/140 C
links 166/189 m 6223/11357 CP mate 127/160
literacy 237/328 machine 493/984 mating 118/172
little 1536/2087 CP machinery 496/607 matter 743/1250 CP
live 1141/1679 CP machines 263/516 matters 259/317 P
lived 1062/1290 P magazine 316/448 mature 313/389
lively 146/153 magazines 117/199 maturity 148/165
liver 158/305 mah (pronun) 204/212 May 6133/12728 C
lives 680/925 mahn (pronun) 142/142 mayor 161/228
livestock 323/426 main 1385/1878 CP me 177/212
living 1168/1654 P mainland 212/339 mean 423/574 CP
load 103/177 P mainly 967/1256 CP meaning 751/1027 CP
located 1930/2735 P maintain 536/666 CP means 1317/1796 CP
loh (pronun) 152/155 maintained 544/631 CP meant 178/210 P
long 4041/6701 CP maintaining 217/241 P meanwhile 164/232
longer 737/920 CP maintains 131/144 P measure 493/646 CP
longest 231/262 maintenance 165/213 measured 334/503CP
loo(pronun) 107/111 maize 129/161 measurement222/387 CP
look 284/320 CP major 4083/7478 CP measurements186/316C
looked 106/119 majority 661/1039 C P
looking 151/164 P makers 109/137

402
Appendix B Alphabetical Frequency List

measures 431/548 motions 114/256 CP nephew 135/141

CP mountain 767/1438 P nerve 180/453
measuring 242/322 CP mountainous 300/346 P nerves 100/167
meat 265/457 mountains 889/1707 P nervous 270/491
medium 455/693 CP mounted 279/422 nest 165/351
mee (pronun) 191/192 mouth 472/689 nests 123/191
meet 356/402 P movies 120/173 net 161/243 CP
meeting 236/270 mph 184/352 P network 345/499 C
melting mud 158/248 never 926/1089 CP
184/271 C muh(pronun) 174/175 nevertheless 579/581 CP
men 1410/2000 CP murder 229/284 new 6738/15987 CP
merged 179/188 murdered 151/167 newer 114/117
message 152/217 my 559/685 P newly 394/437
met 541/622 news 285/608
meters 152/212 C newspaper 303/452
mi 2747/5872 P unlisted (too common) newspapers 216/341
mid 1185/1612 P made C next 1011/1258 CP
middle 1644/2680 P make CP nickel 147/226 C
might 603/842 CP many CP nickname 146/159
mild 316/357 C most CP night 623/815 CP
miles 193/237 CP much CP nine 387/453 C
milk 174/504 must CP nineteenth 126/134 C
mill 149/250 P nitrogen 202/436 C
mills 195/289 noh (pronun) 186/192
mind 434/636 CP non 442/599
mirror 161/286 none 324/392
mixed 442/544 C nontheless 192/214
mixture 338/463 C nor 373/461 CP
mode 246/322 north 3299/7340 P
modes 123/165 northeast 453/597 P
moh (pronun) 147/147 northeastern 453/514
moist 189/216 C n 2447/3142 northern 1916/3324
moisture 192/297 nah (pronun) 152/152 northward 149/183 P
molten 111/181 C name 3023/3687 CP northwest 636/908 P
money 433/825 named 1809/2111 P northwestern 365/427 P
month 254/326 CP names 314/423 nose 140/192
monthly 127/145 narrow 587/713 C not 5323/11796 CP
months 727/923 CP nay (pronun) 124/126 nothing 224/265 C
mood 114/132 nearly 1121/1455 CP Nov 1641/1784
moon 329/765 CP neck 205/262 November 445/612
more 6171/14182 CP nee (pronun) 337/348 now 3031/4448 CP
moreover 219/255 need 664/821 CP nuclei 141/345
morning 160/193 needed 540/695 CP nucleus 216/496 P
mostly 789/991 CP needs 422/557 P nuh (pronun) 139/141
mother 651/924 neighboring 247/270 number 2584/4587 CP
motifs 137/203 neighbors 147/173 C numbered 218/248
motion 88/1405 CP neither 323/381 CP numbers 844/1401 P

403
Appendix B Alphabetical Frequency List

numerous 1390/1707 C ones 367/418 P

nuts 108/226 onset 111/137
onto 285/335 CP
open 1047/1565 C
unlisted (too common) opened 649/804 CP
near CP opening 408/509 C
nearby orange 295/478 CP p 2357/3404
no CP order 2031/2978 CP pace 107/117
ordered 262/323 pah (pronun) 100/101
orders 246/372 P paid 240/313
ordinary 350/423 CP pain 231/434
ore 347/654 C painful 135/151 C
ores 116/202 C pair 279/403 C
oriented 223/269 CP pairs 224/354 C
others 1697/2270 CP pale 157/169
our 430/595CP pan 152/218
out 2639/3957 CP panels 118/165
outbreak 190/209 paper 669/1085 CP
o 1256/1816 outdoor 112/133 papers 222/288 C
oak 158/261 C outer 405/593 CP parent 149/250
obtain 332/373 CP output 310/542 parents 356/517
obtained 618/772 CP outside 739/899 C park 443/783 P
obtaining 130/136 P outstanding 476/548 parks 223/361
occur 1087/1694 CP outward 114/140 P part 3433/5171 CP
occurred 633/845 C overall 252/293 CP parties 324/605
occurrence 152/164 overcome 153/163 C partly 303/375
occurring 249/283 P overseas 169/238 partner 146/162
occurs 935/1300 CP overthrow 177/206 partners 258/302 C
Oct 1685/1835 overthrown 121/150 parts 1479/2197 C
October 484/636 P own 2301/3177 CP party 1272/3662 P
off 1200/1641 CP owned 313/425 P past 831/1108
offer 274/349 owner 129/182 path 215/296 CP
offered 352/417 owners 126/170 patient 193/450
offering 160/180 ownership 153/235 patients 227/535
offers 241/257 oz 102/146 pattern 560/801 P
office 872/1381 patterns 631/925
offices 221/255 paved 104/109 P
offshore 138/179 unlisted (too common) pay 356/486
offspring 108/171 of CP payment 114/137
often 3529/5736 CP on CP payments 109/176
oil 858/2052 CP only CP peace 791/1420
oils 110/211 or CP peaceful 124/139
old 1947/2894 CP other CP peak 433/536 CP
older 503/634 C over CP pearl 114/204
oldest 577/652 peasant 207/285
once 1359/1708 CP peasants 183/302
one 8070/16409 CP pee (pronun) 115/116

404
Appendix B Alphabetical Frequency List

pen 138/169 placed 884/1107 CP powerful 985/1293 CP

people 2484/5106 CP places 435/529CP powers 661/1083 P
peoples 545/990 placing 128/135 C prairie 113/218
per 901/2150 CP plain 409/666 premier 297/430
perceived 133/168 plains 396/746 press 375/623 P
percent 669/1128 plan 540/834 pressed 164/187
perennial 266/364 plane 302/630 CP pressure 825/1713 CP
perfected 120/130 planes 153/255 pressures 205/267 C
perform 288/350 CP planned 301/365 prevailing 122/136
performance 485/698 P planning 327/511 prevalent 134/147
performances 231/285 plans 302/422 prey 221/500
performed 651/868 P plant 988/2297 CP price 337/577 C
performer 107/131 planted 171/237 prices 246/437
performers 109/156 plants 1105/2581 C primarily 1453/1786 P
performing 230/258 P plate 227/555 P prime 669/1344
perhaps 898/1119 CP plateau 298/575 print 121/225 C
person 890/1525 CP plates 207/386 printed 229/345
personal 949/1284 play 1011/1599 printing 270/592
persons 835/1257 played 972/1323 C prints 149/238
petroleum 509/1033 P player 325/841 P prior 359/421 P
pharmaceuticals 107/119 players 199/426 private 1067/1617
phase 361/627 CP playing 324/401 privately 118/130
phases 138/183 P pleasure 139/176 prize 942/1239
phenomena 273/434 CP plus 252/317 C prized 134/160
phenomenon 309/385 C point 1527/2388 CP procedure 217/302 CP
phosphate 104/201 C pointed 227/260 C procedures 267/376
photo 330/684 CP points 572/910 CP profits 117/162
photograph 118/182 CP poison 106/204 C prominence 225/243
photographic 198/311 C poisonous 137/214 prominent 752/849
photographs 172/261 policies 534/786 promise 162/187
photography 184/393 policy 794/1447 promised 142/170
phrase 107/136 pollution 205/441 C promote 268/302
physical 1054/1730 CP poor 675/922 CP promoted 292/310
physically 104/118 P poorly 149/163 promoting 114/124
physician 268/394 possess 221/266 C prompted 134/141
physicians 304/413 possessed 141/172 proof 117/153 P
physicist 344/432 CP possesses 102/110 proper 400/500 P
physics 584/1150 CP possession 165/199 properly 162/170 C
picture 365/650 possessions 107/126 properties 571/1102 CP
pictures 281/430 potatoes 177/217 property 540/1007 CP
piece 239/310 P pottery 199/361 proposed 574/731 P
pieces 449/593 C pound 102/215 protect 404/491 CP
pine 205/414 P pounds 111/129 CP protected 311/374
pink 204/256 C poverty 230/347 protective 179/221 C
pipe 104/211 C powder 128/217 C prove 147/173 P
pitch 124/257 P power 2435/5436 CP proved 606/733 P
place 1566/2140 CP powered 241/394 provide 1090/1506 CP

405
Appendix B Alphabetical Frequency List

provided 1037/1398 P real 601/823 CP

provides 601/737 CP rear 183/278
providing 365/430 reason 466/633 CP
published 1697/2229 CP reasons 292/327 CP
publisher 111/128 rebuilt 234/259
publishing 251/344 receive 383/491
pulled 104/122 P received 1427/1715 CP
pulp 120/233 r 4064/885 receives 221/266
punishment 149/240 race 356/549 P receiving 264/305
pupil 170/181 races 158/302 recognize 228/249 CP
pupils 108/118 racing 107/265 P recognized 972/1214 C
purchase 177/229 radius 103/178 P record 723/1137 CP
purchased 165/179 rah (pronun) 184/186 recorded 465/636 C
purity 110/125 C rail 312/442 recording 198/434 C
purple 212/288 C railroad 562/870 P recordings 124/177
purpose 511/609 CP railroads 395/566 records 447/646
purposes 441/528 railway 160/223 recover109/109 C
pursue 110/128 rain 327/519 C recovered 186/212
pursued 222/247 rainfall 473/657 P recovery 212/281 C
pursuit 122/131 raise 237/264 C red 1341/2432 CP
pushed 156/194 C raised 668/818 CP reddish 179/195 C
put 476/566 CP raising 386/423 C reduce 441/543 CP
ran 258/301 reduced 692/877 CP
random 112/184 C reduces 124/147 CP
unlisted (too common) range 1767/2641 CP reducing 240/290 C
probably P ranged 114/118 reduction 258/336 C
ranges 522/738 C ree (pronun) 272/275
ranging 493/557 reed 115/201
rank 307/554 C refer 222/239 CP
ranked 118/147 C referred 441/495 CP
ranks 245/313 refers 655/723 C
qantum rate 693/1313 CP refused 367/426
130/341 P rates 333/554 C regained 148/157
quarter 259/328 rather 1352/1793 CP regard 191/209 P
quarters 131/143 ratio 198/280 CP regarded 723/866 C
queen 530/844 raw 277/380 C regarding 129/136 C
quest 162/179 ray 490/840 P regardless 125/128 CP
question 396/551 CP rays 248/589 reinforced 146/202
questions 290/397 CP re 159/205 related 1329/1756 CP
quick 156/175 reach 666/817 CP relating 107/109 CP
quickly 597/697 reached 880/1156 CP relation 250/319 CP
CP reaches 394/451 CP relations 558/822 CP
quiet 111/130 reaching 403/454 CP relationship 506/646 CP
quite 299/341 CP read 374/502 CP relationships 284/361 CP
readily 253/298 C relative
reading 261/368 CP 470/605 P
ready 160/198 C relatively 1029/1350 CP

406
Appendix B Alphabetical Frequency List

relatives 163/193 restore 205/225 round 391/482 CP

release 340/439 P restored 448/507 C rounded 212/236
released 410/508 CP resumed 165/174 route 334/437 P
reliable retail 117/148 routes 202/274
120/139 P retain 199/217 C row 107/138
relied 127/137 retained 366/400 rows 115/136
relief 339/522 retirement 268/313 rubber 278/457
rely 114/115 revenue 137/206 rugged 163/187
remain 779/908 CP revenues 116/161 ruh (pronun) 118/118
remainder 220/240 reverse 130/153 C rule 1066/1707 CP
remained 1338/1696 P reversed 129/142 C ruled 693/873
remaining 465/533 CP review 207/263 CP ruler 355/454
remains 1080/1378 CP reviewed 145/145 rulers 330/441
remarkable 361/399 rice 354/583 rules 391/653 P
remarkably 111/114 C rich 893/1147 C ruling 280/328
remembered 364/396 richest 110/118 C run 524/701 CP
removal 235/294C richly 116/125 running 386/524 P
remove 234/287 CP ridge 167/314 C runs 249/342 CP
removed 442/553 C ridges 132/197 rush 101/154
removing 120/137 C right 1212/2244 CP
renamed 228/240 rights 807/1790
rendered 148/165 ring 330/606 unlisted (too common)
renewed 218/240 rings 152/257 repr CP
renowned 280/296 rise 1064/1416 CP
replace 247/275 CP rises 360/413 CP
replaced 846/1080 CP rising 419/486 C
replacement 111/124 risk 158/249
replacing 129/136 river 2273/5059 P
report 220/280 P rivers 813/1389
reports 203/260 P road 415/644 P
require 470/594 CP roads 372/548 CP
required 868/1223 CP rock 792/1615
requirements 264/332 P rocket 137/387
requires 431/549 P rocks 477/1210
requiring 199/218 rocky 277/388
research 1279/2111 CP rod 102/163 P s 17035/58459
researchers 191/249 rodents 103/171 safe 145/178
resemble 286/328 C roh (pronun) 203/215 safety 256/411
resembles 244/292 role 1406/236 C sah (pronun) 121/124
resembling 161/170 room 291/419 CP said 610/735 CP
resolution 141/229 rooms 131/188 sailed 142/220
resort 233/263 root 289/493 CP sale 204/268
resource 122/173 rooted 105/112 sales 184/295
resources 663/1379 roots 480/673 C salmon 134/243
response 507/712 rose 601/793 C salt 461/813 C
responses 116/169 rough 215/258 CP salts 153/263 C
rest 705/846 CP roughly 257/294 CP same 2217/3352 CP

407
Appendix B Alphabetical Frequency List

sample 100/218 C seconds 134/167 CP shadow 132/159

sand 366/684 section 465/688 CP shallow 399/516
sandstone 102/144 sections 319/375 CP shape 680/902 CP
sandy 161/192 secure 251/284 shaped 847/1057 P
save 126/138 secured 201/226 shapes 267/336 P
saved 127/137 P security 330/584 share 314/349 C
saw 480/602 CP seed 257/491 shared 486/540
say 231/282 CP seeds 333/648 shares 122/164
scale 892/1333 CP seek 261/293 sharing 106/137
scales 196/329 CP seeking 239/262 sharp 343/391 C
scandal seeks 100/110 sharply 218/238
127/155 seem 324/377 CP sheep 327/502
scattered 204/233 seemed sheet 142/237 C
scene 366/517 220/297 sheets 135/207
scenes 410/608 seems 308/348 shell 252/521 P
scenic 105/128 seen 894/1145 CP shells 177/315
scheme seized 260/308 shelter 105/137
154/190 seldom 141/152 P shield 126/213
scholar 255/283 self 1112/1543 C shift 186/268 C
scholarly 113/128 sell 151/193 shifted 172/205 CP
scholars 359/465 selling 190/226 shifting
scholarship 133/150 send 122/138 C 111/122
school 1832/3129 sense 799/1107 CP ship 339/699
schools sensitive 307/391 shipbuilding 170/191
797/1629 sensitivity 113/133 shipping 272/343
science 1251/2173 CP sensory 112/249 ships 438/830 P
sciences 471/693 P sent 614/767 P shock 137/231
scientific 899/1382 CP sentence 102/223 shore 304/373
scientist 255/305 C sentenced 103/111 shores 142/172
scientists 504/714 Sept 1655/1805 short 1967/2756 CP
scope 187/212 September 442/613 shorter 223/254
score 117/176 series 1782/2520 CP shortly 377/417 P
scored 107/141 serious 684/856 CP shot 184/255 P
scores 132/162 seriously 146/155 should 777/1180 CP
screen 229/414 services 726/1254 shoulder 186/230
sea 2057/4549 C set 1869/2680 CP shoulders 134/162
seaport 122/134 sets 334/440 CP show 860/1219 CP
search 531/646 setting 394/450 P showed 442/546 C
seas 258/396 settings 181/235 showing 196/214 CP
season 489/728 P settle 179/198 shown 622/945 CP
seasonal 151/189 settled 1 123/1278 shows 440/549 CP
seasons 216/253 settlement 697/998 shrub 124/234
seat 739/855 settlements 246/338 shrubs 222/306
seats 127/189 settlers 385/567 side 968/1503 CP
sec 134/322 P settling 119/127 sided 126/134
second 2411/3596 CP seven 767/964 CP sides 437/586 CP
secondary 440/766 seventh 118/133 siege 141/187

408
Appendix B Alphabetical Frequency List

sight 128/152 small 3443/6011 CP span 170/247

sign 230/303 CP smaller 959/1251 CP sparsely 114/126
signal 199/558 P smallest 196/230 CP spatial 106/145
signals 190/492 smoke 111/179 speak 323/370 C
signed 388/499 smooth 293/352 P speaker 110/194
signs 207/307 P snake 138/327 speakers 136/269
silent 140/184 snow 275/428 speaking 552/685 CP
silicon 117/265 C sodium 215/528 C special 1157/1687 CP
silk 175/343 soft 449/578 CP specialized 506/728
silver 536/961 C soil 562/1168 C specially 123/134
similarly 289/318 CP soils 277/703 species 1778/4735 C
simpler sold 386/466 CP specific 832/1203 CP
107/121 C soldier 241/271 specifically 264/276
simplest 160/184 C soldiers 249/358 specified 191/236 P
simply 430/494 sole 212/245 specimens 112/144
sin 140/219 solely 138/147 spectrum 204/380
since 2992/4519 CP solid 559/897 P speech 358/746
singer 273/453 soluble 111/163 C speed 604/1120 CP
singing 162/243 solution 386/681 CP speeds 184/274 CP
single 1506/2201 CP solutions 198/295 P spend 139/158 C
sir 1128/1929 solve 108/134 CP spending 133/188
sister 215/250 P solved 101/119 spent 705/771 P
site 1263/1718 something 162/201 P sperm 121/310
sites 490/775 sometimes1696/2222 CP sphere 151/236 P
six 1016/1228 CP somewhat 462/502 C spiral 129/193
sixth 145/169 son 1708/2328 spite 109/122
size 1188/1797 CP song 436/692 split 241/283
sized 171/205 songs 430/713 spoke 139/150
sizes 154/185 sons 310/373 spoken 252/490
skeleton 106/191 sort 106/115 spokesman 126/133
skill 235/284 sought 745/934 C sponsored 242/266
skilled 158/186 sound 617/1420 P spontaneous 111/153
skillful 114/126 sounds 240/470 C sport 201/398
skills 266/367 source 1247/1750 C sports 223/384 P
skin 522/1061 P sources 722/1013 C spot 106/153
skull 134/249 south 2872/6531 P spots 127/151 P
sky 260/406 southeast 620/923 P spread 769/1015 C
slave 265/456 southeastern 498/605 spreading 173/233
slavery 233/564 southern 1887/3226 spring 560/779
slaves 238/434 southward 151/192 springs 199/307
slender 245/276 southwest 566/766 spurred 145/156
slight 122/128 CP southwestern 483/556 sq 1182/3547
slightly 492/542 CP sovereign 113/147 square 408/652 CP
slope 101/158 P sovereignty 151/237 stability 237/282
slopes 179/239 P space 947/2566 CP stable 318/447 CP
slow 410/516 P spacecraft 142/426 staff 392/586
slowly 360/419 C spaces 154/203 stage 948/1538 CP

409
Appendix B Alphabetical Frequency List

stages 341/478 C storms 127/187 subdivisions 220/229

stand 302/346 P story 1235/1592 subject 1037/1366 CP
standard 746/1005 CP straight 321/396 CP subjected 142/151 C
standards 408/574 P strait 240/352 subjects 602/782
standing 325/406 CP strange 109/119 C submarine 128/239 P
stands 317/362 P strategic 225/337 subtle 166/176
star 622/1323 P strategy 170/313 P subtropical 198/254
stars 420/1263 P stream 290/436 succeed 143/149
start 234/287 CP streams 287/426 succeeded 987/1281 C
started 325/357 CP street 337/458 succeeding 220/227
starting 234/272 CP streets 171/210 P successes 143/156
state 3263/9450 CP strength 528/723 P successful1 287/1543 P
stated 166/214 CP strengthen 104/108 successfully 414/466 P
statement 201/261 CP strengthened 145/160 succession 389/498
statements 105/157 C stress 275/427 sudden 147/167 C
states 4887/11320 CP stressed 198/226 suddenly 111/120 P
statesman 280/308 stresses 106/125 suffered 413/472 P
station 317/494 stretched 105/114 suffering 193/221 C
stations 277/476 CP strict 189/207 sugar 399/741 P
statistical 152/247 C strictly 128/131 CP sugarcane 132/163
stay 159/167 strike 227/355 CP suggest 254/278 C
steadily 157/170 strikes 138/177 P suggested 349/392 P
steady 157/179 striking 293/336 suggesting 101/105
steam 254/611 C string 206/314 P suggests 243/269
steel 697/1268 C strings 120/262 P suh (pronun) 122/122
steep 132/151 strip 204/289 C suit 128/209 P
stellar 103/193 stripes 109/157 suitable 257/299 CP
stem 185/313 stroke 145/190 suited 155/171
stems 206/297 strong 1407/1836 CP sulfur 163/345 C
step 289/387 CP stronger 172/194 CP sum 139/205 CP
steps 232/307 C strongest 124/136 C summer 721/1063
still 2009/2844 CP strongly 488/543 C summers 191/234
stimulate 107/129 struck 159/225 P sun 679/1443 CP
stimulated 225/280 structural 333/505 sung 124/180
stock 366/607 structure 1361/2326 P sunlight 134/177
stomach 115/211 structures 732/1219 superb 119/125
stone 798/1520 P struggle 466/588 supplied 292/341
stones 215/351 P struggles 109/115 supplies 320/461
stood 144/162 student 415/578 CP supply 665/1033 CP
stop 204/248 P students 472/944 P surface 1233/2798 CP
stopped 130/146 studied 1243/1394 CP surfaces 345/453 P
storage 257/417 C studies 1321/1807 C surgery 165/419
store 178/252 P study 2011/3142 CP surgical 128/218
stored 218/349 C studying 408/426 CP surpassed 107/117
stores 118/157 style 1924/3617 surrender 159/211
stories 662/1040 styles 391/642 surrendered 106/124
storm 159/219 stylistic 120/131 surrounded 383/434 C

410
Appendix B Alphabetical Frequency List

surrounding 576/652 take 942/1195 CP testing 195/307 P

survey 262/323 C taken 1051/1311 CP tests 224/453 C
survival 211/249 C takes 524/599 CP theme 334/398
survive 351/415 taking 470/528 CP themes 439/538
survived 355/407 talent 154/167 themselves 831/1101 P
survives 109/113 tall 365/485 P theory 1405/3209 CP
surviving 290/320 tan 117/167 P thereafter 542/584
sustained 152/165 target 152/236 P therefore 834/1047 C
sweet 234/417 task 222/271 CP thick 433/558 CP
swept 109/134 tasks 132/170 thickness 109/143 P
swift 148/190 taste 250/324 C thin 452/633 CP
swimming 143/220 P taught 866/969 things 317/425 CP
symbol 384/442 CP tax 337/848 think 120/149 CP
symbolic 237/295 C taxation 115/154 thinkers 104/139
symbolism 147/199 taxes 200/349 thinking 174/212 C
symbols 266/377 CP tea 129/282 C third 1424/1949 CP
sympathetic 112/124 teach 149/178 thirds 233/308
symptoms 263/477 teacher 455/572 thirty 132/175
syndrome 112/192 teachers 197/353 thought 1177/1686 C
synthesis 253/402 C teaching 486/716 thousand 287/332 P
synthesized 122/146 C teachings 147/176 thousands 576/721 C
synthetic 283/460 team 304/535 threat 242/300
system 2740/6630 CP teams 148/249 threatened 279/335
systematic 231/280 P technical 476/625 three 3440/5310 CP
systems 1282/2840 CP technique 658/932 CP throat 128/182
techniques 909/1391 CP throughout1810/2517 CP
technological 216/291 P thrust 127/244 P
unlisted (too common) technologies 136/208 tidal 108/172
see CP technology 843/1346 P tied 147/169 P
served tee (pronun) 222/231 P ties 232/295
several CP teeth 297/655 tightly 112/119
she CP telegraph 128/204 timber 232/325
so CP telephone 148/296 time 4090/7407 CP
some CP tell 148/176 CP times 2322/3235 CP
soon CP temperate 371/506 tin 203/338 CP
such CP ten 590/698 P tiny 360/464 CP
tend 454/580 CP tip 263/330 P
tended 219/267 tissue 331/720 C
tends 187/213 tissues 254/489
tenure 124/135 today 1728/2452 CP
term 1892/2666 CP toes 111/167
t 1916/2645 termed 225/280 together 1225/1605 CP
table 290/469 CP terms 1018/1481 toh (pronun) 199/200
tactics 112/171 terrain 175/209 told 122/132 P
tah (pronun) 142/146 terrestial 139/225 tone 250/375
tail 531/876 P test 382/637 C tongue 165/262
tails 129/153 P tested 147/165 tons 230/490 CP

411
Appendix B Alphabetical Frequency List

too 736/948 CP trends 168/196

tool 227/318 P trial 334/541 C
tools 375/633 P trials 139/181
top 673/1043 CP tried 550/654
touch 156/176 trip 159/171 P
tough 136/157 triple 141/205
tour 204/263 troops 594/1037
toured 124/135 troubled 100/105
tourism 383/505 true 851/1131 CP
tourist 252/277 truly 124/136 P
tourists 152/166 trust 148/268 u 3361/7428
toward 1300/1799 P truth 242/338 uh (pronun) 1412/1526
tower 227/363 P try 149/169 CP uhm (pronun) 137/139
towers 127/175 trying 166/177 C uhn (pronun) 248/251
town 996/1343 tube 274/603 CP uhs (pronun) 235/244
towns 402/609 tubes 205/351 ultimate 210/249
trace 151/179 C tuh (pronun) 145/146 ultimately 338/387 C
traced 229/245 tur (pronun) 159/163 unable 326/382 P
traces 156/164 turn 849/1065 CP uncertain 154/161
track 172/328 CP turned 827/981 C uncle 146/190
tract 171/261 turning 274/312 P undergo 127/164 CP
trade 1391/2882 turns 182/218 CP undergraduate 132/183
traders 167/205 twelve 101/130 C underground 258/346 C
trading 350/486 twentieth 220/238 underlying 214/249
traffic 196/319 twenty 120/151 understand 173/209 CP
trail 100/182 two 5736/10916 CP understanding502/665CP
train 179/232 P type 1429/2313 CP understood 309/369 CP
trained 532/598 types 1286/2050 CP undertaken 104/113
training 621/893 undertook 126/135
traits 142/220 unlisted (too common) underwater 104/173 C
travel 439/590 than CP underwent 104/108
traveled 378/410 C that CP unemployment 165/283
traveling 201/220 C the CP unfinished 130/139
travels 196/238 their CP unified 227/271 P
treason 116/144 them CP unique 516/611
treasury 207/276 then CP unit 548/842 CP
treat 158/200 CP there CP unite 107/114
treated 360/439 CP thereby united 4674/9912 CP
treaties 164/240 these CP units 521/902 CP
treating 143/170 C they CP unity 313/431 CP
treatise 203/247 this CP unknown 419/487 CP
treatment 688/1169 C those CP unless 196/236 P
treaty 590/1128 though CP unlike 555/633 C
tree 666/1657 P through CP unprecedented 129/140
trees 669/1299 P thus CP unrest 105/126
tremendous 121/137 P to CP unstable 129/157 C
trend 236/272 took CP unsuccessful 335/367

412
Appendix B Alphabetical Frequency List

unsuccessfully 137/139 variety 1323/1777 C wall 513/855 CP

unusual 320/363 CP variously 105/106 walled 108/125
unusually 103/106 CP vary 547/623 CP walls 503/759 CP
upper 964/1401 CP varying wanted 139/176 CP
upright 110/128 319/355 war 4534/11217 C
uprising 168/208 vast 622/796 C warfare 258/481
upward 185/219 P vee (pronun) 139/139 warm 440/608 P
uranium 134/367 CP vegetable 179/289 warmer 130/151
urged 163/178 vegetables 211/285 C warning 100/131
us 130/170 CP vegetation 386/616 warrior 136/168
usage 131/147 vehicle 217/393 P warriors 128/167
useful 471/593 CP vehicles 170/295 C wars 778/1185
uses 761/1062 CP veins 111/178 wartime 139/154
vertebrates 121/216 waste 191/381
very 1319/2048 CP water 2069/6005 CP
unlisted (too common) vessel 160/274 C waters 507/816
under P vessels 373/648 C wave 335/717
until CP via 199/250 wavelength 119/269
up CP victim 129/179 C wavelengths 101/228 P
upon CP victims 166/196 waves 298/821 P
use CP view 771/1072 CP way 1731/2477 CP
used CP viewed 302/345 CP ways 654/837 CP
using CP views 471/571 we 215/291 CP
usually CP village 502/603 weak 269/340 CP
villages weakened 177/199 C
283/388 weakness 121/145
vivid 156/163 wealth 400/510
voice 341/451 wealthy 309/342
volcanic 315/502 P weapon 104/217
voyage 154/235 weapons 272/552
voyages 110/168 wear 130/182 C
weather 334/620 C
unlisted (too common) weaving 100/161
various CP week 186/227 P
v 1455/2305 weekly 153/180
vacuum 140/220 CP weeks 332/437 C
valley 706/1306 P weigh 205/250 CP
valleys 270/449 weighing 119/149 C
valuable 442/536 P weighs 167/189 C
value 727/1325 CP weight 763/1203 CP
valued 199/233 weights 102/197 P
values 467/735 CP welfare 220/364
vapor 115/229 C well 3766/5839 CP
variable 264/406 w 3408/4398 wells 152/259
varied 374/432 C wage 111/165 west 2616/4841 P
varies 393/447 CP wake 105/142 western 2343/4260
varieties 341/568 walk 119/171 P westward 191/250 P

413
Appendix B Alphabetical Frequency List

wet 169/230 CP word 812/1226 P

whatever 135/143 C words 512/877 CP
wheat 343/645 work 4163/7036 CP
wheel 142/336 worked 1260/1450 C
whereas 753/920 CP worker 146/212
whereby 123/132 workers 557/1069
whether 647/866 CP working 894/1134 CP
white 2138/3721 CP works 2887/4872 C
whites 139/264 workshop 105/121
whole 610/801 C world 5787/12090 CP y 939/1234
wide 1515/1876 P worlds 138/154 C yah (pronun) 125/126
wider 162/170 P worldwide 586/771 year 2689/4357 CP
widespread 632/738 C worms 128/220 yearly 131/139 P
widow 117/141 worn 103/163 years 4912/8977 CP
width 175/216 P worth 159/219 P yellow 683/1036 C
wife 691/838 would 1664/3303 CP yellowish 111/122
wild 527/839 wounded 153/195 yet 728/895 CP
wilderness 106/151 write 322/375 CP yield 260/327 CP
wildlife writing 818/1311 C yielded 151/162 C
141/237 wrote 1671/2168 P yields 183/240 C
will 1606/3300 CP you 216/292 CP
win 324/411 young 1606/2300 CP
wind 448/911 P unlisted (too common) younger 421/504
window 110/139 P want CP youngest 118/126
windows 121/206 was CP your 160/189 CP
winds 226/402 were CP youth 344/393
wine 239/358 C what CP
wing 373/561 when CP
wings 266/455 where CP
winner 131/165 P which CP
winning 376/459 while CP
winter 725/1058 P who CP
winters 194/245 whom
wire 186/329 CP whose CP
wisdom 127/177 why CP
withdraw 106/121 C widely
withdrawal 140/188 with CP
withdrew 198/225 within CP
without 1515/1963 CP z 163/240
wolf 144/224 zero 191/353 CP
woman 630/922 P zinc 199/340 C
women 1842/2727 C zone 335/664
wood 843/1481 CP zones 225/334
wooden 266/369
woods 157/197 x 567/1644
woody 104/163 XIV 186/264
wool 173/294 C

414