Applied Corpus Linguistics A Multidimensional Perspective Language And Computers 52 Language Computers Ulla Connor

Applied Corpus Linguistics A Multidimensional
Perspective Language And Computers 52 Language
Computers Ulla Connor download
https://ebookbell.com/product/applied-corpus-linguistics-a-
multidimensional-perspective-language-and-computers-52-language-
computers-ulla-connor-2206164
Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Corpus Linguistics And The Analysis Of Sociolinguistic Change Language
Variety And Ideology In Advertising Routledge Applied Corpus
Linguistics 1st Edition Joan Osullivan
https://ebookbell.com/product/corpus-linguistics-and-the-analysis-of-
sociolinguistic-change-language-variety-and-ideology-in-advertising-
routledge-applied-corpus-linguistics-1st-edition-joan-
osullivan-43716296
Corpus Linguistic Approaches To Deception Detection Routledge Applied
Corpus Linguistics Mathew Gillings
https://ebookbell.com/product/corpus-linguistic-approaches-to-
deception-detection-routledge-applied-corpus-linguistics-mathew-
gillings-120896804
Corpus Applications In Applied Linguistics Ken Hyland Chau Meng Huat
Michael Handford Editors
https://ebookbell.com/product/corpus-applications-in-applied-
linguistics-ken-hyland-chau-meng-huat-michael-handford-
editors-50219180
Corpusbased Research In Applied Linguistics Studies In Honor Of Doug
Biber Viviana Cortes
https://ebookbell.com/product/corpusbased-research-in-applied-
linguistics-studies-in-honor-of-doug-biber-viviana-cortes-56878636

Translationdriven Corpora Corpus Resources For Descriptive And Applied
Translation Studies 1st Edition Federico Zanettin
https://ebookbell.com/product/translationdriven-corpora-corpus-
resources-for-descriptive-and-applied-translation-studies-1st-edition-
federico-zanettin-5127222
Applied Water Science Volume 1 Fundamentals And Applications Inamuddin
https://ebookbell.com/product/applied-water-science-
volume-1-fundamentals-and-applications-inamuddin-44870592
Applied Water Science Volume 2 Remediation Technologies Inamuddin
https://ebookbell.com/product/applied-water-science-
volume-2-remediation-technologies-inamuddin-44870596
Applied Soil Chemistry Inamuddin Mohd Imran Ahamed Rajender Boddula
https://ebookbell.com/product/applied-soil-chemistry-inamuddin-mohd-
imran-ahamed-rajender-boddula-44870598
Applied Knowledge In Paediatrics Mrcpch Mastercourse Martin Hewitt
https://ebookbell.com/product/applied-knowledge-in-paediatrics-mrcpch-
mastercourse-martin-hewitt-44957184

Preface
The papers published in this volume were originally presented at the
Fourth North American Symposium on Corpus Linguistics and Language
Teaching, co-sponsored by the American Association of Applied Corpus
Linguistics and the Indiana Center for Intercultural Communication, held on 1-
3 November 2002 at the Athletic Club in Indianapolis, Indiana. The
conference drew more than 100 participants from 14 different countries.
Altogether, 52 papers and 12 posters were presented.
The 15 papers in this book are divided into two sections: (1) analyses
of spoken and written language corpora and (2) pedagogical applications of
corpora. The first section opens with Anne O’Keeffe’s paper that uses a corpus
of phone calls to an Irish radio show to explore vague categorization and
shared socio-cultural knowledge. Employing a bottom-up approach, O’Keeffe
shows how a corpus can be used to identify a wider range of vague categories
than a preset list would allow. Once identified, the forms were analyzed and
showed that vague language categorization was a by-product of a close
relationship at a societal level of interaction.
The next two papers are based on the Hong Kong Corpus of Spoken
English. Martin Warren reports on the analysis of discourse intonation to study
how speakers assert dominance and control in conversation. His findings
suggest that the choice of a certain tone is at least partly determined by the
discourse types as well the roles of the speakers; no difference was found
between native English speakers and Hong Kong Chinese English speakers in
the corpus in terms of their tone choices. Winnie Cheng’s paper analyzes
patterns of lexical collocations and intonation in public speeches made by a
government official in Hong Kong. She found that these features were often
used to establish a dynamic relationship between the official and his audience,
and to promote ideology and political agendas.
For the purposes of analyzing both written and oral discourse,
Douglas Biber et al. use an approach that combines corpus-linguistic and
discourse-analytic perspectives to examine patterns in two corpora: the T2K-
SWAL (TOEFL-2000 Spoken and Written Academic Language) Corpus and
the Longman Spoken and Written English Corpus. Three different registers
(i.e. classroom teaching, textbooks, and academic research articles) were
analyzed for the use of the Vocabulary-Based Discourse Units. These units
were then subjected to the analysis of their primary linguistic characteristics,
using Multi-Dimensional techniques. Interesting patterns were found across
the registers in the use and type of units. JoAnne Neff et al. conducted a
contrastive study of argumentative essays by expert and novice writers in
English and Spanish, and of similar essays written by Spanish EFL students, to

iv
determine which linguistic and rhetorical features were transferred from the
L1, and which resulted from other factors such as writing experience and
education. Their paper also includes a section comparing the long-term results
from SPICLE with large English-language corpora. Susana Sotillo and Julie
Wang-Gempp report a study using a corpus of online political discussions to
examine class, ideology, and discursive practices. Employing theories of
critical discourse analysis, they identified rhetorical and linguistic devices to
characterize a variety of socio-economic, cultural, and political issues in four
political discussion threads.
In the second section of this book, nine papers deal with applications
of corpus linguistics to a variety of teaching situations. This section opens with
Sylviane Granger’s plenary address at the conference titled “Computer learner
corpus research: Current status and future prospects.” Granger explains the
advantages of computer learner corpora as compared to other learner corpora,
such as size, the ability to include information on a large number of variables,
and automation of analysis. She continues with practical advice for learner
corpus builders regarding collection of data, analysis, and interpretation. The
final sections of Granger’s paper deal with the contributions of computer
learner corpora for second language acquisition and foreign language teaching.
The other eight papers in this section offer a wide variety of
pedagogical applications of corpus linguistics. Boyd Davis and Lisa Russell-
Pinson discuss the use of corpora and concordancing in courses for ESL,
sheltered-content ESL, and content-area teachers in a K-12 setting. They
describe a project that has created a corpus of oral narratives in native
languages of students in the school system to be used in the training; the
project has also trained teachers in the use of the materials, instructed teachers
in computer-based techniques, and worked with teachers and administrators in
the school system to increase cultural competencies. In the next paper,
Wolfgang Teubert argues for the importance of the teaching of collocations to
foreign language learners rather than using the single word approach. He
shows how parallel corpora need to be compiled and how vocabulary,
including the collocations of a language, needs to be taught from the source
language perspective. William Fletcher continues the pedagogical thread by
focusing on identification of source texts for corpora on the Web. He describes
a system used to identify representative documents efficiently, and provides
examples of its use with written texts from the British National Corpus. The
paper by Mark Davies deals with the use of corpora in historical linguistics
courses. Davies describes how students, using the recently-completed “Corpus
del Espanol” – a “web-based, 100 million word, fully-annotated corpus of
Spanish texts from the 1200s-1900s” – and parallel corpora in late Latin, Old
Spanish, and modern Spanish, can compare linguistic structures and study
historical developments in the Spanish language using real-life language
examples.

v
The final four papers in the section focus on the pedagogical
applications of corpus linguistics for the teaching of English. Eileen
Fitzpatrick and M.S. Seegmiller’s paper introduces The Montclair Electronic
Language Database of writing by undergraduate ESL students. The data base
is tagged for grammar and allows for various analyses of academic student
writing. Joybrato Mukherjee presents the results of a survey among English
language teachers in German secondary schools before and after training
workshops on corpus linguistics. The results show that the use of corpora has
not been central to the practice of English language teaching in Germany.
Based on the results of the survey, Mukherjee proposes an agenda for
workshops on corpus linguistics for English language teachers. The workshop
is potentially compatible with other EFL contexts in addition to Germany.
John Osborne’s paper suggests ways in which both top-down and bottom-up
approaches to corpora can be used in EFL teaching. Using data from learner
and native-speaker corpora, Osborne shows examples about lexical overuse,
grammatical anomalies, non-count nouns, and connector use that are helpful
for teachers of EFL students. In the last paper, Pieter de Haan and Kees van
Esch describe a research project being undertaken in The Netherlands with the
goal of establishing an adequate instrument for measuring the development of
EFL student’s writing skills. Their initial findings indicate that certain lexical
and discourse features, such as word length, sentence length, essay length, and
type/token ratios, correspond to the overall assessment of student writing.
We would like to express our gratitude to many individuals who
helped in the preparation of this book. We thank the Steering Committee of the
Conference (Michael Barlow, Sarah Briggs, Fred DiCamilla, Gene Halleck,
Paul Heacock, Aymerou Mbaye, Charles Meyer, Randi Reppen, Tony Silva,
and Rita Simpson) in the planning of the conference and in the preliminary
selection of the papers for the conference. We wish to thank Charles Meyer,
series editor of Language and Computers, for his encouragement and
assistance throughout the editing of this book. We also wish to thank Jing Gao
and Kyle McIntosh, graduate students in the English Department and research
assistants at the Indiana Center for Intercultural Communication at IUPUI, for
their diligent editorial work in preparing this book for publication.
Indianapolis, Indiana, March 2004
Ulla Connor and Thomas A. Upton

‘Like the Wise Virgins and All that Jazz’: Using a Corpus to
Examine Vague Categorisation and Shared Knowledge
Anne O’Keeffe
Mary Immaculate College, University of Limerick, Ireland
Abstract
This paper will use a corpus to explore vague categorisation (e.g., prostitutes, sailors and
the like) in a specific context where the participants are strangers, but where they share
the same socio-cultural reference points and so can assume a critical level of shared
socio-cultural knowledge when they use vague language. Unlike most work on vague
language, this study looks at vague items which are not necessarily pre-textual or
prototypical, but which emerge from shared knowledge. The corpus comprises 55,000
words of calls to an Irish radio phone-in show. Vague category markers are isolated and
described in terms of form and domain of reference. It is argued that the shared knowledge
required in order to construct vague categories has a common core of socio-culturally
ratified 'understandings' and that the range of domains of reference of these categories is
relative to the depth of shared knowledge of the participants and relative to their social
relationship.
1 Introduction
Much theoretical debate surrounds the epistemic (i.e. knowledge) status of
vagueness. According to the epistemic theory of vagueness, there is no absolute
state of ‘borderline.’ If someone is borderline bald, for example, this theory holds
that s/he is either definitely bald or not bald, but we (as the speakers) cannot at
that point determine (Williamson 1994). However, Jackson (2002) argues that the
role of language in communicating our thoughts about how things are makes a
strong case against this absolutist theory. Other recent philosophical arguments
look at vague language in context (see, for example, Pinkal 1985; Manor 1995;
Kyburg and Morreau 2000). Kyburg and Morreau (2000), for example, take the
stance that ‘contextuality’ and ‘accommodation’ are characteristic of vague
language between speakers and hearers in context:
…just as a handyman can fit an adjustable wrench to a nut, we think, a
speaker can adjust the extension of a vague expression to suit his needs,
relying on the hearer to recognize his intentions and to accommodate him.
(Kyburg and Morreau 2000: 577)
The linguistic study of vague language has been greatly influenced by
Grice’s (1975) Co-operative Principle (CP) and its associated conversational

Anne O’Keeffe
2
maxims. Most notable is the work of Cruse (1975, 1977) who points to the
relativity of vagueness: “a speaker wishing to refer to something in his
surroundings is frequently, if not usually, faced with a range of lexical items of
different levels of specificity, all of them equally appropriate from the point of
view of their inherent sense” (Cruse 1977: 153). Cruse (1977) explores the notion
of unmarked or neutral levels of specificity in various contexts which are not
necessarily covered by Grice’s maxims. He presents a system of markedness in
terms of level of specificity. Of salience to the present study are some incidental
comments made by Cruse in this 1977 paper. Firstly, he makes the point that
under-specification de-emphasises the feature that is omitted, while over-
specification emphasises or intensifies the added feature as an example of under-
specification, he mentions expression of compassion or pity. Apart from under-
specifying simply for reasons of unwillingness to give information, Cruse also
points out that a speaker may underspecify because s/he is an expert in a
particular field, or has at least an everyday familiarity with some class of things:
“the speaker is in effect suggesting that the referent has such a high degree of
givenness in his universe, that he cannot make what is an unmarked reference
without underspecifying” (Cruse 1977: 163).
It is this notion of ‘givenness’ which Cruse associates with vague
language use that is of interest in this paper. We will examine the use of vague
categorization in a very self-contained context where speakers within the same
society draw on their shared knowledge in the frequent use of vague
categorisation. The data will be taken from a small corpus of radio phone-in data
from an Irish radio show called Liveline (see 3 below). When the Liveline
presenter and callers underspecify, they are drawing on assumptions and
expectations about the ‘givenness’ of the shared social and cultural knowledge
and information of their co-participants. Take the following example where a
caller detailing her experience of Maori body tattoo draws on the givenness of our
knowledge of the human physique:
1)
Caller: And their tattoos were absolutely weepingly beautiful. They
were extraordinary. And those men were tattooed all that I
could see okay so starting with the forehead face ears neck
hands et cetera et cetera.
Presenter: Yeah.
Here the caller can take a linguistic shortcut using the vagueness marker et cetera.
This allows her to say ‘the forehead face ears neck hands et cetera et cetera’
instead of forehead, face, ear, neck, hands plus a tiresome list of all bodily parts
that were tattooed. Our shared knowledge of the human body combined with the
speaker’s knowledge of the givenness of this information facilitates such under-
specification. At a more culture-specific level, we find the following type of
example in the data where the ‘given’ or implicit information is not as universal
as in the previous example:

Using a Corpus to Examine Vague Categorisation 3
2)
A caller is reminiscing about his schooldays in an Irish boarding school fifty
years ago.
Caller: …you were supposed to be on a rugby pitch or something like
that you know …
Presenter: Right.
Here, in order to complete the referential set a rugby pitch or something like that,
the listener needs to have shared information from an Irish social context of the
type of games that are usually played in an Irish Catholic boarding school fifty
years ago. An ‘outsider’ (i.e. someone from outside of Irish society) hearing this
utterance can engage with it to a certain degree, either by under-specified broad
human knowledge, or by (possibly incorrectly specified) cultural analogy with
his/her own culture.
The aim of this paper is to examine vague categorisation in context using
a self-contained corpus of data as a measure of the range of shared or given
information of the participants. In other words, by isolating and analysing all of
the vague categories that are constructed by the speakers in the data, it is hoped to
find indexical information about these participants. In so doing, we may gain an
insight into the nature of the shared knowledge that binds this group. It will also
be argued that a corpus provides a very useful tool for the study of vague
language in use.
2 Previous Research
Vague language is defined in a number of ways. Franken (1997) distinguishes
between ‘vagueness’ and ‘approximation’ while Channell (1994) restricts the
definition of vagueness to ‘purposefully and unabashedly vague’ uses of
languages. She divides vagueness into three categories: 1) vague additives (which
include vague approximators such as about and tags referring to vague categories
such as and things like that), 2) vagueness by choice of words (e.g., yoke; thingy)
and vague quantifiers (e.g., piles of), and 3) vagueness by implicature (e.g., the
sentence Sam is six feet tall has the potential to be vague as he may be six feet
and a quarter of an inch tall; see Channell 1994: 18). On the other hand, Zhang
(1998) makes a case for four separate categories: ‘fuzziness’, ‘generality’,
‘vagueness’ and ‘ambiguity’. Unlike Zhang, Chafe (1982) puts vagueness and
hedging together into the category of ‘fuzziness’ all of which are seen as
‘involvement devices’ more prevalent in spoken rather than written language. The
notion of vagueness as an involvement device is consistent with the stance taken
here: that to be vague is to draw on what is given and shared within the
participation framework of the Irish radio audience.
Similar to Channell (1994), Powell (1985) focuses on the notion of
purposeful vagueness. She deals with vague quantifying expressions and argues

Anne O’Keeffe
4
that ‘a maximally efficient exchange of information may be vaguely encoded, and
purposively so, if the principal function of the exchange is essentially non-
descriptive’ (Powel 1985: 32). She also shows that vague quantifying expressions
may encode a speaker’s judgement and that this dimension of use is principally
evaluative in function. The following example from the radio data clearly fit this
model.
3)
Presenter: Why did you decide on boarding school?
Caller: Well we live in the country and the nearest school to us was
going to be fifteen miles away where we= our boys would be
big into sports and all that.
The presenter asks a straightforward referential question as to the caller’s motives
for sending her children to boarding school. The non-descriptive answer gives
two motives 1) the distance from the school and 2) her children’s love of school
team sports and school activities which might not have been sated had they lived
at home given the distance of the family home from the school and the need to
stay on after school hours for games, training and school outings. Here we see
that ‘our boys would be big into sports and all that’ serves as a shortcut to motive
number 2 above. What is of note for this investigation of vague language in use is
the level of assumed knowledge anticipated on the part of the caller in using this
vague linguistic shortcut. This reference is ratified by the presenter who finds the
caller’s explanation adequate and unambiguous (either for her or for the listeners
on whose part she arbitrates). It is also interesting to note that when the caller
says ‘we live in the country’, this is implicitly understood within an Irish context.
It would be taken to mean that we are farmers living outside a town or village
away from a school bus route. In another social context this would often carry
different implications.
2.1 Categories and Categorisations
In this paper, analysis will be limited to vague category markers as such as those
illustrated in examples 1, 2 and 3 above (as opposed to looking at individual
vagueness items such as nouns, quantifiers and so on). Vague category markers
go by various terms across different studies for example, ‘general extenders’
(Overstreet and Yule 1997), ‘generalized list completers’ (Jefferson 1990), ‘tags’
(Ward and Birner 1992), ‘terminal tags’ (Dines 1980; Macaulay 1991), ‘extension
particles’ (DuBois 1993) and ‘vague category identifiers’ (Channell 1994).
Most research into the nature of categories has been concerned with
lexicalised categories, that is those that are encoded as a single lexical item (e.g.,
bird, furniture) (see, for example, Mervis and Rosch 1981; Rosch 1978; Rosch et
al. 1976). Many of these studies look at categories in terms of prototypes
(exhibiting the highest concentration of characteristic properties) compared with

peripheral category members (containing fewest characteristic properties). Of
more relevance to this study, Barsalou (1983, 1987) looks at the question of
whether categories are stable or subject to change. In particular he talks about the
dynamic nature of ad hoc category formation, for example places to look for
antique desks. In such examples, categorisation is non-lexicalised and without
clear boundary. This challenges the notion that categories are stable easily
recognisable and arrived at ‘pre-textually’ (after Overstreet and Yule 1997).
Overstreet and Yule (1997: 85-86) reflect that:
If only common (i.e. lexicalised) categories are studied then little insight
will be gained into the discourse processes involved in categorisation when
a single lexical item is not available to the discourse participants for the
referential category.
Building on the ad hoc categories of Barsalou (1983), they stress the spontaneity
of categorisation and the context-dependent nature of the categories themselves
when one looks at examples from actual discourse as opposed to stylised
examples. Overstreet and Yule (1997: 87) suggest a continuum from lexicalised
to non-lexicalised categories based on the degree to which categories are: a)
conventionally and linguistically established and b) constrained by contextual
factors. They refer to the set of forms that generate non-lexicalised categories as
“general extenders” which they see as integral to the process of establishing
categories that are locally contingent in discourse. In this analysis these forms
will be termed “vague category markers.”
The vague category markers in the corpus will be seen as recognisable
chunks of language that function in an expedient way as linguistic triggers
employed by speakers and decoded by participants who draw on their store of
shared knowledge. It is argued here that the meanings of vagueness categories are
socio-culturally grounded and are co-constructed within a social group that has a
shared socio-historic reality. However, it is wise to issue the caveat that without
access to the speakers for personal reflection, we cannot know for certain whether
they choose to take linguistic shortcuts: a) to be “deliberately and unresolvably
vague” (Powell 1985: 31) or b) to be expeditious and adhere to conversational
norms of quantity.
3 Data
Data for this study are drawn from an Irish radio phone-in show called Liveline
broadcast every weekday on Radio Telefís Éireann (RTÉ) between 1.30pm and
2.45pm. The transcribed corpus comprises approximately 55,000 words. The
programme has been running for almost 18 years and according to recent research
has an audience of 365,000,1
almost 10% of the Irish population. Its longevity and
prominence on Irish airwaves makes it rich for analysis on many levels. The data
were taken from a sample of programmes in 1998, and comprises 44 phone calls

Anne O’Keeffe
6
(from a total of five programmes) spread throughout that year. Programme
selection dates were spread throughout the year at intervals that would avoid daily
or seasonal skewing (i.e. spread around different days of the week and months of
the year at more or less equal intervals). Once dates were chosen, the relevant
programmes were recorded from the RTÉ radio archive and the researcher had no
prior knowledge of what topics would be covered on these programmes. In the
data, topics for discussion meander from call to call and include the following
miscellany: female facial hair problems; tattoos; the peace process in Northern
Ireland; how ears were pierced in the old days; constitutional referenda,
experiences of working abroard; cursory tales about sunbathing without sun
block; reminiscing about boarding schools; warnings about the decline of fidelity
and moral decay in general; things that can go wrong when working in Saudi
Arabia and the growing trend of litigation in Irish society among others. Unlike
many talk radio shows, the presenter in Liveline does not normally provide
counselling and she generally avoids engaging in strong debate. Her role appears
to be more that of conduit between the caller and the audience (see also O’Keeffe
2002; McCarthy and O’Keeffe 2003).
4 Analysis
The analysis focuses on any forms that make vague reference to sets or
categories. Research tells us that vague category markers are found in clause-final
positions and mostly comprise a conjunction and a noun phrase; however,
because a bottom up approach to identifying all vague categories in the data was
used, there was no pre-selection criteria based on form. This poses a challenge for
corpus analysis in that these data must first be checked manually. However, since
we are dealing with a small corpus, this is not an impossible task. While corpus
tools assist in checking the accuracy of the manual searches, there still remain
questions of validity and reliability. In order to enhance these aspects of the
study, two raters were used. One of the raters was from the Republic of Ireland
and one from Great Britain. Retrospectively, it proved very important to have one
rater who was familiar with the cultural references in the data. However, the non-
Irish rater proved to be crucial to validity and reliability of the study as he was in
a position to see exactly when a vague category was exclusively within an Irish
reference domain (see section 5.2). Surprisingly, it was not always as easy for the
Irish rater to see the range limits in her own cultural reference domain. Such a
study could not have been conducted without this inter-rater reliability check.
In this analysis therefore, the 55,000 word corpus was searched
exhaustively by rater 1 to identify manually any forms that were used in vague
categorisation (as a follow up, Wordsmith Tools was used to generate accurate
quantitative results). These were crossed checked by rater 2. The categories or
“sets” which were found in the data were then logged for subsequent analysis.
Any forms which were co-textual (i.e. referring back to a previously identified set
or category made explicit in a previous turn) were not included – see extract 4 as

an example where this is the case. The form anything like that appears to be a
vague category marker, but on closer examination, it refers back to a catalogue of
headaches which the caller details earlier in the turn:
4)
A caller talking her experiences of the side effects of taking a contraceptive pill.
Caller: Am well I’d nasty headaches very nasty headaches am I was on
it for a month. I went on to it for the second month and a couple
of days into it I was out one night I wasn’t feeling myself and I
went home and the following day I’d ah very very bad
headaches and I knew there was something wrong myself
because I’d never experienced anything like that and I don’t
suffer from migraine so am I went to bed got up on Monday
went to work felt dreadful in myself as well as having the
headaches+
5 Results
In all, 138 vague category markers were identified. Each form was classified
according to its reference set as the following example shows for the set: a lot of
undesirables criminals and people like that…(while this might appear to be a
global reference, it was used in a specifically Irish context by the caller):
Table 1: Sample breakdown of initial analysis of categories
Form a lot of undesirables criminals
and people like that…
Reference set (i.e. the set or category that
is alluded to)
Criminals and social undesirables
Broad category (i.e. the broad category
of reference; e.g., Irish historical, global)
Irish social
No. of occurrences 2
5.1 Forms of Vagueness
The following distribution of forms was identified in the Liveline data.2

Anne O’Keeffe
8
Table 2: Distribution of vagueness forms used to mark categories in Liveline data1
Form Example Raw
result
Result/
million
words
thing(s) We’re going to get a clatter of phone
calls talking about there was one nurse
I can’t remem= <chuckles> was it
nurse Caddin wasn’t she involved in
the most extraordinary things in
Dublin?
46 836
X like that a lot of undesirables criminals and
people like that…
21 381
…[that/that]
[kind/sort/ty
pe] of X
unhappy homes all that kind of thing 17 309
and so on Conviction about social justice and so
on
12 218
Or
any/somethi
ng
And it worked very well in fact the day
boys were very useful because you
could often get them to get fags for you
up town or [Yeah] buy a bar of
chocolate or something and smuggle it
in you know.
9 164
Or whatever the expense of insurance and ah people
for instance organising voluntary
sporting activities now find that you
know if yeah have a gymkhana or
whatever …
9 164
Et cetera the development of piers, roads et
cetera and et cetera
6 109
(and) all that
…
maybe they are like the w= the wise
virgins and all that jazz
3 54
(and) stuff …out at discos and stuff 3 54
this that and
the other
with this that and the other thing 3 54
and so forth talking about married men an= ma=
and so forth
2 36
or that I didn't know anything about lights or
that and they told me that the lights
was affecting his eyes
2 36
for the X
that’s in it
a bit of respect for the day that’s in it. 1 18

or some
other one of
X
the ozone layer or some other one of
these quare things up there in the sky
1 18
and
everything
the whole attitude in the school is like
rugby at the moment it’s the rugby
season and the Cup and everything
1 18
or any of X I'd like Bertie or any of them get on
and address what we're voting on on
Friday
1 18
Or _ing . I just saw a lot of kids now by kids I
mean up to maybe age of twenty-four
or so [Mm] enjoying themselves or
doing whatever they do to that
particular form of sound they use as
music.
1 18
TOTALS 138 2,505
These results are presented in Figure 1:
Figure 1: Forms of Vague Category Markers found in Liveline Data (occurrences
per million)
By including any form in the data that is used to construct a vague category, we
find a somewhat unorthodox collection. As mentioned above, Channell (1994)
noted that most vague category markers were clause-final (conjunction) + noun
phrase pattern; however, the results here deviate from this, for example adverbial
phrases and so on/and so forth, with no noun phrases. In particular, we also note
0
100
200
300
400
500
600
700
800
900
t
h
i
n
g
(
s
)
X
l
i
k
e
t
h
a
t
…
[
t
h
i
s
/
t
h
a
t
]
[
k
i
n
d
/
s
o
r
t
/
t
y
p
e
]
o
f
X
a
n
d
s
o
o
n
O
r
a
n
y
/
s
o
m
e
t
h
i
n
g
O
r
w
h
a
t
e
v
e
r
E
t
c
e
t
e
r
a
(
a
n
d
)
a
l
l
t
h
a
t
…
(
a
n
d
)
s
t
u
f
f
t
h
i
s
t
h
a
t
a
n
d
t
h
e
o
t
h
e
r
a
n
d
s
o
f
o
r
t
h
O
r
t
h
a
t
f
o
r
t
h
e
X
t
h
a
t
’
s
i
n
i
t
o
r
s
o
m
e
o
t
h
e
r
o
n
e
o
f
X
A
n
d
e
v
e
r
y
t
h
i
n
g
O
r
a
n
y
o
f
X
O
r
_
i
n
g

Anne O’Keeffe
10
the inclusion of thing(s) which is not necessarily clause final. Figure 2 provides
several examples.
2 gs just won't laughs am some things have to be faced
3 with rose tinted spectacles and saying things were great you don't
4 st I mean I don't mean to be dismissive things like social justice
5 want certain things to go away but some things just won't
6 le but it's um an Islamic country. It's things are so so so
7 eah. +which are doing different things and I can be left ou
8 yes ah of doing unspeakable things to one another ah to
9 before they were in a rural area I mean things have just changed.
10 roller coaster you know I mean I think things are moving very fast
12 Oldest swinger in town kind of thing? Absolutely.
13 mo tour of Italy and all these kind of things about four years ago
15 Which is dependent upon all kinds of things happening in the
16 e whole idea is that a fresher look at things ah by looking at the
17 I did fifty five years ago and a lot of things have passed under
18 RU and they get criticised for a lot of things I don't think it was
19 get to understand that you owe a lot of things to other people in
20 in the most extraordinary things in Dublin?
21 so yo= you know there's a lot of funny things happening in
22 rse it does but she hears a lot of good things about it too. It's
23 it's associated with all sorts of seedy things like venereal
25 sound mild in comparison to some of the things that other people
Figure 2: Sample concordance lines for thing(s)
Fronek (1982), writing on the word thing(s), notes that “the poverty of its
semantic content makes it a very good candidate for the various degrees of
semantic re-categorisation and for use as a function word” (Fronek 1982: 636).
Many linguists (see Bloomfield 1933; Hockett 1958; Gleeson 1956; Lybbert
1972) have focused on the capacity of the word to acquire grammatical function
because of its “semantic emptiness” (Fronek 1982: 636). Fronek illustrates that
there is no sharp distinction between the lexical and grammatical classes. Thing is
an extremely flexible function word capable of shedding most of its semantic
content and thus becoming suitable as a pro-form while also capable of behaving
like any other noun (ibid: 652-3). However, Fronek notes that especially the
plural indefinite things can have such a vague indeterminate referent as to be
almost indefinable. Its notional content is so minimal that from the semantic point
of view it is virtually redundant (ibid: 645). This assertion is disputed here, at
least in the context of the current study, where it is held that what might seem to
the analyst to be indeterminately vague is communicatively and pragmatically
adequate to the collective users in context. Take the following example:
5)
The caller is talking about a boarding school he attended many years ago and he
has just mentioned that at one point it became a mixed gender school.
Caller: … that was sort of a <unintelligible word> an indicator of what
things were to come in the future you know.

Presenter: Yeah yeah. Well I mean there now in a very built-up area
whereas before they were in a rural area I mean things have just
changed.
Caller: Th= that’s right. Yeah that’s right.
The first use of thing (by the caller in line 1) refers to things that were to change
in relation to the school and the presenter ratifies her understanding of this with
yeah yeah (line 2). The presenter then uses things (line 2) to refer to broader
changes in the area around the school (Newbridge College, situated in County
Kildare outside Dublin), which fifty years ago was rural and which now is a very
built up satellite town of Dublin. To a non-Irish observer, it is fair to say that
these two uses of thing are almost opaque, but to someone who has access to the
social information of the participants, this is a normal inexplicit reference to
given, shared information.
5.2 Reference Domains
The 138 vague language forms were distributed across three broad domains:
national references (i.e. Irish), global, and European, as well as a fourth, minor
category, biblical. Rater 1 sorted the items into these broad categories and rater 2
cross checked these. As discussed above, rater 2 was from outside of Ireland and
was better placed to identify exclusively Irish references.
Figure 3: Broad reference domains of categories
As we can see in Figure 3, these sets fall mostly into two reference domains: Irish
and Global. A further breakdown of the Irish category is profiled in Figure 4:
0
20
40
60
80
100
Irish Global European Biblical

Anne O’Keeffe
12
Figure 4: Breakdown of the reference domains at a national level (i.e. Irish)
The General set is all references that are not related to Northern Ireland issues or
historic collective Irish knowledge. They are contemporary social reference
points spanning a multitude of social issues and information. This could be seen
as the most core or most common information held within the group of
participants. Examples from this category include: typical accidents that happen
to people in Ireland; small midland towns in Ireland; typical contemporary issues
that are discussed in Ireland; social activities typical of an Irish teenager.
5.3 Categorisation as Generic Indexical Information
The categories co-constructed within the participation framework of Liveline give
clues as to the profile of the audience. Clearly, it is an Irish-centred one, with the
main core of reference points centring on general Irish social knowledge. When
these “general” data are scrutinised more closely with the help of concordance
line analysis, we find that the locally contingent categorisation can be divided
into four categories 1) social practices and attitudes; 2) social responsibilities and
realities; 3) work, financial and consumerist practices and 4) social types. Raters
1 and 2 devised these four categories and then independently sorted the items.
Table 3: Categories within in General Irish reference domain
Category Example
Social practices and attitudes The process of “word of mouth” in
Ireland
Social responsibilities and
realities
Negative social realities that come
with the Celtic tiger economy
Work, financial and
consumerist practices
Car rental companies in Ireland
Social types Irish criminals and social
undesirables
0
10
20
30
40
50
60
70
80
General Northern Ireland Historic

Their results were then compared and any anomalies were scrutinised and
resolved. Table 3 provides and example for each of these. Figure 5 shows how
these are distributed in the corpus of data.
Figure 5: Percentage breakdown of general Irish references
Based on these core reference points (i.e. most general or “common
denominator”), it is fair to assert that they index or place the participants of the
radio phone-in show Liveline as a socially-aware, middle class group. Most
telling in this respect are the frequent categorisations in relation to “others” in
Ireland who are socially disadvantaged (for an in-depth treatment of this see
O’Keeffe 2002).
5.4 Stereotypes and Prototypes
As discussed earlier, much research has looked at semantic prototypes in the
construction and stabilising of categories. However, it is of note that when we
look at their construction within the stable participation framework of Liveline,
we find that many of the social references are dependent on (and symptomatic of)
stereotyping. In example 6 we find a typical example:
6)
Talking about why people send their children to boarding schools.
Presenter: And you I mean ab= some people were there saying oh well
sending them away unhappy homes all that kind of thing
For the majority of people listening to this caller at the time, we can only assume
that they deconstruct the meaning of the category unhappy homes all that kind of
0
5
10
15
20
25
30
35
40
Social practices
and attitudes
Social
responsibilities
and realities
Work, financial
and consumerist
practices
Social types
%

Anne O’Keeffe
14
thing based on a stereotype as opposed to direct first-hand experience, and this is
the case for many of the examples found in the data.
4 ell sending them away unhappy homes all that kind of thing.
Mm I know
6 mean there was an awful lot of pain and that kind of thing.
7 ools are from unhappy families there is that kind of element I
suppose but mo
8 ke venereal diseases or prostitution or that kind of thing? Well
I I
10 der. And had he been subject to that kind of physical
torture?
11 n't I. And you know Marian if you're in that kind of am hostile
environment
18 ld is going to be and so on I mean that that sort of issue I think
we need to
19 create divisions and conflicts and all that sort of thing.
Yeah.
20 react quite strongly to stress and all that sort of stuff so I
have I'm now
2 g out of this ah situations of hardship and so on I think we
wouldn't say tha
4 xecuted and the other was to get lashes and so on and so forth. Yo=
yo= did y
10 ing through ah this system and the pain and so on and so on. But
having said
15 h ah this system and the pain and so on and so on. But having said
that let me
112 it's associated with all sorts of seedy things like venereal
diseases or
120 Won't that be the most subversive thing that has been done to
both sets
123 e the point of road rage ah this is the thing I'm concerned with
where local
132 or you know the the the danger is these things get worse and worse
and become
133 t okay so let people be beware of these things. Okay? Yeah.
But the
134 h am tough-minded view would say "these things happen. It's too
bad''.
139 roller coaster you know I mean I think things are moving very fast
but all
151 yes ah of doing unspeakable things to one another ah to
come
Figure 6: Samples of negative semantic prosody associated with categorisation
5.5 Categorisation and Semantic Prosody
As noted above, Cruse (1977) tells us that under-specification de-emphasises the
feature that is omitted. On examination of the semantic prosody of the categories
constructed in the data, we find many that are negative. It could be posited

therefore that one of the motivations for using vague categories could be to avoid
over-specification in negative domains. In other words, many of these uses could
serve as euphemisms. Within the context of the work of Louw (1993) and
Sinclair (1996) on semantic prosody, we could say that vague category markers
cluster with lexis which have negative prosody. The degree to which the under-
specification of negative categories is culturally marked cannot be measured or
proven here but it is put forward that it may be so. Figure 6 provides a sample of
some of the negatively marked lexical items which collocate with vague category
markers in the data.
6 Categorisation and Hedging
It could be argued that the construction of vague categories serves as a form of
hedging as the following example illustrates, where the presenter is asking a
difficult question, that is whether the caller was in receipt of a government “hand
out” (note: the Gaeltacht refers to areas of Ireland where Irish is the first
language).
7)
Presenter: Didn’t get a Gaeltacht grant or anything like that?
Caller: No I didn’t get anything not a grant aid whatsoever.
The vague category marker clearly functions to downtone the accusation implicit
in the question. We posit that when speakers want to hedge the force of negative
utterances they can choose to construct a vague category as a discourse strategy.
This supports Cruse’s point cited earlier that under-specification de-emphasises
the feature that is omitted (cf. Cruse 1977).
7 Categorisation as a Generic Activity
Warren (1993) tells us that inexplicitness (of which the construction of vague
categories is one exponent) depends on overlapping factors: 1) the physical
setting and 2) shared knowledge. Transposed onto a national radio audience level,
this assertion fits the findings in this study where the majority of the vague
categories constructed have their reference domain in physical or social space in
Ireland, and all are bound by an almost uncontested ratification by participants in
the construction of their meanings. In other words, these vague categories are
perfectly transparent for their users (though this may not be the case for the
analyst). This has interesting implications for the study of spoken genres. It points
to the speaker-addressee interdependence in the co-construction of meaning and it
points to the bi-directionality of spoken discourse. Take the following example:

Anne O’Keeffe
16
8)
Caller: I have Emm she’s fourteen and her brother slags her now he’s
sixteen he would be going ‘‘look you have you have hair unde=
you have a moustache” and all this so I do have to give out to
him.
Presenter: Yeah.
A non-lexicalised category of things that an Irish teenage boy might say to tease
his sister who has a facial hair problem (and even how it might be said) is vividly
invoked by the caller with minimal lexical effort: he would be going “look you
have you have hair unde= you have a moustache” and all this. This is perfectly
understood by the addressee (and we assume by the audience as hearers), but
crucially it is facilitated by the triangulation between all three: speaker; addressee
and hearer(s) because they know the range of common knowledge that the other
knows.
Over time the participants develop a sense of the internal range of shared
knowledge which can be drawn on. In other words, the range of shared
knowledge accrues within the participation framework. This store of shared
information allows speakers to draw on generic resources with minimal lexical
effort. Consider the following example:
9)
A caller who owns a hostel in the West of Ireland is telling a cautionary tale about
a man who pretended to be a member of the staff and who stole some guests’
luggage. Note: Gardaí refers to the Irish police force.
Caller: …in the hostel Marian there’s one very clear practice with
hostelers and that is honesty and trust in one another.
Presenter: Yeah.
Caller: They would not take a simple tea bag unless they ask for it. Not
one.
Presenter: Right. Okay.
Caller: And once this trust is broken down hostels will cease to exist.
Presenter: Okay well I suppose it is fair to say that am the Gardaí could
pursue it but I guess … I don’t know ho= how the decision is
made. … in the order of things the people weren’t that offended
et cetera et cetera et cetera and there are drug barons et cetera
out there you know yourself how the argument goes ...
Here we find generic activity where the presenter can invoke a whole line
argument through the delexicalised category: “there are drug barons et cetera out

there you know yourself how the argument goes.” Here, she is drawing on the
shared knowledge of the caller and the audience that in the Irish media there has
been much debate as to how the Gardaí should allocate resources, for example,
whether they should prioritise serious crime issues such as drugs and criminal
gangs or whether they should invest more in basic safety for the average citizen
by following up on smaller crime issues such as this one. This is again an
example of the dynamic and collaborative nature of spoken discourse and how
dependent it is on its physical and social contexts and the shared knowledge of its
participants.
8 Conclusion
Analysis of the vague categories in this paper supports Barsalou’s (1983)
assertions about the dynamic nature of ad hoc categorisation as well as Overstreet
and Yule (1997), who stress the spontaneity of categorisation and the context-
dependent nature of the categories themselves. By looking at the reference
domains of the vague categories which are used by the participants, we find that
they index a substantial pool of shared knowledge at an Irish societal level. At a
broader level, this supports Bakhtin’s criticism that many models of linguistic
analysis have failed to understand the nature of utterances because they adopt a
passive model of meaning and understanding. They perceive language as a speech
flow from the speaker to a passive recipient instead of recognising the active role
of the other in the process of speech communication (translations of Bakhtin’s
work in Morris 1994: 80).
The range of shared “core” knowledge which was identified suggests
that the participants draw mostly on shared societal information in the context of
a national radio phone-in show and that this reflects their social relationship. The
participants are strangers to each other but they do share much societal common
ground which allows them to construct vague categories which are mutually
understood. It is also argued that this process in turn helps to create and sustain
the pseudo-intimacy that is required for this type of radio interaction, where it is
important for callers to feel part of a group that has a shared socio-cultural
background.
At a methodological level, this paper shows how corpus linguistic
methods can be used to support the initial detailed examination of forms in a
small corpus. By taking a manageable amount of data and scrutinizing it for all
forms that invoke vague categories, we have been able to identify a wider range
of forms than if we had approached it with a preset list of form types. The benefit
of this bottom-up approach is that we have been able to identify all of the vague
categories in the data as well as all of the forms used to construct them in this
specific context. Now that the forms of vague categorization have been identified
and analysed in this corpus of data, it will be possible to compare them to other
corpora. For example, preliminary research by McCarthy and O’Keeffe (2002)
looked at some of these forms in relation to a sub-corpus of the Cambridge and

Anne O’Keeffe
18
Nottingham Corpus of Discourse in English (CANCODE) which comprised
55,000 words of casual conversation data from close friends. Preliminary findings
show that certain forms are not as frequent in casual conversation between friends
and that the reference sets in the CANCODE data indexed much narrower ranges
of shared knowledge (e.g., within sub-cultural groupings and workplace shared
knowledge). In other words, a correlation between range of shared knowledge
and speaker relationship was evident.
Overall this study has shown us that by looking at a corpus of language
in use within a particular social context one can access indexical information
through patterned use of the language. In this case, we have been able to show
that vague language categorisation is a by-product of a close relationship at a
societal level of interaction.
Notes
1. Source: JNLR/MRBI radio figures released February 2003, quoted in
Oliver (2003).
2. Round brackets indicate lexical items that may co-occur and items
within square brackets are alternative but mutually exclusive (e.g., that
[kind/sort/type] of X implies that kind or sort or type of X).
References
Barsalou, L. (1983), Ad hoc categories, Memory and Cognition, 11: 211-277.
Barsalou, L. (1987), The instability of graded structure: Implications for the
nature of concepts, in U. Neisser (ed.), Concepts and conceptual
development, Cambridge: Cambridge University Press, pp. 101-140.
Bloomfield, L. (1933), Language, New York: Holt, Rinehart and Winston.
Chafe, W. (1982), Integration and involvement in speaking, writing, and oral
literature, in D. Tannen (ed.), Spoken and written language: Exploring
orality and literacy, Norwood, NJ: Ablex, pp. 35-53.
Channell, J. (1994), Vague language, Oxford: Oxford University Press.
Cruse, D.A. (1975), Hyponymy and lexical hierarchies, Archivum and
Linguisticum, VI: 26-31.
Cruse, D.A. (1977), The pragmatics of lexical specificity, Journal of Linguistics,
13: 153-164.
Dines, E. (1980), Variation in discourse-and stuff like that, Language in Society,
1: 13-31.
DuBois, S. (1993), Extension particles, etc., Language Variation and Change, 4:
179-203.
Franken, N. (1997), Vagueness and approximation in relevance theory, Journal of
Pragmatics, 28: 135-151.
Fronek, J. (1982), Thing as a function word, Linguistics, 20: 633-654

Gleeson, H.A. (1956), Introductory linguistics, Philadelphia: Chilton.
Grice, H.P. (1975), Logic and conversation, in P. Cole and J.L. Morgan (eds),
Syntax and semantics: Speech acts, New York: Academic Press, pp. 41-
58.
Hockett, C.F. (1958), A course in modern linguistics, New York: Macmillan.
Jackson, F. (2002), Language, thought and the epistemic theory of vagueness,
Language and Communication, 22: 269-279.
Jefferson, G. (1990), List construction as a task and resource, in G. Psathas (ed.),
Interaction competence, Lanham, MD: University Press of America, pp.
63-92.
Kyburg, A and M. Morreau (2000), Fitting words: vague language in context,
Language and Philosophy, 23: 577-597.
Louw, B. (1993), Irony in the text or insincerity in the writer? The diagnostic
potential of semantic prosodies, in M. Baker, G. Francis, and E. Tognini-
Bonelli (eds), Text and technology: In honour of John Sinclair,
Amsterdam: John Benjamins, pp. 157-176.
Lybbert, E.K. (1972), Substitutes as replacements, Linguistics, 91: 5-16.
Macaulay, R.K.S. (1991), Locating dialect in discourse: The language of honest
men and bonnie lasses in Ayr, New York: Oxford University Press.
Manor, R. (1995), Pragmatic considerations in semantic analyses, Pragmatics
and Cognition, 3 (2): 225-245.
McCarthy, M.J. and A. O’Keeffe (2002), Vague language and participation
framework: Indices of identity among group, culture and nation, paper
read at the 35th Annual Conference of the British Association for
Applied Linguistics, University of Cardiff, 12th -14th September, 2002.
McCarthy, M.J. and A. O'Keeffe (2003), “What's in a name?” Vocatives in casual
conversations and radio phone in calls, in P. Leistyna and C. Meyer
(eds), Corpus analysis: Language structure and language use,
Amsterdam: Rodopi, pp. 153-185.
Mervis, C.B. and E. Rosch (1981), Categorization of natural objects, Annual
Review of Psychology, 32: 89-115.
Morris, P. (ed.) (1994), The Bakhtin reader, London: Arnold.
O'Keeffe, A. (2002), Exploring indices of national identity in a corpus of radio
phone-in data from Irish radio, in A. Sánchez-Macarro (ed.), Windows
on the world: Media discourse in English, Valencia, Spain: University of
Valencia Press, pp. 91-113.
Oliver, E. (2003), No ordinary Joe, Irish Times, February, 15th
, p. 5.
Overstreet, M. and G. Yule (1997), Locally contingent categorization in
discourse, Discourse Processes, 23: 83-97.
Pinkal, M. (1985), Logic and lexicon, Dordrecht: Kluwer Academic Publishers.
Powell, M.J. (1985), Purposeful vagueness: An evaluative dimension of vague
quantifying expressions, Journal of Linguistics, 21: 31-50.
Rosch, E., C.B. Mervis, W.D. Gray, D.M. Johnson, and P. Boynes-Braem (1976),
Basic objects in natural categories, Cognitive Psychology, 2: 491-502.

Anne O’Keeffe
20
Rosch, E. (1978), Principles of categorization, in E. Rosch and B. Llyod (eds),
Cognition and categorization, Hillsdale, NJ: Lawrence Erlbaum, pp. 27-
48.
Sinclair, J. McH. (1996), The search for units of meaning, Textus, IX: 75-106.
Ward, G. and B. Birner (1992), The semantics and pragmatics of “and
everything”, Journal of Pragmatics, 19: 205-214.
Warren, M. (1993), Inexplicitness: a feature of naturalness in conversation, in M.
Baker, G. Francis, and E. Tognini-Bonelli (eds), Text and technology: In
honour of John Sinclair, Amsterdam: John Benjamins, pp. 37-53.
Williamson, T. (1994), Vagueness, London: Routeledge.
Zhang, Q. (1998), Fuzziness-vagueness-generality-ambiguity, Journal of
Pragmatics, 29: 13-31.

A Corpus-driven Analysis of the Use of Intonation to Assert
Dominance and Control
Martin Warren
The Hong Kong Polytechnic University
Abstract
This study reports on one way in which discourse intonation analysis can enrich our
understanding of spoken discourse based on work carried out across sub-corpora in the
Hong Kong Corpus of Spoken English (HKCSE). It focuses on the use of intonation by
speakers to assert dominance and control in different discourse types. Brazil (1997)
argues that the use of the rise tone is a means of asserting dominance and control at
certain points in a discourse and that while participants in conversations have the option
to freely exchange this role throughout the discourse, in spoken discourses other than
conversations such behavior would be seen to be usurping the role of the designated
dominant speaker. By means of a corpus-driven study, the use of tones is examined in data
comprised of Hong Kong Chinese speaking English with native English speakers across a
variety of discourse types in the HKCSE. The findings suggest that the choice of a certain
tone is at least partly determined by both the discourse type and the designated roles of the
speakers. It has been found that in certain kinds of discourse there are intonational
choices that are typically used predominantly, although not exclusively, by the designated
dominant speaker to assert dominance and control over the unfolding discourse. This
discourse behavior does not appear to be confined to the native speakers; both sets of
speakers in the HKCSE exhibited similar behavior in terms of their tone choices. Also, it
would appear from this study that the extent of the power vested in the designated
dominant speaker to assert dominance and control through tone choice varies across
different discourse types.
1 Introduction
The Hong Kong Corpus of Spoken English (HKCSE) is a 2-million word corpus
(i.e. approximately 200 hours) of naturally occurring spoken discourses between
adult Hong Kong Chinese and native speakers of English. The HKCSE consists
of four sub-corpora each of which represents a major spoken discourse type in the
context of Hong Kong. Briefly, the four are: conversations collected in a wide
variety of social settings, academic discourses, business discourses, and public
discourses. The participants in the discourses are all required to give their consent
prior to the recording and they are monitored in terms of age, gender, occupation,
educational background, place of birth, time spent living/studying overseas (for
the Hong Kong Chinese) and mother tongue. The HKCSE is unique in two
respects. First, it is the largest corpus of naturally-occurring spoken English
discourses compiled in Hong Kong. Second, it is both orthographically and
prosodically transcribed, which is a major task rarely undertaken with a corpus of

22 Martin Warren
this size. This study, through focusing on the use of intonation across a variety of
text types contained within the HKCSE, will in part serve to illustrate the
additional value to the corpus linguist of having a prosodically transcribed corpus
to work with. The discourse intonation system (Brazil 1985, 1997) adopted to do
the prosodic transcription is briefly outlined below.
2 Discourse Intonation
The discourse intonation system developed by Brazil (1985, 1997) and others (see
Coulthard and Brazil 1981; Coulthard and Montgomery 1981; Sinclair and Brazil
1982; Hewings 1990; Cauldwell 2002) was chosen to prosodically transcribe the
HKCSE because it is primarily concerned with the function of intonation and its
communicative value in English. This system is of particular relevance to the
researchers working with the HKCSE to further our understanding of discourse,
intercultural communication and pragmatics (see Cheng and Warren 1999,
2001a,b,c). More importantly, discourse intonation is comprised of a set of
choices available to speakers that are not formulated with reference to grammar
and do not have fixed attitudinal meanings. The latter is in contrast to those, such
as Cruttenden (1997), who seek to describe tones based on the fixed attitudes that
they convey. A study by Chun (2002: 15-45) of the various approaches to
intonation points out that discourse intonation is a break with other traditions in
the field. The approach is in contrast to those (see Chomsky and Halle 1968;
Liberman and Prince 1977; Pierrehumbert 1980) who saw rule-driven generative
phonology as a natural follow-on to their work in generative grammar. Later work
by the generative phonologists has tried to assign meaning to intonation, but the
data used were experimentally acquired for the most part (see Pierrehumbert and
Hirschberg 1990). Discourse intonation is also in opposition to those, such as
O’Connor and Arnold (1973) and Crystal (1975, 1995), who seek to describe
tones based on the fixed attitudes that they convey. In terms of the latter break
with existing approaches (see Cauldwell 1997), discourse intonation can in part be
traced back to the work of Halliday (1963, 1967) who was concerned with
developing a phonological typology based on meaning-making grammatical
choices, although in discourse intonation the link to grammatical forms has gone
(Chun 2002: 36). The choice of discourse intonation as the preferred system for
the prosodic transcription of the HKCSE is also in line with those (e.g., Couper-
Kuhlen and Selting 1996) who call for the examination of the functions of
intonation in naturally-occurring discourses to better determine their pragmatic
and situated meanings.
In Brazil’s description of discourse intonation (1997) speakers can select
from four systems: prominence, tone, key and termination. Within the prominence
system, speakers can choose to make a syllable(s) (and thus the word it is in)
prominent (i.e. by means of stress) and so indicate that it is an informative item in
that particular context. In terms of tones, speakers can basically select between
“referring” (fall-rise/rise) or “proclaiming” (fall/rise-fall) tones based on their

The Use of Intonation to Assert Dominance and Control 23
perception at that point in the discourse as to whether the information is common
ground between the participants or new. A tone being the pitch movement that
begins at the tonic syllable (i.e. the last prominent syllable in a tone unit). A
speaker’s choice of high, mid or low key (i.e. the relative pitch level of the first
prominent syllable within a tone unit) serves to indicate contrastive, additive or
equative (“as to be expected”) information, respectively. Lastly, the choice of
high, mid or low termination (i.e. the relative pitch level of the last prominent
syllable within a tone unit) impacts the subsequent interaction so that high
termination in interrogatives constrains the hearer to respond, mid termination
imposes no constraint and low termination does not predict a response. In
declaratives, the choice of high termination denotes the meaning “this will
surprise you” and mid-termination the meaning “this will not surprise you”
(Brazil 1997: 58).
All of these intonational choices, and there are thirteen in all from the
four systems described above (Hewings and Cauldwell 1997), are motivated by
real-time, situation-specific decisions by the speaker to add additional layers of
meaning to words as they are being spoken. The study presented here
concentrates on one of these systems: tone. In particular it looks at the use of the
rise tone based on the roles assigned to speakers across a range of discourse types.
The rationale for this focus is outlined in the next section.
3 The Use of Intonation to Assert Dominance and Control
In discourse intonation there are five tones that speakers may choose from. Four
of these are used to distinguish between information that is common ground and
information that is new (see Figure 1). The fifth tone is level tone which is
associated with tone units which precede an encoding pause or otherwise
truncated tone units (Brazil 1997).
fall-rise tone
“refer”: either
rise tone
Intonation choice: either
fall tone
“proclaim”: either
rise-fall tone
(Adapted from Brazil 1997: 83)
Figure 1: The referring and proclaiming tone choices available to speakers

24 Martin Warren
Within the tone system illustrated above, a speaker can choose from one
of four tones. As stated earlier, the basic choice for a speaker is between a
referring tone and a proclaiming tone depending on whether the speaker assumes
that the information is common ground between the participants or not. Once this
basic choice has been made, the speaker has a further choice between two kinds
of referring tones and two kinds of proclaiming tones. The distinction (Brazil
1997: 82-96) between the two referring tones is that the fall-rise indicates that this
part of the discourse will not enlarge the common ground assumed to exist
between the participants and the rise tone reactivates something which is part of
the common ground. In terms of the two proclaiming tones, the fall tone shows
that the area of speaker-hearer convergence is about to be enlarged while the rise-
fall tone indicates addition to the common ground and to the speaker’s own
knowledge at one and the same time (Brazil 1997).
It should be noted that a speaker's choice between selecting fall-rise or
fall tones is by no means unique to one particular kind of spoken discourse.
Brazil suggests that this is the basic choice to be found in all forms of spoken
discourse. However, there are tone choices which he characterizes as being
“participant specific” in specialized discourse types (Brazil 1985: 129-132). It is
important to point out that specialized discourses are discourses in which there is
a designated dominant speaker, and so they differ fundamentally from
conversations because conversations are a discourse type in which the speakers
have equal speaking rights. Thus in specialized discourse types there are
participant specific roles which then impact the choice available to speakers
within the discourse intonation system. These participant specific tones are the
rise tone and the rise-fall tone. The decision to choose one of these two tones is
made in terms of fall-rise/rise or fall/rise-fall and, more importantly, Brazil
explains the rationale behind these choices by describing the role relationships
pertaining between the participants in a discourse. In discourse types where one
speaker is dominant, in the sense of having greater responsibility for the discourse
and greater freedom in making linguistic choices, that speaker monopolizes the
fall-rise/rise choice. This observation would apply to the teacher in classroom
talk, the interviewer in an interview, the doctor in a doctor/patient consultation,
and so on. The rise-fall tone is by far the least prevalent of the tones, but again
Brazil claims that it tends to be the dominant speaker(s) in a discourse, in which
the participants are of unequal status, who alone makes this selection. The types
of discourse in which one participant is dominant, and thus is designated “all-
knowing” by the institutionalized relationships in force, would limit the selection
of the rise-fall tone to that participant. According to Brazil (1997), the rise-fall
tone is a very rare occurrence and, as a result, dominance and control of a
discourse through tone choice is achieved almost exclusively through speakers
choosing to use the rise tone. It is the use of this tone, therefore, that is the focus
of this study.
In conversations, however, the selection of the rise tone is not restricted
by the existence of institutionalized inequalities between the participants, and if a
speaker, for whatever reason, wishes to assert dominance and control through the

selection of this tone, he/she has the option to do so. Consequently, in
conversation the rise tone is selected by all, some or none of the participants
depending on the moment by moment decisions of those involved and not on the
basis of a restrictive set of conventions. Brazil (1985: 131) argues that in
conversation there is “an ongoing, albeit incipient, competition for dominance.”
However, he adds that this does not necessarily imply aggressiveness or rudeness
on the part of speakers, rather it can be characterized as “to remind, underline,
emphasize, insist or convey forcefulness” (Brazil 1997: 98) when a speaker
selects a rise tone, and so overtly assumes the status of the dominant speaker. The
important point is that dominant speaker status is neither predetermined nor fixed
in conversation and is typically interchangeable among the participants as the
discourse unfolds, unlike the situation with specialized discourses.
It needs to be made clear that while the words “dominance and control”
have a generally negative semantic prosody (Sinclair 1991) the fact that speakers
choose to use the rise tone to exert dominance and control locally in a discourse is
not an inherently negative behavior. While it is possible that the overuse of this
tone by a participant not deemed to be in an institutionalized dominant role might
be heard to be usurping the designated dominant speaker, this would require
repeated rather than isolated use of this tone by the speaker. Brazil (1997) argues
that the function of the rise tone is primarily to facilitate the smooth exchange
between the discourse participants. In any discourse, there is the need at times for
speakers to exert dominance and control. What is of interest here is whether this,
to an extent, is predetermined by the roles assigned to speakers in particular
discourse types.
This “corpus-driven” (Tognini-Bonelli 2002) study has several aims in
relation to an examination of the use of intonation to exert dominance and control.
First, it examines Brazil’s claim that the choice of the rise tone is monopolized by
the designated dominant speaker in discourses other than conversations. Second,
it examines whether similar patterns of tone choice are found across the two sets
of speakers in the HKCSE. Last, it examines whether the monopolizing, if it
exists, of the rise tone by the designated dominant speakers is uniform across the
different discourses drawn from the HKCSE.
Before addressing these aims, it is useful to first describe the main uses
of the rise tone in exerting dominance and control.
4 The Uses of Rise Tone
A number of controlling or dominating uses are given (Brazil 1997: 89) for the
rise tone. These are illustrated with examples from the HKCSE below.
4.1 Continuative Use of Rise Tone
The continuative use of the rise tone serves to convey to the hearer that the
speaker is underlining the expectation that she/he will be allowed to continue to

26 Martin Warren
speak (Brazil 1997). Extract 1 is taken from the sub-corpus of business discourses
in the HKCSE and is a discourse type termed informal office talk between a Hong
Kong Chinese woman and her male colleague.
Extract 1:
a1
: // for work // I’ve been to er // Fuzhou // Xiamen // and //
Nanjing // and er // for holiday // I’ve been to Shanghai //
I’ve also er // er went to er // Koasiung // in Taiwan //
(HKCSE)
Speaker a, in extract 1, is listing the places she has visited in mainland China and
Taiwan for both work and pleasure. To indicate to her colleague that her list is
ongoing she chooses the rise tone as she names each location and then ends with a
fall-rise tone. Through her choice of tone she asserts control of the discourse at
these points in order to hold on to her turn and so complete her list.
4.2 Use of Rise Tone to Exert Pressure on Hearer to Speak
Speakers can choose the rise tone in certain contexts to put pressure on the hearer
to respond to what they have said (Brazil 1997). Extracts 2 and 3 are the openings
of two separate service encounters recorded at an information counter at Hong
Kong airport involving two different female service providers interacting with
two different customers.
Extract 2:
a: // yes // can I help you //
(HKCSE)
Extract 3:
a: // good evening sir // can I help you //
(HKCSE)
In both of the above extracts the same sequence of tone choices can be seen.
Brazil (1997: 95) also gives an example of a service provider’s use of the rise
tone with can I help you and argues that the question is probably perceived as
warmer when the rise tone is used than if it was uttered with the fall-rise tone.
This example confirms that a speaker’s choice to assert dominance and control
should not be associated automatically with either positive or negative behavior.
4.3 Use of Rise Tone to Openly Remind the Hearer(s) of Common
Ground
The dominant speaker in a discourse can choose to assert dominance through the
use of the rise tone to openly assert that the hearer needs to be reminded of

something that is common ground between the participants (Brazil 1997). Extract
4 is taken from a work placement interview involving an undergraduate being
interviewed by the human resources manager of a 5-star Hong Kong hotel. The
purpose of the interview is to determine whether the student is suitable to be
placed in the hotel and, if so, which department of the hotel he will be placed in.
Extract 4:
b: // um // u I // I want to ask if I // have // any chance to work //
er // as a bartender // u or // something in the bar //
(HKCSE)
Earlier in the interview, speaker b was asked which department he would
like to work in during his placement, and he told the interviewer that he would
like to work in the bar. The interview is nearing its conclusion and the interviewer
has asked speaker b if he has any questions and above is his first question. The
interviewee repeatedly chooses the rise tone to reactivate common ground and in
so doing reminds the interviewer that he is very keen to work as a bartender or in
any other capacity in the hotel’s bar.
5 Distribution of Rise Tone across Different Discourse Types
In order to examine a cross-section of text types in the HKCSE to determine the
use of the rise tone by speakers to exert dominance and control, three discourse
types were selected: conversations, service encounters and business meetings.
These text types were chosen on the basis that, in theory, one of them, namely
conversations, does not have an institutionalized designated dominant, while the
other two do. The discourses were also chosen to have parity between the
participants in terms of the number of words spoken in order to make it possible
to make direct comparisons when analyzing the data.
For each discourse type, a search was conducted to determine the
frequency with which each participant used the rise tone when speaking, and the
results for each discourse type are tabulated below.
5.1 Conversations
The conversations contained in the HKCSE are collected in a variety of settings
such as homes, restaurants, cafés, pubs and cars. All of the participants are friends
or related to one another and regularly converse with each other in English.
Table 1 shows that the overall distribution of rise tone between the two
sets of speakers is very even. This confirms the claim by Brazil (1985) that in a
conversation any participant can choose to employ this tone in order to exert
dominance and control at specific points in the discourse. However, it needs to be
added that the fact that any participant can choose to employ the rise tone does
not mean that its use is invariably evenly spread between the participants. A

28 Martin Warren
conversation might well be dominated by one of the speakers for a variety of
possible reasons, and this can be seen in several of the conversations
(conversations 1, 2, 3 and 5) in Table 1 where one of the speakers has made
greater use of this tone than the other speaker. The important point is that this is
not pre-determined in conversations, and so across a number of conversations the
use of the rise tone becomes evened out across the two sets of speakers in the
HKCSE.
Table 1: Conversations
Conversations Hong Kong Chinese
rise tone
Native Speaker
rise tone
1 19 26
2 70 47
3 56 33
4 23 21
5 47 88
6 64 56
7 126 113
8 60 52
Total 465 441
5.2 Service Encounters
Service encounters are one of the discourse types contained in the business sub-
corpus of the HKCSE. These discourses were collected at check-in counters and
information counters at Hong Kong’s airport, hotel reception desks and retail
outlets. They all involve Hong Kong Chinese service providers interacting with
native English speaker clients/customers.
In Table 2, the use of the rise tone across the two sets of speakers can be
seen to be distributed unevenly. The general picture is that the service provider
uses the rise tone more frequently than the client/customer in a ratio of 5:3. It can
therefore be argued that in this discourse type the service provider is the more
dominant speaker in terms of the use of this tone. However, the service provider
by no means monopolizes the use of the rise tone. In four of the twenty-two
service encounters it is the client/customer who chooses the rise tone more
frequently. This might be due to the fact that the roles assigned to the participants
in service encounters are not viewed as so rigid in terms of who is the dominant
speaker.

Table 2: Service Encounters
Service Encounters Service Provider
rise tone
Client/ Customer
rise tone
1 7 11
2 2 3
3 5 0
4 10 2
5 4 2
6 1 0
7 1 0
8 2 2
9 2 1
10 17 10
11 2 1
12 3 7
13 3 0
14 1 0
15 2 0
16 4 0
17 14 7
18 7 5
19 1 0
20 2 0
21 3 0
22 10 13
TOTAL 103 64
5.3 Business Meetings
The meetings analyzed are all formal business meetings taken from the sub-
corpus of business discourses of the HKCSE. The meetings are formal in the
sense that they all had an agreed agenda with a chair responsible for the progress
of the meeting.
Table 3 shows that the chairs in the business meetings choose to employ
the rise tone more frequently than the meeting members in an overall ratio of 3:1.
Also, this pattern of behavior holds across the two sets of speakers in the HKCSE.
Of the discourse types analyzed, business meetings display the greatest disparity
between the participants when it comes to the use of the rise tone. However, even
the chairs of formal business meetings do not completely monopolize this tone to
the exclusion of the other discourse participants.

30 Martin Warren
Table 3: Business meetings
Meetings Chair
rise tone
Members
rise tone
1 (Chair: Hong Kong Chinese,
Members: 5 Hong Chinese, 1
native English speaker)
13 2
2 (Chair: Native English speaker,
Members: 3 Hong Kong
Chinese, 4 native speakers)
62 20
3 (Chair: Native English speaker,
Members: 4 Hong Kong,
Chinese 4 native speakers)
125 44
4 (Chair: Hong Kong Chinese,
Members: 3 Hong Chinese, 1
native English speaker)
89 36
TOTAL 289 102
5.4 Speaker Distribution of Rise Tone across Discourse Types
It is probably simplest to present the findings from across the three different
discourse types on a continuum (see Figure 2). At one end of the continuum the
use of the rise tone is evenly spread between the discourse particpants (i.e. 50:50
in a two-party discourse) and at the other extreme end of the continuum the use of
the rise tone is entirely monopolized by a designated dominant speaker (i.e.
100:0).
business
meetings
service
encounters
conversations
------------------------------------------------------------------------
50:50 67:33 75:25 100:0
Figure 2: Speaker distribution of rise tone
Based on the findings in this study using data drawn from the HKCSE, each of
the discourse types is plotted on the continuum in Figure 2. Conversations are at
one extreme end where the use of the rise tone is chosen equally by participants
enjoying equal staus. As we move towards the other end of the continuum we
find that the degree to which designated dominant speakers use the rise tone more
frequently than the other discourse participants steadlily increases. The first

discourse type on the continum is the service encounter, followed by the business
meeting.
A word of caution is needed at this point. While Brazil argues that the
use of the rise tone is made by speakers asserting dominance and control at
certain points in the discourse, the effect of the use of this tone is probably
cumulative. In other words the isolated use of the rise tone by a speaker might
pass unnoticed whereas repeated use might be perceived by the hearer as the
assertion of dominance and control. In the data we have examined, all speakers
choose this tone but there is a clear tendency for the rise tone to be chosen more
frequently by the designated dominant speaker. Also, there are, of course, many
other ways in which speakers can exert dominance and control in spoken
discourse (e.g., through interruptions, opening and closing the discourse, and
determining turn-taking organization) and the use of the rise tone is only one
contributing element to such behavior takes at a local level and it should not be
viewed as either the major or the determining factor.
6 Conclusions
The analyses have shown that the speakers in each of the three discourse types
exhibit different behavior in their use of the rise tone depending on their roles and
the discourse type. In all of the discourse types examined, the designated
dominant speakers have been consistently more likely to choose the rise tone and
in doing so exercise dominance and control. These findings confirm the claim
made by Brazil (1997) that in certain kinds of discourse the right to choose the
rise tone is constrained by speaker roles. However, it is also clear from the
findings that there are degrees of dominance and control from one discourse type
to another. It would appear, for example, that the role of the chair at a business
meeting might be one of greater dominance and control than that of, for example,
the service provider with regard to choosing to use the rise tone. This, it is argued,
suggests that the extent of dominance and control vested in the designated
dominant speaker is not fixed but rather seems to vary across discourse types. In
other words, there is a continuum with conversations at one end in which the
power relationship is equal and then other discourse types which have a
designated dominant speaker can be plotted on the continuum depending on the
extent of the power difference that is manifested in the speakers’ roles. Also
while no discourse type revealed a designated dominant speaker with a complete
monopoly on the use of the rise tone, there does seem to be connection between
the use of the rise tone and the participant role of designated dominant speaker.
Lastly, there is no evidence to suggest that the two sets of speakers (i.e. Hong
Kong Chinese and native English speakers) in each of the discourse types in the
HKCSE diverged with respect to their usage of the rise tone, both sets of speakers
exhibited similar behavior in terms of their tone choice in the data examined.

32 Martin Warren
Acknowledgements
The work described in this paper was substantially supported by a grant from the
Research Grants Council of the Hong Kong Special Administrative Region
(Project No. B-Q396). Thanks are due to Richard Cauldwell who has been
consultant to the project with respect to the prosodic transcription of the data.
Notes
1. Throughout the HKCSE, Hong Kong Chinese speakers are identified by
lower case letters and all other speakers by upper case letters. Females
are denoted by the letter ‘a’ and males by the letter ‘b’.
References
Brazil, D. (1985), The communicative value of intonation, Birmingham, UK:
English Language Research.
Brazil, D. (1994), Pronunciation for advanced learners of English, Cambridge:
Cambridge University Press.
Brazil, D. (1997), The communicative role of intonation in English, Cambridge:
Cambridge University Press.
Cauldwell, R.T. (1997), Tones, attitudinal meanings, and context, Speak Out!
Newsletter of the IATEFL Pronunciation Special Interest Group, 21: 30-
35.
Cauldwell, R.T. (2002), Streaming speech: Listening and pronunciation for
advanced learners of English, Birmingham, UK: Speechinaction.
Cheng, W. and M. Warren (1999), Inexplicitness: what is it and should we be
teaching it? Applied Linguistics, 20: 293-315.
Cheng, W. and M. Warren (2001a), The use of vague language in cross-cultural
conversations, English World-Wide, 22 (1): 81-104.
Cheng, W. and M. Warren (2001b), She knows more about Hong Kong than you
do isn’t it: Tags in Hong Kong conversational English, Journal of
Pragmatics, 33 (9): 1419-1439.
Cheng, W. and M. Warren (2001c), The functions of actually in a corpus of
intercultural conversations, International Journal of Corpus Linguistics,
6 (2): 257-280.
Chomsky, N. and M. Halle (1968), The sound pattern of English, New York:
Harper.
Chun, D.M. (2002), Discourse intonation in L2: From theory and research to
practice, Amsterdam: John Benjamins.
Coulthard, M. and D. Brazil (1981), The place of intonation in the description of
interaction, in D. Tannen (ed.), Analyzing discourse: Text and talk,
Washington, DC: Georgetown University Press, pp. 94-112.

Coulthard, M. and M. Montgomery (eds) (1981), Studies in discourse analysis,
London: Longman.
Couper-Kuhlen, E. and M. Selting (1996), Towards an interactional perspective
on prosody and a prosodic perspective on interaction, in E. Couper-
Kuhlen and M. Selting (eds), Prosody in conversation, Cambridge:
Cambridge University Press, pp. 11-56.
Cruttenden, A. (1997), Intonation (2nd
edition), Cambridge: Cambridge
University Press.
Crystal, D. (1975), The English tone of voice, London: Edward Arnold.
Crystal, D. (1995), The Cambridge encyclopaedia of the English language,
Cambridge: Cambridge University Press.
Halliday, M.A.K. (1963), The tones of English, Archivum Linguisticum, 15: 1-28.
Halliday, M.A.K. (1967), Intonation and grammar in British English, The Hague:
Mouton.
Hewings, M. (1990), Papers in discourse intonation, Birmingham, UK: English
Language Research.
Hewings, M. and R.T. Cauldwell (1997), Foreword, in D. Brazil, The
communicative role of intonation in English, Cambridge: Cambridge
University Press, i-vii.
Liberman, M. and A. Prince (1977), On stress and linguistic rhythm, Linguistic
Inquiry, 8: 249-336.
O’Connor J.D and G.F. Arnold (1973), Intonation of colloquial English (2nd
edition), London: Longman.
Pierrehumbert, J. (1980), The phonology and phonetics of English intonation,
Unpublished doctoral dissertation, MIT, Cambridge, MA.
Pierrehumbert, J. and J. Hirschberg (1990), The meaning of intonational contours
in the interpretation of discourse, in P. Cohen, J. Morgan, and M.
Pollock (eds), Intentions in communication, Cambridge, MA: MIT Press,
pp. 271-312.
Sinclair, J. (1991), Corpus, concordance and collocation, Oxford: Oxford
University Press.
Sinclair, J. and D. Brazil (1982), Teacher talk, Oxford: Oxford University Press.
Tognini-Bonelli, E. (2002), Functionally complete units of meaning across
English and Italian: Towards a corpus-driven approach, in B. Altenberg
and S. Granger (eds), Lexis in contrast: Corpus-based approaches,
Amsterdam: John Benjamins, pp. 73-96.

// Æ FRIENDS // ÌÊ LAdies and GENtlemen //: Some
Preliminary Findings from a Corpus of Spoken Public
Discourses in Hong Kong
Winnie Cheng
The Hong Kong Polytechnic University
Abstract
This paper describes the analysis of twelve public speeches made by The Honorable Tung
Chee-hwa, Chief Executive of Hong Kong Special Administrative Region (HKSAR)
between October and December 2001. The speeches were examined at two levels of
meaning making, namely collocational and intonational. Through the use of a
concordancer it was possible to discover, and provide evidence for, the ways in which a
public speaker constructs a relationship with the audience and the ways in which the
speaker conveyed particular meanings and ideological positions by means of making
lexico-grammatical and intonational choices, both directly and indirectly.
1 Introduction
In the English Department at the Hong Kong Polytechnic University a team has
been compiling the two-million-word Hong Kong Corpus of Spoken English
(HKCSE). The HKCSE comprises four sub-corpora: conversations, business
discourses, academic discourses and public discourses (Cheng and Warren 1999,
2000). The HKCSE is transcribed both orthographically and prosodically (see
Warren in this volume for more details). While the orthographic transcription of
spoken data is well established and the conventions quite well-known, the number
of spoken corpora that are also prosodically transcribed is very small and the
representation of prosodic features in the data is less standardized. An exception
is the London-Lund Corpus of Spoken English (Svartvik 1990).
The sub-corpus of public discourse in the HKCSE comprises 50 hours of
data recorded in Hong Kong between 2001 and 2003, including public speeches,
forum discussions, press conferences, radio and TV broadcasts, which total
approximately half a million words. This study examined one of the types of
public discourses in the HKCSE – public speeches, and specifically speeches
made on separate occasions by the Honorable Tung Chee-hwa, Chief Executive
of Hong Kong Special Administrative Region (Hong Kong SAR) between
October 2001 and December 2001.
In Hong Kong, very few studies have been conducted on public
discourses. Flowerdew (1997a,b, 1998, 2002a,b) has been a pioneer in this area
with studies that examined the speeches of Chris Patten, the last governor of
Hong Kong under the British colonial regime, as well as compared Chris Patten’s
discourse with that of Tung Chee-hwa, the Chief Executive of the Hong Kong

Winnie Cheng
36
SAR. For instance, Flowerdew (1997, 2002a) compared the themes of the
discourse of the two public figures and found that both Chris Patten and Tung
Chee-hwa focused on four themes: free market economy, freedom of the
individual, the rule of law, and democracy, and found that Tung Chee-hwa put
new emphasis on the development of a “knowledge-based” economy and less
emphasis on democratic development.
The study described in this paper focused on a smaller collection of
discourses, compared to Flowerdew, and investigated both the lexico-
grammatical and intonational features relating to Tung Chee-hwa’s public
speeches. Lexical analysis focused on how meanings, experiences and ideologies
in public speeches are encoded and conveyed directly by the choice of particular
words and indirectly by patterns of lexical co-occurrences (Stubbs 1996: 97-98).
This paper does not only discuss the use of individual words in the speeches, but
also, and more importantly, how certain words occur within different collocations
and different grammatical structures. Identifying patterns of lexical co-occurrence
in particular texts means looking for “which words collocate, and which words
occur in which grammatical constructions” (Stubbs 1996: 97-98); in other words,
the choices that a speaker makes of collocates and lexical and grammatical
patterns to produce different meanings (see Sinclair 1991; Hunston 2002;
Tognini-Bonelli 2002). The analysis of the communicative role of intonation
(Brazil 1985, 1997) represents an attempt to study the public speeches at a
different level of meaning making and expression of ideologies. Analysis of
discourse intonation aims to identify the moment by moment judgments made by
the speaker in the public speeches based on his assessment of the current state of
understanding operating between himself and his audience.
2 Description of Data
The data analyzed in this paper comprised twelve public speeches made by The
Honourable Tung Chee-hwa, Chief Executive of Hong Kong Special
Administrative Region (HKSAR), in a range of contexts between August 2001
and December 2001 (see Table 1 for details). The speeches were randomly
selected from those available during the period of time. All of the speeches except
P012, which was a program recorded from TV, were obtained from the Website
of the Hong Kong Government Information Centre. The length of the twelve
public speeches ranged from 25 minutes 36 seconds to 2 minutes 30 seconds,
totaling 127 minutes 6 seconds of recorded data.
3 Analysis of Collocates of Most Frequently-Occurring Words
The specialized corpus comprising twelve speeches by the Chief Executive was
interrogated by WordSmith Tools (Scott 1999) to generate a wordlist. Wordsmith
Tools is software designed to interrogate a corpus in a variety of ways, and this
study has used it for generating wordlists and concordance lines. The wordlist of

A Corpus of Spoken Public Discourses in Hong Kong 37
the first 50 most frequently-occurring words in the specialized corpus is shown in
the Appendix.
Table 1: Twelve public speeches of the Chief Executive of Hong Kong SAR,
October-December 2001
Code Event and date Type of speech Duration
P009 Opening ceremony of the “Tourism
Hong Kong” Exhibition at the Hong
Kong Cultural on August 24, 2001
Remarks at an
opening ceremony
4 min
P008 Reception held at Government House on
August 30, 2001 to welcome Mr.
Michael Eisner, the Chairman and Chief
Executive Officer of The Walt Disney
Company
Remarks at a
reception
5 min 30sec
P002 Welcoming Reception for the Global
Summit of Women 2001 on September
13, 2001
Remarks at a
reception
8 min 8sec
P007 Condolence remarks delivered on
September 29, 2001 at the International
Day of Remembrance organized by the
American Chamber of Commerce for the
victims of the tragic events in the United
States on September 11
Condolence
remarks at a
commemoration
2 min 30sec
P012 The Chief Executive's Policy Address
2001 on October 10, 2001
Annual policy
speech
6 min 22 sec
P001 Joint Chambers’ luncheon on October
12, 2001
Invited speech at a
luncheon
25 min 36sec
P003 Opening Dinner of the World Economic
Forum’s East Asia Economic Summit
2001 at the Hong Kong Convention and
Exhibition Center on October 29, 2001
Invited speech at a
dinner
26 min
P004 Opening Ceremony of INTEGER Hong
Kong Pavilion on November 2, 2001
Remarks at an
opening ceremony
7 min
P005 Opening Ceremony of the Twelve World
Productivity Congress on November 6,
2001
Remarks at an
opening ceremony
10 min
P006 Award Presentation Ceremony of the
2001 Hong Kong Awards for Industry
on November 19, 2001
Remarks at an
award presentation
ceremony
9 min
P010 “Hong Kong Salutes the World”
reception held at Government House at
the afternoon of December 8, 2001
Remarks at a
reception
8 min
P011 90th
anniversary dinner of the University
of Hong Kong on December 18, 2001
Remarks at a
dinner
15 min
Total 127 min 6 sec

Winnie Cheng
38
The wordlist showing the top 50 most frequently-occurring words in the
public speeches is more lexically dense than the wordlists generated by reference
corpora such as the Bank of English (Sinclair 1987). The latter typically contain
few lexical words in the first 250 most frequently-occurring words, whereas the
wordlist of public speeches has a number of such words contained in it, namely
Hong Kong, Government, world, people, economy, economic, education, and
business. Arguably, the frequent occurrence of these words is already saying that
we are more likely looking at the public discourse of a Hong Kong politician or
businessman rather than someone else operating in any domain in any city or
country.
By running the twelve computer-readable public speech texts through
the Wordsmith Tools Concordancer, the specialized corpus was examined through
focusing on the most frequently-occurring words in order to explore the
realization of the “idiom principle” (Sinclair 1991: 109-121) in the speeches. The
idiom principle is one of the main principles of the organization of language that
involves patterns of mutual word choice, two examples of which are collocation
and idiom. While both collocation and idiom represent lexical co-occurrence of
words, a collocation consists of words each of which keeps some meaning of its
own whereas an idiom gives a single unit of meaning (Sinclair 1991: 172). The
analysis was based on the premise that identifying collocations and idioms in the
speeches would shed light on the key themes running through the speeches. In
others words, the words that have a high frequency in the corpus are the most
likely source of the “idiom principle” at work.
3.1 “Government”
The most frequently-occurring lexical word – government – is used 83 times in
the public speeches. The word collocates to the left with the definite article 39
times, with the hearer presumably relying on exophoric reference to identify
which government the speaker is referring to. The internal lexical variation of the
more explicit the SAR government occurs 6 times. Interestingly, the combination
the government has a tendency to collocate to the right with the modal verb will
(13 times). This indicates that in Hong Kong, as elsewhere, there is a tendency for
the government to promise future courses of action and this is borne out by this
observation.
It is not surprising to observe that, in the context of Hong Kong, the
word government on occasion collocates with central to denote the Beijing-based
national government. This occurs 8 times in the speeches and the concordance
lines, as shown in Table 2, illustrate the use of semantic prosody (Sinclair 1991)
by the speaker.

Table 2: Sample concordance lines for central government
Sinclair (1991) uses the term “semantic prosody” to describe the ways in which
certain words have a tendency to occur in a certain semantic environment; for
example, the verb happen tends to be “associated with unpleasant things,
accidents and the like” (Sinclair 1991: 112). In this case, central government in
these public speeches has a strongly positive semantic prosody because the
central government is consistently portrayed by the speaker as a source of support,
for example, enjoy firm support of the central government, we can leverage the
support of the central government, and the strong support of the Central
Government. Positive semantic prosody is also evident in I am happy to announce
that the Central Government has agreed and we will work with central
government to expand the business.
3.2 “Educate”
According to Stubbs’ (1996: 172) analysis of the COBUILD corpus, education is
by far the most common word form of educate and typically collocates with
terms denoting kinds of institutions. In the public speeches examined in this study,
however, it has been found that while education is also the most common word
form of educate, occurring 43 times, it collocates primarily with reform (also
reforms, reforming) (12 times) and invest (also investing, investment) (11 times).
Table 3 shows sample concordance lines for education. In both cases, it is the
Hong Kong government doing the reforming and investing. For education reform,
the speeches have made it explicit that there are the series of education reform, an
all round education reform, the interface between secondary and tertiary
education and reform the academic structure, as well as further enhance our
tertiary education with systemic reforms. For investment in education, a range of
investment needs and plans are described in the speeches, for example,
investment in lifelong education, investment in formal education, and investing
heavily on education for several years. While acknowledging that investment and
reforming education is never an easy task, the Chief Executive makes the
commitment that education is a long term investment.
# Concordance
1. most generous and accommodating central government the times are indeed
2. we also enjoy firm support of the central government I firmly believe that with
3. I am happy to announce that the Central Government has agreed that foreign
4. can leverage on the support of the central government and the huge potential
5. ratic procedures we will work with central government to expand the business
6. railway already planned by the central government such a connection would
7. years with the firm support of the central government and the determination of
8. wan with the strong support of the Central Government we’ve made progress in

Winnie Cheng
40
Table 3: Sample concordance lines for education
# Concordance
1. accelerate following the series of education reform that has begun (pause) ed
2. rdingly the government spending on education has increased forty six percent in
3. help them acquire an all round education education reform is a complex pr
4. teaching and learning in our basic education system secondly we must further
5. secondary students through current education reforms second increase to sixty
6. and learning at all levels of basic education providing teacher support and tim
7. we must do so by expanding the education opportunities for our people we
8. new economy what count most are education and brain power women perform
9. for investing investment in lifelong education the little challenge that remains
10. lustrious members of Hong Kong’s educational and legislative establishments
From looking at the use of reform, reforms and reforming in the public speeches,
it has been found that these words almost invariably collocate with education to
an extent that suggests that only education can undergo this kind of process. This
point is returned to later in this paper.
3.3 “Economy”
Another word, economy, occurs 45 times and often (17 times) collocates to the
left with knowledge (14 times) or knowledge-based (3 times). This is almost the
exclusive use of the word knowledge in these speeches and is different from the
Bank of English in which, in a random selection of 40 concordance lines,
knowledge does not collocate with economy even once.
A related word form is economic which occurs 39 times. This word in
the Hong Kong SAR Chief Executive’s public speeches seems to be interesting in
terms of its semantic prosody. Often economy is neutral as in, for example,
economic fundamentals or economic activities, but can also have either positive
semantic prosody (7 times), for example, co-operation (3 times), development (3
times), growth (1 time) or more frequently negative semantic prosody (12 times)
with restructuring (7 times), downturn (4 times), uncertainties (1 time). The
semantic prosody of economic restructuring is discussed below. Table 4 shows
sample concordance lines for economic restructuring.
Table 4: Sample concordance lines for economic restructuring
# Concordance
1. in helping us through our economic restructuring and as a major provider of
2. Hong Kong to undergo economic restructuring there’s a sense of anxiety in
3. and youths in the face of economic restructuring the SAR ~overnment has to
4. ride on the challenge of economic restructuring taking Hong Kong into new
5. to weather the process of economic restructuring and finally to make the gov
6. faced with the need for economic restructuring as a result of globalization

No doubt the noun phrase economic restructuring has come to have
negative connotations in the minds of many people, and this is borne out by the
collocate to weather (line 5) and the clause the need for Hong Kong to undergo
economic restructuring there’s a sense of anxiety (line 2) in which the phrase
economic structuring occurs. Also, it is interesting that in these public speeches,
education seems consistently to undergo a process of reform but the economy
undergoes a process of restructuring. A look at 40 random concordance lines
from The Bank of English confirms that reforms are almost always associated
with the political domain whilst restructuring is a process associated with
companies, business, and so on.
3.4 “Business”
The next most frequently-occurring word is business with 39 occurrences (see
Table 5 for sample concordance lines). For the most part the word business is
used to refer to an area of work or activity with no strong collocations. The word
business collocates with, one word to the right, single occurrences of meetings,
connections, opportunities, operation systems, gateway, partners, activities, firms,
city, leaders, corporations, sector, ventures, and environment. A pattern is
observed only in business-friendly, a compound adjective, (7 times) co-occurring
with Hong Kong or city. The positive semantic prosody generated when the Hong
Kong Government and business co-occur is worth noting. In these co-occurrences,
the government is consistently portrayed as the facilitator of business through
verbs such as provide, improve, further, develop, enhance and so on.
Table 5: Sample concordance lines for business
# Concordance
1. you are visiting the Mainland for business meetings or playing golf your entry
2. people will begin to travel again and business will begin to invest again (pause)
3. ability and marketing knowledge and business connections to explore the opportun
4. ith central government to expand the business opportunities of professional servic
5. glish and Chinese are used widely in business in the provision of services and in
6. ed highly sophisticated and advanced business operations systems such as our cont
7. percent during the same period our business and indeed our society as a whole
8. hancing the presence of international business in both places many of you have as
9. Asia and have been for a long time a business gateway linking the world with the
10. a good citizenry a most supportive business sector and a most generous and acc
11. nd to make Hong Kong into a more business friendly city the government is com
12. K Integer team local developers and business partners construct a demonstration
3.5 “Hard”
There are other words in the public speeches that are interesting to investigate
although their frequencies of occurrence are not as high as those of the words just

Winnie Cheng
42
discussed. One of these words is the adjective hard. In his study of the word hard,
Sinclair (1991) found that the word attracts certain other words in strong
collocation – for example, hard work, hard luck, hard facts and hard evidence.
The search for hard in the twelve public speeches has shown that in these
political speeches, there is a strong collocation of hard, as an attributive adjective,
but only with work. Table 6 shows sample concordance lines for hard. In all
collocations of work and hard, it is the government (my colleagues and I, we) that
are, have already been or need to work hard. One might want to speculate why
the others, hard luck, hard facts and hard evidence, are absent from data based on
political speeches.
Table 6: Sample concordance lines for hard
# Concordance
1. enhancement we need to work hard to improve our hard and soft infrastructure
2. we have already been working hard and investing heavily on education for se
3. duals this we are always trying hard to do better as for example in the implem
4. secondly we must enhance the hard and soft infrastructure and generally mak
5. ance on which we are working hard to improve these include fostering a mor
6. in Hong Kong after a year of hard work the INTEGER Hong Kong Pavilion
7. economy we are also working hard to establish centres of excellence among
8. colleagues and I are working hard often burning proverbial midnight oil to
9. by corporate failures in the hard hit sectors a further decline in the stock
10. n the making and this I find it hard to overlook seeing confidence brimming
Flowerdew (2002a) investigated a large collection of speeches by Tung Chee-hwa
between July 1997 and May 2002 and found a prominent theme in this
politician’s speeches, value, which typically collocates with Chinese or
Confucian. However, in the present study, when the word value was searched for
in the twelve public speeches, a very different result has been found, which is
presented in Table 7. The collocates that were discussed in Flowerdew (2002a)
are not found in the twelve speeches at all. Instead it has been found that added
and high and the resulting idiom are related to the fields of economics and
business as evident in such collocates as economic activities, productivity,
technology-based activities and competitiveness, rather than related to culture.
So as to make better sense of the initial findings, the words education,
government, economy, economic and business were further investigated to seek to
identify the source of what Stubbs (1996: 172) terms “cultural keywords,” which
means that “words occur in characteristic collocations, which show the
associations and combinations they have, and therefore the assumptions which
they embody.” An obvious source for these cultural keywords was the Chief
Executive’s annual policy address made during the period in which the corpus
was compiled (i.e. 2001). The annual policy address is a statement of the Chief
Executive’s vision and aims for the coming twelve months and, in theory at least,
forms the basis of the Hong Kong Government’s policy agenda and overall
priorities for that period of time. In brief, “The 2001 Policy Address: Building on

Table 7: Sample concordance lines for value-added
# Concordance
1. a knowledge economy with high value added economic activities secondly w
2. a economic powerhouse where high value added economic activities is the norm
3. possibility of a sharp decline of the value of the yen a stable yen will help maint
4. services and processes that will add value to and boost the productivity of o
5. economy and continue to move up ~ value chain innovation and technology are th
6. competitiveness and achieving high value added depends very much on your abi
7. glad i’m glad to see concrete high value-added technology-based activities tak
8. improve efficiency and productivity add value and enhance the competitiveness
9. OEM production and move up the value chain promote innovation and technol
10. mote peace and prosperity through value creation performance and productivity
11. industries to move upmarket and add value we have established a five billion
our Strengths Investing in our Future” that the Chief Executive presented to the
Legislative Council on October 10, 2001 was studied in order to see whether in
effect the language used in the twelve public speeches by the Chief Executive
embodies this agenda. It has been found that topping the agenda are these themes:
“to succeed in our education reforms,” “push forward economic restructuring”
and “the transition to a new knowledge-based economy.” However, other items
that are also high on the agenda in The 2001 Policy Address, namely “a more
democratic Hong Kong,” “raising the standards of good governance” and
“investing in human talents,” are not to be found in the twelve speeches analyzed.
One interesting observation though is that all of these priorities listed do have
“the firm support of the central government.” It seems that in the subsequent
twelve public speeches, the Chief Executive has re-prioritized the list of themes
and some items have either been dropped or are on the backburner, for the time
being at least.
Table 8: Sample concordance lines for human capital
# Concordance
1. world have a clear view of their human capital structure as much as they have
2. do are five first we must invest in human capital in a sustained manner and on
3. overall game plan of investing in human capital an unfortunate fact about know
4. ment all join together to invest in human capital in a serious and sustained man
5. phasis of APEC leaders was on human capita] building because only through
6. quest for resources to a quest for human talents information technology has not
7. far sighted and a well supported human capital policy will cer- certainly succe
8. success and least economic and human dislocations and adapt ourselves to
9. ent support to acquire additional human capital that they’re likely to need well
10. our share to enhance the pool of human capital of your workforce (pause) the
11. to combine ideas and concepts human and capital resources research support
A word search was conducted for “democratic Hong Kong,” “good governance”
and “human talents” to see if these were anywhere to be found in the twelve

Winnie Cheng
44
public speeches given their emphasis in the annual policy address. However,
there was no trace of them except that “human talents” in “investing in human
talents” seems to have been reformulated as “human capital”, as in invest(ing) in
human capital, human capital building, to acquire additional human capital, and
to enhance the pool of human capital, which may very well have been the
government’s human capital policy in 2001. Table 8 above shows sample
concordance lines for human capital.
4 Analysis of the Discourse Intonation of Words and Collocates
In this study, the ways in which the speaker/politician communicates his
meanings and ideologies to his audience or the public are not only revealed
through word choices but also through discourse intonation. The system adopted
for the prosodic transcription of the data is primarily developed by Brazil (1985,
1997) and is termed “discourse intonation.” It is particularly suitable because it
enables the researcher to deal with the four different parts of the system
individually (McCarthy 1991: 144). By emphasizing the situation-specific
communicative value of intonational choices, discourse intonation is particularly
useful for those interested in examining intonation from discourse analytical or
pragmatic perspectives. The system is best introduced by looking at an example
which illustrates the various systems and choices that make up discourse
intonation as manifested in the public speeches examined in this study.
Extract 1
1. // ÌÊ the eVENTS // ÌÊ of sepTEMber eLEven // ÌÊ have CREATed
2. NEW unCERtainties // ÌÊ and acCENtuated the CYclical economic
aROUND
3. DOWNturn the world // Ì and HERE in HONG KONG //
(HKCSE)
In extract 1, the discourse can be seen to be divided up into tone units, the
boundaries of which are marked off with // at each end. A tone unit is that stretch
of the discourse which contains the “systemically-opposed” (Brazil 1997: 3)
features of intonation, and Brazil states that the tone unit is both planned by the
speakers and decoded by the hearer as a whole. What, then, are the systemically
opposed intonation features contained in the tone unit? One of the features is
prominence which is indicated in extract 1 by means of upper case letters. A
speaker has the choice to make syllables (and hence the words they are in) in a
tone unit prominent, or not, by means of stress based on situation-specific
decisions to make words informative at that point in the discourse. Thus, for
example on line 2, the speaker has chosen to make cyclical and downturn
prominent and economic non-prominent. Another feature is tone, indicated by the
arrows at the start of each tone unit. According to Brazil (1997: 68-70), the

speaker makes real-time decisions as to which tone to choose based on whether
the information is common ground between the participants or whether it
constitutes new information for the hearer. Brazil calls the tones used for the
former “referring tones” (i.e. fall-rise and rise) and the latter “proclaiming tones”
(i.e. fall and rise-fall). In extract 1, the speaker chooses fall-rise tones for the
information regarding the events of 9/11, the uncertainties they have generated,
and the world economic downturn, but the last tone unit, which includes Hong
Kong in the picture so far described, is presented to the audience as new
information since a fall tone is chosen.
Two other features are key and termination (Brazil 1997: 40) which
indicate the relative pitch (i.e. high, mid or low) chosen by the speaker. Key
choices, which are shown in the transcription by the words being above, on or
below the line, show the pitch level choice at the onset of a tonic unit. High key
selection has contrastive value, mid key has additive value and the selection of
low key has equative value, that is with the meaning “as to be expected” (Brazil
1985: 75-84). In extract 1, the speaker chooses high key on around which has the
situation-specific meaning of “this is not what you expected to hear.” Lastly,
Brazil (1997) states the speaker also chooses key at the end of the tonic segment,
and he terms this system termination. The high, mid and low termination choices
are respectively marked above, on and below the line and underlined in the
transcription. By means of this choice, the speaker can seek to constrain the next
speaker’s response if s/he selects high or mid key due to the seeming preference
for “pitch concord” (Brazil 1985: 86) found in spoken discourse across turn
boundaries. If the speaker chooses low key termination, no attempt to elicit a
response is made, leaving the next speaker to initiate a new topic or for the
discourse to come to a close. In declaratives, the choice of high termination
carries the meaning “this will surprise you” while mid termination has the
communicative value of “this will not surprise you” (Brazil 1997: 58). In the
public speeches in this study, the use of termination by the speaker is not of
interest as the discourses are monologic but not dialogic.
Having very briefly explained discourse intonation, the paper further
analyzes some of the collocates examined earlier in terms of the patterns, if any,
of intonational features associated with them, and the tone units in which they
occur. These collocates are central government, knowledge/knowledge-based
economy, economic restructuring, business friendly, and education
reform/reform/ reforming. It is believed that describing the analysis of the
discourse intonation of some of the collocates will provide further insights into
how the politician communicates particular meanings and ideologies to his
audience and the public.
First of all, in terms of the speaker’s choice of prominence, it is
interesting to note that all of the words examined were chosen to be made
prominent by the speaker, which is in line with the finding that these are
contextually important frequently-occurring words which relate to the current
political agenda of the Hong Kong government.

Winnie Cheng
46
When it comes to the speaker’s choice of tone, there is also an
interesting pattern of speaker behavior. The speaker utters these combinations of
words on 52 occasions, and on 43 of these he chooses referring tones (i.e. either
fall-rise or rise tones) for the tone units in which they are made prominent. These
tones indicate that the speaker assumes the information to be common ground
between the discourse participants. This may well reflect the reality of the
common ground between the participants or, given that Brazil (1997)
acknowledges that intonation – like many other aspects of spoken discourse – is
open to exploitation by speakers, it might be an example of a politician asserting
common ground as a tactic for providing legitimacy for what he is saying. In
other words, he gives new information as if it is common ground in order to
present it as well-known or well-established and therefore less contentious. It
would be interesting to explore this further in future research studies.
The key choices are also interesting in terms of these combinations of
words. For the most part, mid key, which is the unmarked choice with additive
communicative value, is chosen by the speaker, and there are no instances of low
key being chosen. However, there are a few instances of high key being chosen,
which might seem to be counter-intuitive given its contrastive value, when in all
other instances they are said with mid key, and all occur in tone units with
referring tones. It is therefore interesting to look at these instances of high key
choice in a little more detail.
In extract 2, the speaker selects high key on education which has a
contrastive value in the sense of going against what might be expected.
Extract 2
eduCAtion
1. // ÌÊ FOllowing the SEries // ÌÊ of reFORMS // Ì THAT
beGAN LAST year //
(HKCSE)
In this section of the speech, the speaker has just said that government
expenditure is going to accelerate, and so choosing high key on education has the
communicative value that this is against the hearer’s expectations in terms of
increased government expenditure being linked to the government’s education
reforms. For many people in Hong Kong these days, government reforms are
more likely to be associated with reduced funding rather than increased funding
and so the choice of high key is appropriate.
In extract 3, the speaker has chosen high termination when he says
restructuring. This choice of high termination is perhaps motivated by the wide
variety of possible items that could follow economic in this context. Such a key
choice is interpreted to add the situation-specific meaning of “there are many
ways in which tourism helps us economically, but here I am focusing on this
surprising one.”

Extract 3
CLEARly
1. // ÌÊ we REcognise TOUrism as a MAjor FACtor // Ê in
reSTRUCTuring
2. HELPing US THROUGH our ecoNOmic //
(HKCSE)
In extract 4, the speaker chooses high key when he is talking about the support of
the central government.
Extract 4
1. // Ì FInally we have CONfidence // Ì beCAUSE we can LEverage // Ê
on
supPORT
2. the of the central GOvernment //
(HKCSE)
Again, this use of high key carries with it the situation-specific sense that it goes
against the hearer’s expectations. This example might be thought to argue against
some of the earlier findings about the semantic prosody associated with this
speaker’s use of the words central government. However, in this context the
speaker is not saying that the support of the central government is unexpected, but
rather that what is unexpected is that this support can be leveraged by Hong Kong
to help it become a major economic center in the region.
The above analyses of the intonational choices made by the speaker with
regard to some of the combinations of words and collocates examined in this
study serve to illustrate the additional insights to be made when the corpus
linguist has access not only to the orthographic transcription of spoken discourses
but also the prosodic transcription.
There are advantages in being able to conduct an analysis of a spoken
corpus that includes aspects of discourse intonation. It is by means of discourse
intonation that speakers indicate their perceived relationship with the audience in
terms of what is assumed to be common ground, or what is enlarging the assumed
common ground, through the choice of referring or rising or falling tones.
Speakers are also able to choose to make certain lexical items prominent and thus
indicate to the hearers that they are more informationally important within the
immediate discourse context by means of stress. Lastly, the choice of relative
pitch through high, mid or low key and high, mid and low termination displays
the speaker’s evaluative stance with regard to what he/she is saying and the
audience to whom it is addressed. All of these systems within discourse
intonation carry a communicative value and also serve to create and reinforce the
relationships between the speakers and hearers. As has been pointed out, these
choices are also open to exploitation by speakers so that a politician might assert
common ground when none exists or that information falls within the hearer’s

Winnie Cheng
48
expectations when in fact it does not, and so on. These forms of exploitation are
all part and parcel of an expert discourser’s repertoire, especially one might argue
in political discourse. What remains to be investigated is the weighting to be
attached to each of these systems within discourse intonation itself and, then, the
weighting of discourse intonation as a whole in conveying the total meanings
contained in a spoken discourse. In other words, while there is no doubting that
discourse intonation has communicative value, the weight of its contribution to
meaning and relationship making has yet to be fully understood.
5 Conclusion
A few conclusions can be drawn from the findings related to collocational and
intonational analyses. First, some of the findings are less generalizable to other
contexts where public speeches are made by political leaders because those
examined in this study are specific to Hong Kong. Second, political speeches
probably are the site of collocates and idioms with a “shorter shelf life” because
political agendas are dynamic. Political discourses and perhaps other genres in the
public domain tend to transmit short-term priorities or agendas, albeit reflecting
underlying values, as opposed to longer term values and beliefs. So, for example,
we might expect the same semantic prosody to persist regarding central
government, and for reform and restructure to remain associated with particular
domains, but others, such as human capital, might not persist for long. In the
following year, for instance, health might be reformed or the economic climate
might improve so that with a word such as economic, there will be more
collocates of positive than negative semantic prosody.
The third conclusion is that politicians need to generate “buzz
collocations” and “buzz idioms” to drive home their messages, but these also
need to be reworked and repackaged from time to time, which may add to the
temporary nature of some patterns of language use in the context of political
speeches. The findings, therefore, underline the genre-specific nature of
collocations and idioms and the realization of the idiom principle.
The fourth conclusion is derived from discourse intonational analysis,
which has suggested that there are patterns of intonation choices made with
regard to the key words and collocates examined. Studies comparing the
intonation of political speeches with other monologues could well yield
interesting findings and serve to confirm or refute some of the points made here.
To conclude, the next time you overhear a politician saying, “We are
working hard with the support of the central government on our education
reforms in order to provide value-added human capital for our knowledge
economy,” you are probably listening to the Chief Executive of the Government
of the Hong Kong Special Administrative Region (SAR) making a public speech.

Exploring the Variety of Random
Documents with Different Content

"That's the worst of you," she said. "You have a fatal facility. You
have always got what you meant to get. You've never had to
struggle. Probably that means that you have never had high enough
aims. What will the world say about you in forty years?"
"Darling, it may say exactly what it pleases. If in forty years' time
there is anybody left who remembers me at all, and he tells the
truth, he will say that I enjoyed myself quite enormously. But why be
posthumous? Have another peppermint and tell me about your golf."
Edith did not have any more peppermints, so she took a cigarette
instead.
"I have a feeling that we are all going to be posthumous with regard
to our present lives long before we are dead," she remarked. "We
can't go on like this."
"I don't see the slightest reason for not doing so," said Dodo. "I
remember we talked about it one night at Winston when you fished
in my tea-gown."
"I know, and the feeling has been growing on me ever since. There
have been a lot of straws lately shewing the set of the tide."
"Which is just what straws don't do," said Dodo. "Straws float on the
surface, and move about with any tiny puff of air. Anyhow, what
straws do you mean? Produce your straws."
She paused a moment.
"I wonder if I can produce some for you," she said. "As you know, I
was to have dined with the Germans to-night and was put off. Is
that a straw? Then, again, Jack told me something this evening
about an Austrian ultimatum to Servia. Do those shew the tide you
speak of?"
"You know it yourself," said Edith. "We're on the brink of the
stupendous catastrophe, and we're quite unprepared, and we won't
attend even now. We shall be swept off the face of the earth, and if
I could buy the British Empire to-day for five shillings I wouldn't pay
it."

Dodo got up.
"Darling, I seem to feel that you lost your match at golf this
afternoon," she said. "You are always severe and posthumous and
pessimistic if that happens. Didn't you lose, now?"
"It happens that I did, but that's got nothing to do with it."
"You might just as well say that if you hit me hard in the face," said
Dodo, "and I fell down, my falling down would have nothing to do
with your hitting me."
"And you might just as well say that your dinner was put off this
evening because the Ambassador really was ill," retorted Edith.
Dodo woke next morning to a pleasant sense of a multiplicity of
affairs that demanded her attention. There was a busy noise of
hammering in the garden outside her window, for though she was
the happy possessor of one of the largest ballrooms in London, the
list of acceptances to her ball that night had furnished so unusual a
percentage of her invitations, that it had been necessary to put an
immense marquee against the end of the ballroom fitted with a
swinging floor to accommodate her guests. The big windows
opening to the ground had been removed altogether, and there
would be plenty of rhythmical noise for everybody. At the other end
of the ballroom was a raised dais with seats for the mighty, which
had to have a fresh length put on to it, so numerous had the mighty
become. Then the tables for the dinner that preceded the ball must
be re-arranged altogether, since Prince Albert, whom Dodo had not
meant to ask to dine at all, had cadged so violently on the telephone
through his equerry on Sunday afternoon for an invitation, that Dodo
had felt obliged to ask him and his wife. But when flushed with this
success he had begun to ask whether there would be bisque soup,
as he had so well remembered it at Winston, Dodo had replied icily
that he would get what was given him.

These arrangements had taken time, but she finished with them
soon after eleven, and was on her way to her motor which had been
waiting for the last half-hour when a note was brought her with an
intimation that it was from Prince Albert.
"If he says a word more about bisque soup," thought Dodo, as she
tore it open, "he shall have porridge."
But the contents of it were even more enraging. The Prince
profoundly regretted, in the third person, that matters of great
importance compelled him and the Princess to leave London that
day, and that he would therefore be unable to honour himself by
accepting her invitation.
"And he besieged me for an invitation only yesterday," she said to
Jack, "and I've changed the whole table. Darling, tell them to alter
everything back again to what it was. Beastly old fat thing! Really
Germans have no manners.... Daddy has been encouraging him too
much. If he rings up again say we're all dead."
Dodo instantly recovered herself as she drove down Piccadilly. The
streets were teeming with happy, busy people, and she speedily felt
herself the happiest and busiest of them all. She had to go to her
dressmakers to see about some gowns for Goodwood, and others
for Cowes; she had to go to lunch somewhere at one in order to be
in time for a wedding at two, she had to give half an hour to an
artist who was painting her portrait, and look in at a garden party.
Somehow or other, apparently simultaneously, she was due at the
rehearsal of a new Russian ballet, and she had definitely promised to
attend a lecture in a remote part of Chelsea on the development of
the sub-conscious self. Then she was playing bridge at a house in
Berkeley Square—what a pity she could not listen to the lecture
about the sub-conscious self while she was being dummy—and it
was positively necessary to call at Carlton House Terrace and enquire
after the German Ambassador. This latter errand had better be done
at once, and then she could turn her mind to the task of simplifying
the rest of the day.

There were entrancing distractions all round. She was caught in a
block exactly opposite the Ritz Hotel, and cheek to jowl with her
motor was that of the Prime Minister, and she told him he would be
late for his Cabinet meeting. He got out of the block first by shewing
an ivory ticket, and Dodo consoled herself for not being equally well-
equipped by seeing a large flimsy portmanteau topple off a luggage
trolley which was being loaded opposite the Ritz. It had a large
crown painted on the end of it in scarlet, with an "A" below, and it
needed but a moment's conjecture to feel sure that it belonged to
Prince Albert. Whatever was the engagement that made him leave
London so suddenly, it necessitated an immense amount of luggage,
for the trolley was full of boxes with crowns and As to distinguish
them. The fall had burst open the flimsy portmanteau, and shirts
and socks and thick underwear were being picked off the roadway....
Dodo wondered as her motor moved on again if he was going to
quarter himself on her father for the remainder of his stay in
England.
A few minutes later she drew up at the door of the German
Embassy, and sent her footman with her card to make enquiries.
Even as he rang the bell, the door opened, and Prince Albert was
shewn out by the Ambassador. The two shook hands, and the Prince
came down the three steps, opposite which Dodo's motor was drawn
up. It was open, there could have been no doubt about his seeing
her, but it struck her that his intention was to walk away without
appearing to notice her. That, of course, was quite impermissible.
"Bisque soup," she said by way of greeting. "And me scouring
London for lobsters."
He gave the sort of start that a dramatic rhinoceros might be
expected to give, if it intended to carry the impression that it was
surprised.
"Ah, Lady Dodo," said he. "Is it indeed you? I am heartbroken at not
coming to your house to-night. But the Princess has to go into the
country; there was no getting out of it. So sad. Also, we shall make
a long stay in the country; I do not know when we shall get back. I

will take your humble compliments to the Princess, will I not? I will
take also your regrets that you will not have the honour to receive
her to-night. And your amiable Papa; I was to have lunched with him
to-day, but now instead I go into the country. And also, I will step
along. Auf wiedersehen, Lady Dodo."
Suddenly a perfect shower of fresh straws seemed to join those
others which she and Edith had spoken about last night, and they all
moved the same way. There was the note which she had received
half an hour ago saying that the Prince could not accept the
invitation he had so urgently asked for; there was the fact of those
piles of luggage leaving the Ritz; there was his call this morning at
the German Embassy, above all there was his silence as to where he
was going and his obvious embarrassment at meeting her. The tide
swept them all along together, and she felt she knew for certain
what his destination was.
"Good-bye, sir," she said. "I hope you'll have a pleasant crossing."
He looked at her in some confusion.
"But what crossing do you mean?" he said. "There is no crossing
except the road which now I cross. Ha! There is a good choke, Lady
Dodo."
Dodo made her face quite blank.
"Is it indeed?" she said. "I should call it a bad fib."
She turned to her footman who was standing by the carriage door.
"Well?" she said.
"His Excellency is quite well again this morning, my lady," he said.
That too was rather straw-like.
"Drive on," she said.

Just as impulse rather than design governed the greater part of
Dodo's conduct, so intuition rather than logic was responsible for her
conclusions. She had not agreed last night with Edith's reasonings,
but now with these glimpses of her own, she jumped to her
deduction, and landed, so to speak, by Edith's side. As yet there was
nothing definite except the unpublished news of an Austrian
ultimatum to Servia, and the hurried meeting of the Cabinet this
morning to warrant grounds for any real uneasiness as to the
European situation generally, nor, as far as Dodo knew, anything
definite or indefinite to connect Germany with that. But now with the
fact that her dinner had been put off last night and the ambassador
was quite well this morning, coupled with her own sudden intuition
that the Allensteins were going back post-haste to Germany, she
leaped to a conclusion that seemed firm to her landing. In a flash
she simply found herself believing that Germany intended to provoke
a European war.... And then characteristically enough, instead of
dwelling for a moment on the menace of this hideous calamity or
contemplating the huge unspeakable nightmare thus unveiled, she
found herself exclusively and entrancedly interested in the situation
as it at this moment was. She expected the entire diplomatic world,
German and Austrian included, at her ball that night; already the
telegraph wires between London and the European capitals must be
tingling and twitching with the cypher messages that flew backwards
and forwards over the Austrian ultimatum, and her eyes danced with
anticipation of the swift silent current of drama that would be
roaring under the conventional ice of the mutual salutations with
which diplomatists would greet each other this evening at her house.
Hands unseen were hewing at the foundations of empires, others
were feverishly buttressing and strengthening them, and all the
hours of to-night until dawn brought on another fateful day, those
same hands, smooth and polite, would be crossing in the dance, and
the voices that had been dictating all day the messages with which
the balance of peace and war was weighted, would be glib with little
compliments and airy with light laughter. She felt no doubt that
Germans and Austrians alike would all be there, she felt also that the
very strain of the situation would inspire them with a more elaborate

cordiality than usual. She felt she would respect that; it would be
like the well-bred courtesies that preceded a duel to the death
between gentlemen. Prince Albert, it is true, in his anxiety to get
back without delay to his fortressed fatherland had failed in the
amenities, but surely Germany, the romantic, the chivalrous, the
mother of music and science, would, now and henceforth, whatever
the issue might be, prove herself worthy of her traditions.
Once more Dodo was caught in a block at the top of St. James's
street, and she suddenly made up her mind to stop at the hotel and
say good-bye to Princess Albert. Two motives contributed to this, the
first being that though she and he alike had been very rude throwing
her over with so needless an absence of ceremony and politeness,
she had better not descend to their level; the second, which it must
be confessed was far the stronger, being an overwhelming curiosity
to know for certain whether she was right in her conjecture that
they were going to get behind the Rhine as soon as possible.
Dodo found the Princess sitting in the hall exactly opposite the
entrance, hatted, cloaked, umbrellaed and jewel-bagged, with a
short-sighted but impatient eye on the revolving door, towards
which, whenever it moved, she directed a glance through her
lorgnette. As Dodo came towards her, the Princess turned her head
aside, as if, like her husband, seeking to avoid the meeting. But next
moment, even while Dodo paused aghast at these intolerable
manners, she changed her mind, and dropping her umbrella, came
waddling towards her with both hands outstretched.
"Ah, dear Dodo," she said, "I was wondering, just now I was
wondering what you thought of me! I would have written to you, but
Albert said 'No!' Positively he forbade me to write to you, he called
on me as his wife not so to do. Instead he wrote himself, and such a
letter too, for he shewed it me, all in the third person, after he had
asked for bisque soup only yesterday! And I may not say good-bye
to your good father or anyone; you will all think I do not know how
to behave, but I know very well how to behave; it is Albert who is so
boor. I am crying, look, I am crying, and I do not easily cry. We have

said good-bye and thank you to nobody, we are going away like
burglars on the tiptoe for fear of being heard, and it is all Albert's
fault. In five minutes had our luggage to be packed, and there was
Albert's new portmanteau which he was so proud of for its
cheapness and made in Germany, bursting and covering Piccadilly
with his pants, is it, that you call them? It was too screaming. I
could have laughed at how he was served right. All Albert's pants
and his new thick vests and his bed-socks being brought in by the
porter and the valets and the waiters, covered with the dust from
Piccadilly!"
"Yes, ma'am, I saw it myself," said Dodo, "when I was passing half
an hour ago."
The Princess was momentarily diverted from the main situation on to
this thrilling topic.
"Ach! Albert would turn purple with shame," she said, "if he knew
you had seen his pants, and yet he is not at all ashamed of running
away like a burglar. That is his Cherman delicacy. 'Your new bed-
socks,' I said to him, 'and your winter vests and your pants you must
have made of them another package. They will not go in your new
portmanteau; there is not room for them, and it is weak. It has to go
in the train, and again it has to go on the boat, and also again in the
train.' It is not as if we but went to Winston—ah, that nice Winston!
—but we go to Chermany. That is what I said, but Albert would not
hear. 'By the two o'clock train we go,' he said, 'and my new vests
and my socks and my pants go in my new Cherman portmanteau
which was so cheap and strong.' But now they cannot go like that,
and they will have to go in my water-proof sheet which was to keep
me dry on the boat from the spray, for if I go in the cabin I am ill. It
is all too terrible, and there was no need for us to go like this. We
should have waited till to-morrow, and said good-bye. Or perhaps if
we had gone to-morrow we should not go at all. What has
Chermany to do with Servia, or what has England either? But no, we
must go to-day just because there have been telegrams, and Cousin
Willie says, 'Come back to Allenstein.' And here am I so rude

seeming to all my friends. But one thing I tell you, dearest Dodo; we
chiefly go, because Albert is in a Fonk. He is a Fonk!"
"But what is he frightened of?" asked Dodo. The Princess was letting
so many cats out of the bag that she had ceased counting them.
"He is frightened of everything. He is frightened that he will be
pelted in the London streets for being a Cherman prince, just as if
anybody knew or cared who he was! He is frightened of being put in
prison. He is frightened that the Cherman fleet will surround England
and destroy her ships and starve her. He is frightened of being
hungry and thirsty. He is as a pig in a poke that squeals till it gets
out."
This remarkable simile was hardly out of the Princess's mouth before
she squealed on her own account.
"Ach, and here he is," she said. "Now he will scold me, and you shall
see how I also scold him."
He came lumbering up the passage towards them with a red, furious
face.
"And what did I tell you, Sophy?" he said. "Did I not tell you to sit
and wait for me and speak to no one, and here are you holding the
hand of Lady Dodo, to whom already I have said good-bye, and so
now I do not see her. It is done, also it is finished, and it is time we
went to the station. You are for ever talking, though I have said
there shall be no more talking. What have you been saying?"
Princess Albert still held Dodo's hand.
"I have been saying that your new portmanteau burst, and I must
take your vests and your socks and your pants in my water-proof
sheet. Also I have been saying——"
"But your water-proof sheet, how will your water-proof sheet hold all
that was in my portmanteau? It is impossible. Where is your water-
proof sheet? Show it me."

"You will see it at Charing Cross. And if it is wet on the boat I will
take out again your vests and your socks and your pants, and they
may get wet instead of me."
"So! Then I tell you that if it is wet on the boat, you will go to your
cabin, and if you are sick you will be sick. You shall not take my
clothes from your water-proof sheet."
"We will see to that. Also, I have been saying good-bye to dearest
Dodo, and I have been saying to her that it was not I who was so
rude to her, but also that it was you, Albert. And I say now that I
beg her pardon for your rudeness, but that I hope she will excuse
you because you were in a fonk, and when you are in a fonk, you no
longer know what you do, and in a fonk you will be till you are safe
back in Germany. All that I say, dearest Albert, and if you are not
good I will tell it to the mob at Charing Cross. I will say, 'This is the
Prince of Allenstein, and he is a Prussian soldier, and therefore he is
running away from England.' Do not provoke me, heart's dearest.
You will now get them to send for a cab, and we will go because you
are a fonk. There will be no special train for us, there will be no one
of our cousins to see us off, there will be no red carpet, and it is all
your fault. And as for dearest Dodo, I kiss her on both cheeks, and I
thank her for her kindness, and I pray for a happier meeting than is
also our parting."
That afternoon there began to be publicly felt the beginning of that
tension which grew until the breaking-point came in the first days of
August, and but for Dodo's shining example and precept, her ball
that night might easily have resolved itself into a mere conference.
Again and again at the beginning of the evening the floor was empty
long after the band had struck up, while round the room groups of
people collected and talked together on one subject. But Dodo
seemed to be absolutely ubiquitous, and whenever she saw earnest
conversationalists at work, she plunged into the middle of them, and

broke them up like a dog charging a flock of sheep. To-morrow
would do for talk, to-night it was her ball. Her special prey was any
group which had as its centre an excited female fount of gossip who
began her sentences with "They tell me...." Whenever that fatal
phrase caught Dodo's remarkably sharp ears, she instantly led the
utterer of it away to be introduced to someone on the great red dais,
managed to lose her in the crowd, and "went for" the next offender.
The rumour that the Allensteins had left Charing Cross that
afternoon for Germany was a dangerously interesting topic, and
whenever Dodo came across it, she strenuously denied it, regardless
of truth, and asserted that as a matter of fact they were going down
again to-morrow to stay with her father at Vane Royal. Then
perceiving him not far off, looking at the dais with the expression of
Dante beholding the Beatific vision, she had dived into the crowd
again, and told him that if he would assert beyond the possibility of
contradiction that this was the case, she would presently introduce
him to anyone on the red dais whom he might select. As he
pondered on the embarrassment of such richness, she was off again
to break up another dangerous focus of conversation.
An hour of wild activity was sufficient to set things really moving,
and avert the danger of her ball becoming a mere meeting for the
discussion of the European situation, and presently she found five
minutes rest in the window of the music gallery from which she
could survey both the ballroom and the marquee adjoining it. In all
her thirty years' experience, as hostess or guest, she had never been
present at a ball which seemed quite to touch the high-water mark
here, and she felt that without Lord Cookham's assistance she had
provided exactly the sort of evening that he had designed, in honour
of Jumbo. It had happened like that; everybody was present in that
riot of colour and rhythm that seethed about her, and at the moment
the dais which stretched from side to side of the huge room was
empty, for every one of its occupants was dancing, and she
observed that even Lord Cookham (who had come in an official
capacity) had deserted his place behind the row of chairs, and was
majestically revolving with a princess, making little obeisances as he

cannoned heavily into other exalted personages. The whole of the
diplomatic corps was there, German and Austrian included, and
there was the German ambassador, quite recovered from his curious
indisposition, waltzing with the Italian ambassadress. The same
spirit that had animated Dodo in breaking up serious conjectures
and conversation seemed now to have spread broadcast; all were
conspirators to make this ball, the last of the year, the most brilliant
and memorable. From a utilitarian point of view there was no more
to be said for it than for some gorgeously-plumaged bird that
strutted and spread its jewelled wings, and yet all the time it was a
symbol, expressing not itself alone but what it stood for. The glory of
great names, wide-world commerce, invincible navies, all the
endorsements of Empire, lay behind it. It glittered and shone like
some great diamond in an illumination which at any moment might
be obscured by the menace of thundercloud, but, if this was the last
ray that should shine on it before the darkness that even now lapped
the edge of it enveloped it entirely, that gloom would but suck the
light from it, and not soften nor crush its heart of adamant....
From the moment that the ball got moving Dodo abandoned herself
to enjoying it to the utmost, wanting, as was characteristic of her, to
suck the last-ounce of pleasure from it. She had that indispensable
quality of a good hostess, namely, the power of making herself the
most fervent of her guests, and never had she appreciated a ball so
much. Not until the floor was growing empty and the morning light
growing vivid between the chinks of closed curtains did she realise
that it was over.
"Jumbo, dear," she said, "why can't we double as one does at
bridge, and then somehow it would be eleven o 'clock last night, and
we should have it all over again? Are you really going? What a pity!
Stop to breakfast—my dear, what pearls! I can't believe they're real
—and don't let us go to bed at all. Yes, do you know, it's quite true—
though I've been lying about it quite beautifully—the Allensteins left

for Germany this afternoon, I mean yesterday afternoon. Oh, I don't
want to begin again.... What will the next days bring, I wonder?"
She stood at the street door a moment, while he went out into that
pregnant and toneless light that precedes sunrise, when all things
look unreal. The pavement and road outside were pearly with dew,
and the needless head-light of his motor as it purred its way up to
the door gleamed with an unnatural redness. In the house the floor
was quite empty now and the band silent, a crowd of men and
women eager to get away besieged the cloakroom, and in ten
minutes more Dodo found herself alone, but for the servants already
beginning to restore the rooms to their ordinary state.
She felt suddenly tired, and going upstairs drew down the blinds
over her open windows. She wanted to get to sleep at once, to shut
out the dawning day and all that it might bring.

CHAPTER VII
DODO'S APPRENTICESHIP
The morning papers were late that day, and when they arrived Dodo
snatched at them and automatically turned to the communiqué from
the French front. There was a list of names of villages which had
been lost to the allies, but these were unfamiliar and meant nothing
to her. Then she looked with a sudden sinking of the heart at the
accompanying map which shewed by a black line the new position of
the front, and that was intelligible enough. For the last fortnight it
had been moving westwards and southwards with regular and
incredible rapidity like the advance of some incoming tide over level
sands. Occasionally for a little it had been held up, but the flood,
frankly irresistible, always swept away that which had caused the
momentary check.... In the next column was an account of German
atrocities compiled from the stories of Belgian refugees.
Dodo had come back to London last night from Winston where she
had been seeing to the conversion of the house into a Red Cross
hospital, and just now she felt, like some intolerable ache, the sense
of her own uselessness. All her life she had found it perfectly easy to
do the things which she wanted to do, and she had supposed herself
to be an efficient person. But now, when there was need for efficient
people, what did her qualifications amount to? She could ride, as
few women in England could ride, she was possessed of enormous
physical and nervous energy, she was an inimitable hostess, could
convert a dull party into a brilliant one by the sheer effortless out-
pouring of her own wit and infectious vivacity, but for all practical

purposes from organisation down to knitting, she was as useless as
a girl straight out of the nursery where everything had been done for
her by assiduous attendants. She was even more useless than such
a child, for the child at any rate had the adaptability and the power
of learning appropriate to its age, whereas Dodo, as she had lately
been ascertaining, had all her life been pouring her energy down
certain definite and now useless channels. In consequence those
channels had become well-worn; her energy flowed naturally with
them, and seemed to refuse to be diverted, with any useful result,
elsewhere. She could ride, she could play bridge, she could, as she
despondently told herself, talk the hind leg off a donkey, she could
entertain and be entertained till everyone else was dying to go to
bed. And no one wanted her to do any of those things now; there
was absolutely no demand for them. But when it came to knitting a
stocking herself, or being personally responsible for a thing being
done, instead of making a cook or a groom or a butler responsible
for it, she had no notion how to set about it.
Very characteristically when David's nurse had announced her
intention of being trained for hospital work, Dodo had warmly
congratulated her determination, had given her an enormous tip,
and had bundled her off to the station in a prodigious hurry, saying
that she would look after David herself. But the things that a small
boy required to have done for him filled her with dismay at her own
incompetence, when she had to do them. If he got his feet wet,
fresh socks had to be found for him; if his breeches were covered
with short white hairs from his ride, these must be brushed off;
buttons had to be replaced; there was no end to these ministrations.
Dodo could not get on at all with the stocking she was knitting or
the supervision of the storing of the furniture at Winston, while she
had to produce a neat daily David, and incidentally failed to do so.
She advertised for another nurse without delay, and David was
exceedingly relieved at her arrival.
Dodo was, luckily, incapable of prolonged despair with regard to her
own shortcomings, and by way of self-consolation her thoughts
turned to the fact that before she left Winston she had contrived and

arranged a charming little flat in a wing of the house for herself and
David and Jack whenever he could find time to come there, for he
was in charge of a remount camp, knowing, as he certainly did, all
that was to be known about horses from A to Z. Dodo's mind harked
back for a moment to her own uselessness in envious contemplation
of the solid worth, in practical ways, of her husband's knowledge.
For herself, through all these frivolous years she had been content
with the fact of her consummate horsemanship; she had hands, she
had a seat, she had complete confidence (well-warranted) in her
ability to manage the trickiest and most vicious of four-legged
things. There her knowledge (or rather her instinct) stopped,
whereas Jack, a mere lubber on a horse compared with herself, was
a perfect encyclopædia with regard to equine matters of which she
was profoundly ignorant. He could "size up" a horse by looking at it,
in a way incomprehensible to Dodo; he knew about sore backs and
bran mashes and frogs and sickle-hocks, and now all the lore which
she had never troubled to learn any more than she had troubled to
decipher a doctor's prescription and understand its ingredients, was
precisely that which made Jack, at this crisis when efficiency was
needed, so immensely useful.... However, after all, she had been
useful too, for she had planned that delicious little flat at Winston
(necessary, since the house was to be made into a hospital), which
would give accommodation to them. Everything, of course, was
quite simple; she had put in two bathrooms with the usual
paraphernalia of squirts and douches and sprays, and had converted
a peculiarly spacious pantry into a kitchen with a gas-stove and
white tiled walls. Naturally, since the house was no longer habitable,
this had to be done at once, and her energy had driven it through in
a very short space of time. The expense had been rather staggering,
especially in view of the cost of running a hospital, so Dodo had sent
the bill to her father with a lucid explanatory letter.
The thought of this delicious little flat, which would be so economical
with its gas-stove for cooking, and its very simple central heating, in
case, as Jack gloomily prognosticated, there should be difficulties
about coal before the war was over, made Dodo brighten up a little,

and diverted her thoughts from the on-creeping barbarous tide in
France, and the sense of her own uselessness. After all somebody
had to contrive, to invent, even though plumbers and upholsterers
effected the material conversion, and Daddy paid the bill; and she
had come up to town in order to superintend a similar change at
Chesterford House. That was to be turned into a hospital for officers,
and Dodo was determined that everything should be very nice. The
ballroom would be a ward, so also would be the biggest of the three
drawing-rooms, but the dining-room had better be left just as it was,
in anticipation of the time when the invalids could come down to
dinner again. She intended to keep a couple of rooms for herself,
and one for her maid, since she could not be at Winston all the year
round.... And then suddenly she perceived that behind all her
charitable plans there was the reservation of complete comfort for
herself. It cost her nothing, in the personal sense, to live in a wing at
Winston and a cosy corner of the house in London. There was not
an ounce of sacrifice about it all, and yet she had read with a certain
complacency that very morning, that Lord and Lady Chesterford had
set a noble example to the rest of the wealthy classes, in giving up
not one only but both of their big houses. But now all her
complacency fell down like a house of cards. Jack certainly had given
up something, for his day was passed in real personal work.... He
was on the staff with a nice red band on his cap, and tabs on his
shoulders and spurs. And here, even in the moment that she was
damning her own complacency, she was back in the old rut, thinking
about signs and decorations instead of what they stood for. There
was the black line of the tide creeping over France, and three
columns of casualties in the morning's paper, and one of German
atrocities....
Dodo was expecting Edith to lunch, and since the chef had gone
back to France to rejoin the colours, there was only a vague number
of kitchen-maids, scullery-maids and still-room maids in the house to
manage the kitchen, and even these were being rapidly depleted,
as, with Dodo's cordial approval, they went to canteens and other
public services. She had, in fact, warned Edith only to expect a

picnic, and she thought it would be more picnicky if they didn't go to
the dining-room at all, but had lunch on a table in her sitting-room.
This did not, as a matter of fact, save much trouble, since the
dining-room was ready, and a table had to be cleared in her sitting-
room, but Dodo at the moment of giving the order was on the
dramatic "stunt," and when Edith arrived there was a delicious little
lunch in process of arrival also.
"Darling, how nice of you to come," said Dodo, "and you won't mind
pigging it in here, will you? Yes, let's have lunch at once. The chef's
gone, the butler's gone, and I shall have parlour-maids with white
braces over their shoulders. My dear, I haven't seen a soul since I
left Winston yesterday, and I haven't seen you since this thunderbolt
burst. Do they burst, by the way? All that happened before the
fourth of August seems centuries away now. I can only dimly
remember what I used to be like. A European war! For ten years at
least that has been a sort of unspeakable nightmare, which nobody
ever really believed in, and here we are plunged up to the neck in
it."
Edith seemed to have something in reserve.
"Go on," she said, helping herself to an admirable omelette. "I want
to know how it affects you."
Dodo finished her omelette in a hurry, and drew a basket full of wool
and knitting needles from under the table. Out of it she took a long
sort of pipe made of worsted. She made a few rapid passes with her
needles.
"I have been frightfully busy," she said. "If I'm not busy all the time
I begin wondering if any power in heaven or earth can stop that
relentless advance of the Germans. The French government are
evacuating Paris, and then I ask myself what will happen next? What
about the Channel ports? What about the Zeppelins that are going to
shower bombs on us? And then by the grace of God I stop asking
myself questions which I can't answer, and occupy myself in some
way. I have been terrifically busy at Winston, clearing all the house

out for the hospital we are having there, and just making a small
habitable corner for David and Jack and me at the end of the east
wing, do you remember, where the big wisteria is. Central heating,
you know, because Jack says there will be no coal very soon, and
my darling Daddy is going to pay the bill. Then I came up here,
because this house is to be a hospital for officers——"
Dodo suddenly threw her hands wide with a gesture of despair.
"Oh, how useless one is!" she said. "I know quite well that my
housekeeper could have done it all with the utmost calmness and
efficiency in half the time it took me. When I was wildly exciting
myself about blocking up a door in my room at Winston, so as not to
have vegetable-smells coming up from the kitchen, and thinking how
tremendously clever I was being, she waited till I had quite finished
talking, and then said, 'But how will your ladyship get into your
room?' And it's the same with this awful stocking."
Dodo exhibited her work.
"Look!" she said, "the leg is over two feet long already, and for three
days past I have been trying to turn the heel, as the book says, but
the heel won't turn. The stocking goes on in a straight line like a
billiard cue. I can never do another one, so even if the heel was kind
enough to turn now, I should have to advertise for a man at least
seven feet high who had lost one leg. The advertisement would cost
more than the stocking is worth, even if it ever got a foot to it.
Failing the seven-foot one-legged man, all that this piece of worsted-
tubing can possibly be used for, is to put outside some exposed
water-pipe in case of a severe frost. Even then I should have to rip it
up from top to bottom to get it round the pipe, or cut off the water-
supply and take the pipe down and then fit the stocking on to it.
Then again when David's nurse left, I said I would look after him.
But I didn't know how; the nervous force and the time and the
cotton and the prickings of my finger that were required to sew on a
button would have run a tailor's shop for a week. Oh, my dear, it's
awful! Here is England wanting everything that a country can want,
and here am I with hundreds of other women absolutely unable to

do anything! We thought we were queens of the whole place, and
we're the rottenest female-drones that ever existed. Then again I
imagined I might be able to do what any second-rate housemaid
does without the smallest difficulty, so when other people had taken
up the carpet on the big stairs at Winston, I sent four or five
servants to fetch me a broom, so that I could sweep the stairs. They
were dusting and fiddling about in the way housemaids do, and they
all grinned pleasantly and stopped their work to fetch me something
to sweep the stairs with. I supposed they would bring me an
ordinary broom, but they brought a pole with a wobbly iron ring at
the end of it, to which was attached a sort of tow-wig. I didn't like to
ask them how to manage it, so I began dabbing about with it. And
at that very moment the grim matron leaned over the bannisters at
the top of the stairs and called out, 'What are you doing there? You
look as if you had never used a mop before!' I hadn't; that was the
beastly part of it, and then she came down and apologised, and I
apologised and she shewed me what to do, and I hit a housemaid in
the eye and hurt my wrist, and dislocated all work on that stair-case
for twenty minutes. And then I tried to weigh out stores as they
came in, and I didn't know how many pennies or something went to
a pound Troy. And you may be surprised to hear that a hundred-
weight is less than a quarter, or if it's more it isn't nearly so much
more as you would think. I'm useless, and I always thought I was so
damned clever. All I can do is to play the fool, and who wants that
now? All my life I have been telling other people to do things,
without knowing how to do them myself. I can't boil a potato, I can't
sew on a button, and yet I'm supposed to be a shining light in war-
work. 'Marquez mes mots,' as the Frenchman never said, they'll soon
be giving wonderful orders and decorations to war-workers, and
they'll make me a Grand Cross or a Garter or a Suspender or
something, because I've made a delicious flat for myself in the
corner of Winston, and sent the bill in to Daddy, and will be going
round the wards at Winston and saying something futile to those
poor darling boys who have done the work."

Dodo held up a large piece of hot-house peach on the end of her
fork.
"Look at that, too," she said. "I'm an absolute disgrace. Fancy eating
hot-house peaches in days like these!"
Edith had rather enjoyed certain parts of Dodo's vivacious summary
of herself, but the most of it caused her to snort and sniff in violent
disagreement. Once or twice she had attempted to talk too, but it
was no use till Dodo had blown off the steam of her self-
condemnation. Now, however, she took up her own parable.
"Wouldn't you think it very odd of me," she said in a loud voice, "if I
began writing epic poems?"
"Yes, dear, very odd," said Dodo.
"It wouldn't be the least odder than you trying to sew on buttons or
washing David. You are just as incapable of that as I am of the
other. You only waste your time; you never learned how, so why on
earth should you know how? We're all gone perfectly mad; we're all
trying to do things that are absolutely unsuited to us. I really believe
I'm the only sane woman left in England. Since the war began I
have devoted myself entirely to my music, and I've written more in
these last few weeks than I have during a whole year before. There
have been no distractions, no absurd dances and dinners. I've been
absolutely uninterrupted. Bertie has been taken on for the London
Defence against Zeppelins. He has never seen a Zeppelin and knows
as much about defences as I know about writing sonnets; and
Madge pours out the most awful tea and coffee on the platform at
Victoria. She never could pour anything out; if she was helping
herself to a cup of tea she flooded the tray, and I should think that
in a few days Victoria station will be entirely submerged. That will
mean that troops will have to reach their trains in London by means
of rafts."
"But one can't help doing something," said Dodo. "One can't go on
being useless."

"You don't mend it by being worse than useless. That's why I devote
myself to music. I can do that, and I can't do any of the things that
everybody else is trying to do."
Edith paused a moment.
"There's another reason, too," she said. "I should go off my head if I
wasn't busy about something. I wish there was such a thing as a
clinical thermometer of unhappiness, and you would see how utterly
miserable I am. You can't guess what being at war with Germany
means to me. All that is best in the world to me comes from
Germany; all music comes from there. And yet last night when I was
playing a bit of Brahms, Bertie said, 'Oh, do stop that damned Hun
tune!' Why, there's no such thing as a Hun tune! Music is simply
music, and with a few exceptions the Huns, as he loves to call them,
have made it all."
"He calls them Huns," said Dodo carefully, "because they've already
proved themselves the most infamous barbarians. Did you see the
fresh atrocities in the Times this morning?"
"I did, and I blushed for the wickedness of the people who invented
them and the credulity of the people who believed them. They can't
be true. I know the Germans, and they are incapable of that sort of
thing. I bet you that every German paper is full of similar atrocities
committed by the English."
"Then you'll have to blush for the wickedness and the credulity of
the Germans too, darling!" remarked Dodo. "You will be red."
Edith laughed.
"Yes, I'm sorry I said that," she said. "But in any case what has
Brahms got to do with it? How can any sane person develop racial
hatred like that? Let's have a pogrom of Jews because of Judas
Iscariot. To go back. I'm not sent into the world to empty slops, but
to make symphonies. Very few people can make symphonies, and
I'm one of them. Huns or no Huns, what have artists to do with
war?"

"But, my dear, you can't help having to do with it," said Dodo. "You
might as well say, 'What have artists to do with earthquakes'?' But
an earthquake will shake down an artist's house just as merrily as a
commercial traveller's. You can't be English, and not have to do with
war."
Edith was silent a moment, and suddenly her face began to tie itself
into the most extraordinary knots.
"Give me some port or I shall cry," she said. "I won't cry; I never do
cry and I'm not going to begin now."
The prescription seemed to be efficacious.
"Then there's my boy," she said. "Berts has left Cambridge and I
suppose that before Christmas he'll be out in France. He's about as
much fitted to be a soldier as you are to be a housemaid. Of all the
instances of everybody wanting to do what they are totally incapable
of, the worst is the notion that we can make an army. You can't
make an army by giving boys bayonets. Germany is an army, for
forty years she has been an army. Why compete? Germany will wipe
up our army and the French army like a housemaid, which you want
to be, wiping up a slop. Have you seen what the German advance
has been doing this last week? Nothing in the world can save Paris,
nothing in the world can save France. Out of mere humanitarian
motives I want France to see that as quickly as possible. The war is
over."
Dodo rose.
"Don't talk such damned nonsense, Edith," she said. "That port has
gone to your head and given you vin triste. If anything was wanting
to make me quite certain that we are going to win it, it is the fact
that you say we are not. Do you remember when those beastly
Allensteins were staying with me, and how he knocked out
'Deutschland über alles,' on the table with his fat fingers? The effect
on you was that you played 'Rule Britannia' and 'God Save the King'
as loud as you could on the piano next door. It was extremely rude
of you, but it shewed a proper spirit. Why can't you do it now?"

"Because it's hopeless. Before Germany shewed her strength you
could do that just as you can tweak a lion's tail when he is lying
asleep behind bars at the Zoo. But now we're inside the cage. I don't
say we are not formidable, but we don't make ourselves more
formidable by sending all the best of our young men out to France to
be shot down like rabbits. We were not prepared, and Germany was.
Her war-machine has been running for years, smoothly and slowly,
at quarter-steam. We've got to make a machine, and then we've got
to learn how to run it. Then about the navy——"
Dodo assumed a puzzled expression.
"Somebody, I don't know who," she said, "told me that there was an
English navy. Probably it was all lies like the German atrocities."
Edith threw her hands wide.
"Do you think I like feeling as I do?" she asked. "Do you think I do it
for fun?"
"No, dear, for my amusement," said Dodo briskly. "But unfortunately
it only makes me sick. Hullo, here's David."
David entered making an awful noise on a drum.
"Shut up, David," said his mother, "and tell Edith what you are going
to do when you're eighteen."
"Kill the Huns," chanted David. "Mayn't I play my drum any more,
mummy?"
"Yes, go and play it all over the house. And sing Tipperary all the
time."
David made a shrill departure.
"Of course you can teach any child that!" said Edith.
"I know. That's so lovely. If I had fifty children I should teach it to
them all. I wish I had. I should love seeing them all go out to
France, and I should squirm as each of them went. I should like to
dig up the graves of Bach and Brahms and Beethoven and Wagner

and Goethe, and stamp on their remains. They have nothing to do
with it all but they're Huns. I don't care whether it is logical or
Christian or anything else, but that's the way to win the war. And
you're largely responsible for that; I never saw red before you talked
such nonsense about the war being over. If we haven't got an army
we're going to have one, and I shall learn to drive a motor. If I could
go to that window and be shot, provided one of those beastly Huns
was shot too, I should give you one kiss, darling, to shew I forgave
you, and go to the window dancing! I quite allow that if everybody
was like you we should lose, but thank God we're not."
Dodo's face was crimson with pure patriotism.
"I'm not angry with you," she said, "I'm only telling you what you
don't know, and what I do know, so don't resent it, because I
haven't the slightest intention of quarrelling with you, and it takes
two to make a quarrel. You know about trombones and C flat, and if
you told me about C flat——"
Edith suddenly burst into a howl of laughter.
"Or C sharp," said Dodo, "or a harpsichord. Oh, don't laugh. What
have I said?"
Edith recovered by degrees and wiped her eyes.
"In all my life I have never had so many offensive things said to
me," she remarked, "I can't think why I don't mind."
"Oh, because you know I love you," said Dodo with conviction.
"I suppose so. But there's Berts going out to that hell——"
"Oh, but you said the war was over already," said Dodo. "Besides
what would you think of him if he didn't go?"
"I should think it extremely sensible of him," began Edith in a great
hurry.
"And after you had thought that!" suggested Dodo.
Edith considered this.

"I don't know what I should think next," she said. "What I'm going
to do next is to get back to my scoring."
Edith's remarks about the absurdity of people attempting to do
things for which they had no aptitude made a distinct impression on
Dodo, and she totally abandoned the stocking of which she could
not turn the heel, and made no further dislocation of work by trying
to use a mop. But she found that if she really attended, she could
count blankets and bed-jackets, and weigh out stores and
superintend their distribution. Again, driving a motor was a thing
that seemed within the limits of her ability, and by the time that
Winston was in full running order as a hospital she was fairly
competent as a driver. Awful incidents had accompanied her
apprenticeship; she had twice stripped her gear, had run into a stone
wall, luckily in a poor state of repair, and had three times butted at a
gate-post. Her last accident, after a week really tedious from mere
uneventfulness, had been when she had gone all alone, as a
pleasant surprise, to the station to meet Jack, who was coming
home for two days' leave. She had been both driving and talking at
high speed, and so had not seen that she was close to a very sharp
corner on the marshy common just outside the gates, and preferring
the prudent course, as opposed to the sporting chance of getting
round the corner without capsizing, had gone straight ahead, leaving
the road altogether, until, remembering to apply her brakes, she
stuck fast and oozily in the marsh.
"There!" she said with some pride. "If I had been reckless and
imprudent I should have tried to get round that corner and had an
upset. Didn't I show presence of mind, Jack?"
"Marvellous. And what are we to do now?"
Dodo looked round.

"We had better shout," she said. "And then somebody will come with
a horse and pull us out backwards. It has happened before," she
added candidly.
"But if nobody comes?" asked he.
"Somebody is sure to. It's unthinkable that we should remain here
till we die of exposure and hunger, and the crows pick our whitening
bones. The only other thing to do is that you should jump out and
fetch somebody. I wouldn't advise you to, as you would sink up to
your knees in the mud. But it's a lovely afternoon; let's sit here and
talk till something happens. Haven't I learned to drive quickly?"
"Very quickly," said Jack. "We've covered the last three miles in four
minutes."
"I didn't mean that sort of quickly," said Dodo, "though daresay I
said it. Isn't it lucky it's fine, and that we've got plenty of time? I
wanted a talk with you and somebody would be sure to interrupt at
home. He would want sticking-plaster or chloroform or charades."
"Is all that your department?" asked Jack.
"Yes, they call me Harrods. You never thought I should become
Harrods. Oh, Jack, if you've got an ache in your mind, the cure is to
work your body till that aches too. Then two aches make an
affirmative."
"What?" said Jack.
"You see what I mean. And the odd thing is that though I'm entirely
taken up with the war, I try not to think about the war at all, at least
not in the way I used to before I became Harrods. One is too busy
with the thing itself to think about it. In fact, I haven't looked at the
papers for the last day or two. Has there been any news?"
"Not much. I've been busy too, and I really hardly know. But there's
been nothing of importance."
"Jack, what's going to happen?" she asked.

"Oh, we're going to win, of course. God knows when. Perhaps after
three years or so. But it's no good thinking about that."
Dodo gave a little groan.
"I know it isn't. If I realised that this was going on all that time, I
think I should just get drunk every day. Let's talk about something
else, and not realise it."
"When are you coming to see my camp?" asked he.
"I should think when the war is over and there isn't any camp. I
don't see how I can get away before. How long has it been going
now? Only three months, is it? And I can hardly remember what
things were like before. How did one get through the day? We got
up later, it is true, but then we went to bed later. Did we do nothing
except amuse ourselves? I couldn't amuse myself now. And what did
we talk about? I seem to remember sitting and talking for hours
together, and not finding it the least tedious."
"I shall insist on your having a holiday soon," said Jack.
"Oh no, darling, you won't. I've had fifty-five years' holiday in my life
and three months' work. That doesn't give much of a daily average,
if you work it out; somewhere about five minutes a day, isn't it? I
must have something better than that to shew before I have another
holiday.... Jack, did you say that we must look forward to three years
or more of this? Good Lord, how senseless it all is! What do you
prove by setting millions of jolly boys to kill each other? Oh, I
shouldn't have said that; I would have said, 'What do you prove by
having our jolly boys killed by those damned Huns?' Yes, darling, I
said damned, and I intended to. I told Edith that one day. The way
to win a war is to be convinced that your enemy are fiends. 'Also,' as
that fat Albert would say, 'we must therefore kill them.' But I wish I
really meant it. There must be a lot of nice fellows among the Huns.
They've had a bad education; that's what is the matter with them.
Also, they have no sense of humour. Fancy writing a Hymn of Hate,
and having it solemnly sung by every household! That odious Cousin

Willie has approved of it, and it is being printed by the million. No
sense of humour."
Dodo unconsciously hooted on her motor-horn, and looked wildly
round.
"I didn't mean to do that," she said, "because I don't want to be
rescued just yet. It's lovely sitting here and talking to you, Jack,
without fear of being asked to sign something. What was I saying?
Oh yes, humour! The Huns haven't got any humour, and the lack of
that and of mirth will be their undoing. How wise Queen Elizabeth
was when she said that God knew there was need for mirth in
England now, just at the time when England was in direst peril. That
is frightfully true to-day. We shall get through by taking it gaily. It's
much best not to let oneself see the stupendous tragedy of it all. If I
did that I would simply shrivel up or get drunk."
Dodo began a laugh that was near to a sob.
"I saw three boys this morning," she said, "all of whom had had a
leg amputated. There were three legs to the lot of them. So they put
their arms round each other's necks so as to form a solid body, and
marched down the long walk shouting 'left, right, left, right.' Then
they saw me, and disentangled their arms and grinned, and tried to
salute, and so they all fell down with roars of laughter. My dear, did
you ever hear of such darlings? That was the mirth that Queen
Elizabeth said was so necessary. I wanted to kiss them all, Jack."
"I want to kiss you," he said.
"Then you shall, you dear, if you think it won't shock the magneto. I
do miss you so horribly; you're the only real link between the days
before the war and the war. All other values are changed, except you
and David. What a nice talk we have had, at least I've had the talk,
so you must do your part and find it nice. Now let's hoot, until
several strong cart-horses come to help us."
Dodo performed an amazing fantasy on the horn, while the early
sunset of this November day began to flame in the west, which
reminded her that there were charades this evening. A chance

Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com

Applied Corpus Linguistics A Multidimensional Perspective Language And Computers 52 Language Computers Ulla Connor

More Related Content

Similar to Applied Corpus Linguistics A Multidimensional Perspective Language And Computers 52 Language Computers Ulla Connor (20)

Recently uploaded (20)

Applied Corpus Linguistics A Multidimensional Perspective Language And Computers 52 Language Computers Ulla Connor