Instant Download Introducing Pragmatics in Use 2nd Edition Anne O'Keeffe PDF All Chapter
Instant Download Introducing Pragmatics in Use 2nd Edition Anne O'Keeffe PDF All Chapter
Instant Download Introducing Pragmatics in Use 2nd Edition Anne O'Keeffe PDF All Chapter
com
https://ebookname.com/product/introducing-pragmatics-in-
use-2nd-edition-anne-okeeffe/
OR CLICK BUTTON
DOWLOAD NOW
https://ebookname.com/product/introducing-performative-
pragmatics-1st-edition-douglas-robinson/
https://ebookname.com/product/where-prosody-meets-pragmatics-
studies-in-pragmatics-8-2009th-edition-dagmar-barth-weingarten/
https://ebookname.com/product/introducing-neuropsychology-2nd-
edition-john-stirling/
https://ebookname.com/product/introducing-sociolinguistics-2nd-
edition-miriam-meyerhoff/
Introducing statistics 2nd ed Edition Upton
https://ebookname.com/product/introducing-statistics-2nd-ed-
edition-upton/
https://ebookname.com/product/advances-in-agronomy-82-1st-
edition-donald-l-sparks/
https://ebookname.com/product/ac-dc-first-edition-phil-sutcliffe/
https://ebookname.com/product/introducing-linguistic-
morphology-2nd-edition-laurie-bauer/
https://ebookname.com/product/introducing-linguistic-
morphology-2nd-edition-laurie-bauer-2/
Introducing Pragmatics in Use
Brian Clancy is Lecturer in Academic Writing and Research Methods, Academic Learning
Centre, and Lecturer in Applied Linguistics, Department of English Language and Literature,
Mary Immaculate College, University of Limerick, Ireland.
Svenja Adolphs is Professor of English Language and Linguistics and Head of School at
the School of English, University of Nottingham, UK.
Introducing
Pragmatics in Use
Second Edition
ANNE O’KEEFFE
BRIAN CLANCY
SVENJA ADOLPHS
Second edition published 2020
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2020 Anne O’Keeffe, Brian Clancy, Svenja Adolphs
The right of Anne O’Keeffe, Brian Clancy, Svenja Adolphs to be identified
as authors of this work has been asserted by them in accordance with
sections 77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced
or utilised in any form or by any electronic, mechanical, or other means,
now known or hereafter invented, including photocopying and recording,
or in any information storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
First edition published by Routledge 2011
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record has been requested for this book
ISBN: 978-1-138-48199-2 (hbk)
ISBN: 978-1-138-48200-5 (pbk)
ISBN: 978-0-429-34295-0 (ebk)
Typeset in Akzidenz Grotesk
by Apex CoVantage, LLC
Visit the eResources: www.routledge.com/9781138482005
Contents
Chapter 1 Introduction 1
1.1 What is pragmatics? 1
1.2 Ways of studying pragmatics 2
1.3 The empirical turn within pragmatics 3
1.4 The main functions of software tools used in corpus
pragmatics 5
1.5 The structure of this book 17
1.6 Further reading 19
Chapter 4 Reference 69
4.1 Introduction 69
4.2 Deixis 72
vi CONTENTS
Chapter 5 Politeness 98
5.1 Linguistic politeness 98
5.2 The face-saving approach to politeness 103
5.3 Impoliteness 114
5.4 Discursive politeness 119
5.5 Conclusion 122
5.6 Further reading 123
Appendix 229
References 233
Index 255
Figures
Throughout the book, we draw upon screenshots of the main corpus tools. We
acknowledge kind permission to reproduce these: AntConc; ELAN; #LancsBox; Sketch
Engine and WordSmith Tools (Lexical Analysis Software Ltd) and web-based search
interfaces for the Spoken BNC2014; the BNCWeb; Business Letters Corpus; Corpus of
Contemporary American English (COCA); Google Ngram Viewer; the Michigan Corpus of
Academic Spoken English (MICASE) and the Michigan Corpus of Upper Level Student
Papers (MUCUSP).
We acknowledge the research output drawn from the English Grammar Profile online
resource (Cambridge University Press), which includes examples from the Cambridge
Learner Corpus. We also reproduce, with kind permission, a screenshot sample from
resources that have been developed using the Trinity Lancaster Corpus (Centre for Corpus
Approaches to Social Science, Lancaster University/Trinity College London). Every effort
has been made to contact copyright holders. If any have been inadvertently overlooked the
publishers will be pleased to make the necessary arrangement at the first opportunity.
Although he did not live to see this second edition, we will always be grateful to the late
Ron Carter for his lasting inspiration. And to our partners Ger, Elaine and Nick, respectively,
we say thank you for your patience, support and love, as always.
CHAPTER 1
Introduction
and methods within pragmatics is not problematic. If anything, the vibrant scholarship from
both a micro- and a macro-perspective on the nature, conditions and variables of language
use adds to the breadth and depth of the field as a whole. Because this book involves look-
ing at naturally occurring language use, especially through corpora, there is often cross-
over between micropragmatic items such as deixis or speech acts and how they manifest
at a macro-level across social variables and conditions. While this book does not involve
introspective approaches to research, it benefits greatly from the scholarship in this area
in relation to core areas of pragmatics. Therefore, in defining pragmatics, we embrace the
richness across both component and perspective positions. For us, the best definition of
pragmatics remains a broad one which we cited in the first edition of this book, namely
that of Fasold (1990: 119), who says that it is ‘the study of the use of context to make
inferences about meaning’, where inferences refer to the deductions we make based on
available evidence. In the following section we explore further the notion of context and
how it can be studied.
Within a broad definition of pragmatics, we are looking at language in use and at meaning
in the making. As discussed above, core to this endeavour is trying to account for the vari-
ables of ‘context’ in understanding language in use. Rühlemann (2019) offers some useful
parameters for understanding the contextual variables of language. His non-exhaustive list
is summarised in Table 1.1.
Table 1.1 A range of contextual variables that aid understanding language in use (adapted from
Rühlemann, 2019: 6–7)
Variable Description
sequential The utterance(s) that precede and follow an utterance; the utterance(s) that
context can be expected to follow an utterance.
activity context The recognisable activity that the speaker and the hearer are engaged in at
the time of the utterance.
spatiotemporal The time and place when the utterance was made; the receiving, time and
context place of the utterance (for the listener or reader).
multimodal The speaker’s bodily conduct into which the utterance is integrated (e.g.
context posture, direction of gaze, how close they are to the speaker, whether they
move their head or hand when making an utterance, etc.).
intentional What the speaker intends to say in making an utterance, which often may not
context be apparent on the surface structure of the utterance.
emotive context The speaker’s emotional involvement with the entity the utterance is about.
epistemic context The (possibly infinite) range of the speaker’s and the hearer’s knowledge.
social context The power semantic or role relationship that holds between the speaker and
the hearer.
From Rühlemann’s (2019) list we can see that there is a broad matrix within which we
interpret and make meaning, both as interlocutors (speaker and hearer) and as observers
from the outside.
INTRODUCTION 3
1) Consider the variables in Table 1.1 above. What new variables might be added to
this list? Or, how might any of these be modified or subdivided?
2) In relation to two languages that you are familiar with, discuss how some of these
variables might differ across these languages and related cultures (e.g. in relation
to the activity context).
While it is noted that the empirical turn came late to pragmatics (Taavitsainen and Jucker,
2015), the field is not without a range of methodological models for gathering data (as we
will discuss in detail in Chapter 2). In order to examine language empirically, a valid and
reliable method of obtaining the data is needed. The main approaches in pragmatics are:
to elicit samples of the pragmatic phenomenon; to observe language and note how it is
used in a given context; to interview speakers about how they might use language or about
their opinions on language use, or to examine samples of recorded language that is stored
electronically in a corpus. Jucker et al. (2018) divide these across the following types of
empirical methodological approaches:
We will take a detailed practical look at these methodological approaches in Chapter 2. This
section serves as a general overview.
Over the years, with the growth of studies using empirical data in pragmatics, it cannot
but be noticed that a range of analytical frameworks can be used. We attempt to sum-
marise these in Table 1.2 across six approaches: experimental, ethnographic, ethnometh-
odology/conversation analysis, discourse analysis, critical discourse analysis and corpus
pragmatics.
In reality, approaches to pragmatics research are rarely siloed. By drawing upon syner-
gies in approaches, the researcher can find the optimum means of gathering and analysing
data. This can, for example, mean that an ethnographic approach will use corpus tools to
4
Table 1.2 A broad summary of the main approaches to gathering and analysing data in pragmatics research
Experimental Prompts are used to elicit a pragmatic The analytical focus is mostly on how speech acts manifest across variables such as speaker
approach phenomenon (e.g. speech act) using relationship, power semantic, L1 background which allow for pragmatic conclusions resulting
INTRODUCTION
a discourse completion task (DCT), from micro-analysis. Much of the interpretation of DCTs draws upon conversation analysis.
roleplay or interview. The large sample of responses elicited using the prompt task are analysed for patterns of
language use within or across turns, and this is often formulaic or routinised. Where relevant,
turn organisation can be examined and compared across contextual variables.
Ethnographic The researcher is immersed in a Observations, field notes and transcribed recordings are thematically analysed for an in-depth
approach community and makes observation- understanding of how language is used in a specific context. Language use in a community is
based field notes as well as video or analysed in an iterative way, where the researcher moves between hypothesising about how
sound recordings. interactions take place based on observations and close analysis of the actual interactions. This
leads to a ‘thick’ description of community activities (Marra and Lazzaro-Salazar, 2018: 359).
Ethnomethodology Very short recordings are made or Very detailed transcriptions of the turn-by-turn unfolding interactions are analysed so as to
and conversation identified for a very specific context establish turn preference, order and canonical sequencing within given situations. This approach
analysis (e.g. calls to a radio phone-in or involves looking in micro-detail at short stretches of interaction (e.g. a call opening) and, from
emergency helpline). this, generalisations can be made about role, power or context as they emerge through the turn-
taking order and sequence.
Discourse analysis Short recordings or texts from a Spoken or written texts are analysed for pragmatic features (e.g. pragmatic markers) or
specific context are gathered (e.g. a discourse features (e.g. text organisation).
classroom).
Critical discourse Individual recordings of (usually) public Usually, small samples of language are viewed from a critical perspective (e.g. the power
analysis speech events or texts are gathered (e.g. semantics of the pronouns used in a text; the use of modality and stance in political
a political speech, a newspaper article interviews; vagueness or deixis in news reports, etc.).
on Brexit, Tweets on a specific topic).
Corpus pragmatics Large corpora are accessed, usually Corpus tools are used to recall search items from the corpus. Pragmatic items may derive
online, or small corpora can be built from word or multi-word unit frequency lists or from keyword analysis. Alternatively, pragmatic
through text curation or through functions may be recalled if the data has been pragmatically annotated or if existing knowledge
recordings. These are transcribed of speech act manifestations, such as illocutionary force identifying devices (IFIDs), are used
and sometimes annotated (e.g. as the basis for searches. Concordancing is used to look at these search items in a more
for all instances of a pragmatic contextualised way. There is often a need for close analysis of large amounts of concordance
phenomenon). lines to manually categorise the function of a given form.
INTRODUCTION 5
The frequency of a word or a phrase (multi-word unit) tells us about its profile and use. If a
word or phrase recurs in a given context, it is usually indicative of a salient feature. For this
reason, frequency lists are seen as a good starting point for the analysis of a corpus. As
a first step into corpus data, a researcher will look at the word list results and this is often
done by comparing it with another word list from a different corpus. When looking at a word
list comparatively, the researcher is normally more concerned with the distribution at the top
of the rank order list (i.e. the top most frequent words or phrases). The British National Cor-
pus (BNC) (see Appendix) offers a benchmark for typical rank order, for example, the, of, an,
to and a are the first five most frequent items (in that order). In order to discover instances
of the pragmatically specialised use of language, we might take this set as our baseline and
follow up on any differences that might occur, as we illustrate in Task 1.2.
Consider the word lists in Table 1.3. It shows actual (‘raw’) results for the top 20 most fre-
quent words in three different corpora:1 the EnTenTen15 (internet texts); The BNC1994
(spoken and written) and the TED_en corpus (TED talks). Notice how the rank order of
the first five items is very similar across the three corpora.
6 INTRODUCTION
Table 1.3 Top 20 most frequent words in the EnTenTen15 corpus, The British National Corpus
1994 and the TED_en corpus
Look at the order of the words from the sixth to the twentieth most frequent (these are
shaded in Table 1.3).
1) Circle the words that stand out as being different in their rank order of frequency.
2) Taking the different types of corpora into consideration, speculate as to why these
words you have circled might be more or less frequent than in the other lists.
3) If you have access to these corpora, examine the words you have identified by
looking at them in concordance lines. Check if your intuition is correct (see also the
section on concordance lines and task 1.4 where we follow up on this).
Corpus software can also count phrases (combinations of words) that frequently recur;
for example, Table 1.4 shows the top ten most frequent two-, three- and four-word units in
the Limerick Corpus of Irish English (LCIE) (see Appendix).
Table 1.4 Ten most frequent two-word, three-word and four-word units in LCIE
1 you know 4406 I don’t know 1212 you know what I 230
2 in the 3435 do you know 769 know what I mean 215
INTRODUCTION 7
This corpus software function has added much to our understanding of multi-word
units (MWUs). The many terms that have evolved under the umbrella term multi-word units
(MWUs) reflects the ongoing attempt to find the best methodology for both counting and
accounting for the fact that some words seem to occur in units with other words. These
terms often come with slightly differing definitions and include, inter alia: formulas, formu-
laic units, formulaic sequences, routines, fixed expressions, prefabricated patterns (prefabs),
clusters, chunks, concgrams, strings, n-grams and lexical bundles/lexical phrases (see
Greaves and Warren (2010) and Gray and Biber (2015) for coverage of differing terminol-
ogy, methodologies and research findings).
For the purposes of this book, we opt to use the umbrella term multi-word unit (MWU)
unless we are discussing a particular study or relevant finding. Some key issues and find-
ings from research to date include the following:
• MWUs are very common and are defined differently across a number of studies. Termi-
nology and definitions are tied up with important variations in retrieval methods (What
constitutes a unit?; What is the frequency cut-off for inclusion?, etc.);
• MWUs can be examined as continuous (e.g. you know what I mean) or discontinuous
(e.g. the * of * ) sequences;
• We can talk about the length of an MWU in terms of whether it is across two-, three-,
for-, five- or six-word slots and, within these units, we can examine the fixedness or
variability of the constituent components;
• MWUs vary in frequency across speech and writing. For example, Biber et al. (2004)
found lexical bundles to be more frequent in speech than in writing. Other studies
found that continuous units were more frequent in speech, whereas writing (espe-
cially academic registers) relies heavily on discontinuous units which act as frames
(Biber, 2009);
• MWUs vary across registers and many units have become pragmatically specialised in
terms of their discourse function (see Chapter 8).
The study of MWUs is important for pragmatic research because they are often asso-
ciated with functions such as stance marking, focus, text organisation and referential mean-
ing. O’Keeffe et al. (2007), for example, examined MWUs in the Cambridge and Nottingham
Corpus of Discourse in English (CANCODE) (see Appendix), and found that many items
had pragmatically specialised functions in spoken interactions, summarised in Table 1.5.
8 INTRODUCTION
Table 1.5 Pragmatically specialised functions of MWUs, based on O’Keeffe et al. (2007)
Function Example
Discourse marking you know; I mean; and then; but I mean; you know what I mean; at the end of the day
Face and politeness do you think; do you want (me) (to); I don’t know if; what do you think; I was
(mitigation) going to say
Vagueness and a couple of; and things like that; or something like that; (and) that sort of thing;
approximation (and) all this/that sort of thing
Other items can be appear to be functionally less specialised but often these items
relate to the world of the speaker or writer. That is, they are frames for referential meaning –
to refer to the shared world of the interlocutors (see Biber et al., 2004; O’Keeffe et al.,
2007; Biber, 2009; Greaves and Warren, 2010). We will explore referential frames in Task
1.3 (based on some selected results from O’Keeffe et al., 2007; Biber, 2009; Gray and
Biber, 2015).
Examine the following high-frequency multi-word units (MWUs) based on existing research
which operate as frames (* marks a slot that can be filled by various possible words):
1) Think of words that might go into the empty slot marked by * and then sort the
phrases that you have generated into the following common referential categories.
Try to come up with three examples for each category.
2) Which other category might you add to finish categorising the phrases you have
generated?
3) Can you think of longer phrases that these items might frame across these three
functions (e.g. It was getting late; In the beginning of the century, etc.)?
We will be exploring MWUs in terms of their pragmatic meaning and discourse function
in a number of chapters in this book. Within a corpus pragmatic approach, what interests us
most is how these units function pragmatically across different contexts of language use.
Concordance lines
As discussed above, word frequency lists, when examined comparatively, can point to the
possibility of a word having some specialised meaning or use in a particular context. This
INTRODUCTION 9
can sometimes lead to insights about pragmatically specialised uses when a form is fur-
ther investigated. Therefore, it is essential that frequency analysis is complemented by a
detailed consideration of the environment of a word through the use of concordance tools.
A concordance, as defined by Sinclair (2003: 173), ‘is an index to the places in a text
where particular words and phrases occur’. In a concordance, the search word you enter
will appear in the middle of the search screen (see Figure 1.1). This word is referred to as
the node (in Figure 1.1, have is the node). There are, however, some caveats concerning
concordance lines. The first is that although they provide information on a node (the search
word), they do not interpret it. It is the responsibility of the researcher to use the software
to determine the patterns that are salient and to construct hypotheses as to why these
patterns occur. Therefore, as Baker (2006: 89) states, ‘a concordance analysis is … only
as good as its analyst’. We will now exemplify some of the typical phases that a researcher
might undertake in the process of hypothesis formation, moving iteratively between word
lists, concordances and patterns.
In Task 1.2, when looking at words that seem to be at a higher or lower rank order
across the three lists in the task, you may have noticed the word have. It is in the top
20 most frequent words from the TED_en corpus but it does not appear in the top 20
most frequent words in the other two corpora, EnTenTen15 and the BNC. This points to
the possibility of being used in some specialised way in the TED talks data. When we
create a concordance of have in the TED_en corpus and we sort it to the right of the
node (i.e. organise the words immediately to the right of have in alphabetical order), we
can check for any interesting patterns by scrolling down through the screens of results
(see Figure 1.1).
In scrolling through the screens we are trying to identify patterns, and the norm is to
look for patterns both to the right and left of the node word through sorted searches. To
further examine the patterns of have, we can use the ‘collocates’ function, which is normally
available as part of a corpus tool (see more on collocates below). As Figure 1.2 illustrates,
the software will instantly calculate the most statistically salient words that co-occur to the
left and right of the node word have.
Based on the results in Figure 1.2, any of the top collocates listed merit further inves-
tigation based on their statistical salience.2
By way of illustration, we will examine the first item we by looking at the patterns of
we + have in a concordance (as illustrated in Figure 1.3).
Next, we use the software to calculate the most frequent words immediately to the
right of we + have (often referred to as 1R). The top ten results are shown in Figure 1.4.
The results of this search in Figure 1.4 give us more possible routes of investigation.
Let us follow the first line by examining to. The concordance lines of we + have + to will
show us more patterns. In Task 1.4, we will explore the collocates of we + have + to. For
a researcher, this process would be undertaken by close analysis of concordance lines. In
Table 1.6 we have put together some concordance examples by way of illustration.
Reflecting on the patterns in Table 1.6, we can hypothesise that we have to + [verb],
used in TED talks, is gaining a pragmatically specialised use in a requestive context to make
a call for action or a strong appeal. We note that speakers use we rather than you when
addressing their audience so as to make the request inclusive and global. We may also
speculate that this use is particularly associated with contexts of public discourse on the
issue addressing global climate change.
Figure 1.1 A sample of concordance for have in the TED_en corpus sorted 1R (using Sketch Engine)
INTRODUCTION 11
Figure 1.2 Top ten collocates of have in the TED_en corpus (using Sketch Engine), examining all
word candidates three to the left and three to the right of we
There are many ways of examining and interpreting concordance line data and it is a
process that is central to corpus pragmatics. Here we consider it in greater detail based on
the seminal work of Sinclair (1996). Concordance output facilitates an inductive approach
by helping the user notice patterns relating to how a lexical item or MWU is used in context.
In order to describe the nature of individual units of meaning, Sinclair (1996) suggests four
parameters that are important to the process of interpreting a concordance: collocation,
colligation, semantic preference and semantic prosody:
1) Collocation: refers to lexical patterning and the probability of two words co-occurring
frequently next to or near each other: blonde hair, make an effort, do one’s duty, tor-
rential rain, strictly forbidden, a major incident. Some collocate relationships are strong
because the possibilities of other combinations are few. For example, make/express/
fulfil + wish are strong collocates because wish does not collocate with a wide range
of verbs, whereas the adjective big + car, town, house are weak collocations because
big can collocate with many words (Carter et al., 2011). Corpus software will help you
calculate the strength of a relationship between two words (see Figures 1.2 and 1.4).
Figure 1.3 Concordance sample of we + have as the node (sorted 1R)
INTRODUCTION 13
Figure 1.4 Top ten most frequent words to the right of we + have in the TED_en corpus (using
Sketch Engine)
2) Colligation: refers to the grammatical patterning of words and the likelihood of the
co-occurrence of grammatical choices. By using a certain verb, for example, this may
co-select a particular syntax. For example, we say I was discharged from the hospital
rather than I was discharged out of the hospital. Even though from and out of both
imply exiting from the building, only from colligates with discharge. When we refer to
leaving a hotel, we use a different verb entirely and we use active voice and out of, as
in I checked out of the hotel. The strength of a colligational pattern is also included by
software in its collocates function (see again Figures 1.2 and 1.4 where a number of
grammatical items are included).
3) Semantic preference: refers to how collocates can, through usage, appear to have
a preference for a particular semantic domain. For example, in his discussion of the
expression ‘the naked eye’, Sinclair (1996) finds that most of the verbs and adjectives
that collocate with this expression are related to the concept of ‘vision’. A search
for the collocates of the naked eye using COCA shows visible, invisible, seen, see,
appears, looks all within the top ten most frequent collocates to the left of the search
phrase.
4) Semantic prosodies: associations that arise from the collocates of a lexical item are
not easily detected using introspection (Sinclair, 1987; Louw, 1993). Semantic proso-
dies have mainly been described in terms of their positive or negative polarity (Sinclair,
1991; Stubbs, 1995). For example, naked eye is often found in relation to objects that
cannot be seen with the naked eye. Carter and McCarthy (1999) illustrate the negative
prosody associated with the get passive in the corpus data they examined (e.g. get
arrested, get sued, get nicked). Rühlemann (2010) notes the negative prosody of set in
(e.g. boredom can easily set in).
14 INTRODUCTION
Table 1.6 shows us the most frequent verbs that follow we + have + to in the TED_en
corpus. Examine these verbs and their examples.
1) What hypotheses can you form about possible pragmatically specialised uses of
we + have + to + [verb] in TED talks, based on Table 1.6?
2) How might you follow up on these hypotheses using the TED_en corpus or another
corpus?
Table 1.6 A list of the most frequent verbs that follow we + have + to in the TED_en corpus,
with examples
At this point, let us distinguish between collocation, colligation and multi-word units
(MWUs) (detailed above); we note that collocation and colligation are concerned with the
co-occurrence relationship of one word with another word rather than as a unit. Crucially,
the researcher sets out to examine the collocational or colligational relationship of a given
word in a more top-down manner. That is, they choose to examine it as we did here with the
word have and its patterns. A researcher looking at MWUs takes a more open, bottom-up
approach through corpus software searches for n-gram units within the parameters they
set (e.g. a researcher might opt to find two-, three-, four-, five- or six-word units, with a min-
imum frequency cut-off of 20 per million, etc.) (see Greaves and Warren (2010) and Gray
and Biber (2015) for useful background on this topic). Both top-down collocational and
colligational analyses and bottom-up MWU approaches are important to corpus pragmatics
INTRODUCTION 15
and, in both cases, it is through concordance line analysis that we can analyse meaning,
discourse function and ultimately pragmatic specialisation.
Keyword analysis
Keywords can be described as words (or MWUs) which occur with unusual frequency in a
text or a set of texts in a corpus when compared to another corpus. The corpus used for
comparison is called the reference corpus. Keywords are identified on the basis of statis-
tical comparisons of word frequency lists from the reference corpus and the corpus under
investigation (referred to as the target or study corpus). The frequency of each item in the
target corpus is compared with its equivalent in the reference corpus and the statistical
significance or difference is calculated using chi-square or log-likelihood statistics (see
Dunning, 1993). The choice of the reference corpus used as the basis for comparison in the
calculation of keywords is important because it will affect the output of keywords (Gabriela-
tos (2018) offers detailed coverage of this). When generating keyword lists, it is best to try
more than one reference corpus and to consider the differences in the results. In general
terms, the closer the reference corpus is in terms of genre, the fewer keywords will result
because fewer items will be unusually frequent. Conversely, the more distant a reference
corpus is in terms of genre from the target corpus, the more words will have comparatively
more unusual frequencies, and so more keywords will normally result from the comparison.
These differences (as a result of using different reference data) are in themselves telling,
as Task 1.5 illustrates.
Table 1.7 and 1.8 show two sets of keyword results. Both lists are generated from the
same text but use different reference corpora. The target text was the well-known
1995 BBC 1 Panorama television interview by Martin Bashir with Diana, Princess of
Wales.3
Review and compare the list and consider these questions:
Table 1.7 KEYWORD LIST 1: All of the keywords based on comparison of the Bashir–Diana
Panorama interview with Spoken Media Corpus (arranged vertically in the grids in order of ‘keyness’)
Table 1.8 KEYWORD LIST 2: A sample of the 92 keywords based on comparison of the
Bashir–Diana Panorama interview with Spoken Academic Corpus (arranged vertically in the
grids in order of ‘keyness’)
Some commentary on the keyword lists in Task 1.5 is based on O’Keeffe (2006, 2012) and
Vaughan and O’Keeffe (2015).
On one hand, the results in Table 1.8, based on the more ‘distant’ reference in terms
of its genre, appear more wide-ranging (in all there are 92 keywords from this calculation)
and capture more of the ‘aboutness’ of the target text (Phillips, 1989). We find common
first- and second-person pronouns I, I’m, my, myself, yourself and me arising as keywords
because they are not high frequency in academic lectures and thus arise as ‘unusually fre-
quent’. We also see keyword results (Table 1.8) that reference the more private sphere of the
‘I–you’ domain, including husband–wife relationships, love, bulimia, marital breakdown, etc.,
all of which would not normally be talked about in the more referential world of academia.
Meanwhile, in Table 1.7, we see that by using a reference corpus that was close in genre
to the target text there were far fewer items. In other words, we can say that the results in
Table 1.7 possibly represent more salient keywords because, despite the reference corpus
being close to the target, these words are still used with unusual frequency by comparison.
Therefore, working through concordances of this candidate list might be more productive
(and more doable in scale). Gabrielatos (2018) notes that the size of the reference corpus
is not as important as the representativeness of each corpus, and the principled selection
of corpora to be compared.
Finally, a comment on the challenges of spoken corpora. In Tables 1.7 and 1.8, we
notice that the vocalisation uh appears as a keyword in both lists. This is most likely a
function of the variation in the transcription of vocalisations in the reference corpus which
comprises many media transcripts. Some transcribe the same or similar vocalisation as
INTRODUCTION 17
uhm, erm or ah, among others. This is an important point regarding the analysis of spoken
corpora: if similar words or vocalisations are transcribed differently, this will have a bearing
on keyword calculations (for more on transcription, see Chapter 2).
This book is structured around nine chapters. These move from introductory matters in
this chapter, including the history, origin and emergence of the field and an overview of the
application of different analytical frameworks in empirical research. In this chapter we also
cover the basic functions of corpus tools because it sets the scene for much of what we
talk about in other chapters.
Chapter 2 looks in detail at how pragmatics can be researched. This chapter show-
cases quite a large toolkit available to a researcher who is interested in gathering data
for pragmatics research. It distinguishes between elicited, observed and recorded data,
and guides the reader through the different instruments that can be used under these
three broad headings. All methods have pluses and minuses and we try to present a
balanced view. Although we clearly have a preference for using corpus data, we are keen
to stress that other methods have a lot to offer in themselves and if used in conjunction
with corpus tools (as we explore in detail in Chapter 3, see below). The main point to
take away from Chapter 2 is that there is a need for methodological awareness. Some
methods can be highly controlled so that the researcher can be very precise in the lan-
guage that they elicit but this is at the expense of the degree to which the researcher
compromises on the naturalness of the data. On the other hand, naturally occurring data
can be elicited and recorded through note-taking or digital recording but the researcher
has little control over the data that results and this can pose challenges for pragmatics
research. What is interesting in Chapter 2 is the range of approaches that have emerged
for gathering research data within the empirical turn in pragmatics, as we have already
alluded to in this chapter.
Chapter 3 focuses on corpus pragmatics, a recent coinage for the coming together of
corpus linguistics and pragmatics. The chapter addresses the processes of doing corpus
pragmatics research in a way that accommodates different approaches. Corpus pragmat-
ics usually works from frequencies of forms to their pragmatic function in what is termed
a form-to-function approach (this is exemplifed in this chapter through our analysis in the
section on concordance lines, for instance). In this approach, the frequency lists and con-
cordance lines ultimately lead us to a conclusion about the use of a form (e.g. we have to in
TED talks; see above). The opposite approach is to begin with the function (e.g. a speech
act), and to try to narrow down the range of possible forms used to perform this and to use
these forms to find language instances in a corpus. For example, the words and phrases
typically associated with a speech act (as a result of experimental research, such as illocu-
tionary force identifying devices (IFIDs)) can be used to search a large corpus to retrieve
examples. This approach is referred to as a function-to-form approach. In an ideal world,
function-to-form approaches are facilitated by pragmatically annotated corpus data so that
all instances of a given speech act or pragmatic phenomenon (such as a pragmatic marker)
can simply be recalled. At the time of writing, pragmatic annotation is fast developing (see
Weisser, 2015; Archer and Culpeper, 2018).
18 INTRODUCTION
Chapter 4 furthers our exploration of key concepts within the study of pragmatics
through its focus on reference. The chapter covers its general definition as well as going
into analytical depth in terms of how deictic reference can be examined using corpus prag-
matics. Deixis represents the intersection of grammar and pragmatics, and the chapter
explores many of these grammatical items such as the personal pronouns you and I and the
demonstratives this and that. This chapter showcases the potential of corpus pragmatics
for examining the relationship between the context of the utterance and the referential
practices therein. This relationship is shown to characterise the very nature of our pragmatic
systems.
In Chapter 5, we explore politeness theory through the lens of a number of different
models. Within these paradigms, we examine some key features using corpus pragmatic
techniques. For example, we explore Brown and Levinson’s (1987) concepts of positive
politeness through a case study of vocatives, and negative politeness through a micro-study
of hedging across different contexts of use. In this chapter, we also look at the concept of
impoliteness in the context of naturally occurring data. Finally, we examine discursive polite-
ness where we look beyond linguistic structures to include the individual’s interpretation of
these structures as (im)polite in instances of ongoing verbal interaction. This brings to the
fore the dynamic notion of relational work.
Speech acts are the focus of Chapter 6, which examines the link between linguistic
forms in the shape of speech acts and their function in context. We provide an overview of
Speech Act Theory and discuss the main arguments and underlying assumptions on which
this theory is based. This includes a discussion of direct and indirect speech acts, performa-
tives and constatives, and the broad taxonomy of different speech act categories such as
directives or commissives. The chapter also looks at the way in which context and co-text
impact upon the analysis of speech acts in a discourse framework. Throughout this chapter,
we explore ways of using corpus pragmatics in the form-to-function analysis of speech
acts. This adds further context to issues discussed in both Chapters 2 and 3.
Drawing upon a range of different corpora, Chapter 7 examines pragmatic variation
within a language. As we note, the study of language variation has traditionally focused on
phonological, lexical and syntactical levels, particularly taking an historic view. The system-
atic study of variation at a pragmatic level is a relatively recent development by comparison.
This chapter also highlights the broadening of the variational focus from phonology, lexis
and syntax to variation in social space. This is achieved through explorations of variation
from a macro-social perspective (e.g. factors such as age, gender, ethnicity, social class,
etc.), and from a micro-social perspective (e.g. more ‘local’ factors such as the degree of
social distance between participants (strangers, friends, family) or power (an employee talk-
ing to her or his boss)).
To contrast with the focus of Chapter 7 on variation within a language, we go a
level deeper in Chapter 8 to examine variation in terms of register. In this chapter, we
will explore the notion that specific registers involve the pragmatically specialised use
of language. In doing so, we will draw upon naturally occurring language from a range
of contexts, including casual conversation, healthcare communication, crime fiction, ser-
vice encounters and Shakespearean drama. The chapter again employs a corpus prag-
matic approach to the examination of features characteristic of these specific situations.
This chapter also builds on Chapter 2 in relation to the synergies between conversation
INTRODUCTION 19
analysis (CA) and corpus linguistics, offering some useful examples of how CA can aid in
the analysis of corpus data.
The final chapter in our book, Chapter 9, looks at pragmatics and language teaching,
and considers the degree to which it is teachable and learnable in the context of the ongo-
ing debate in Second Language Acquisition (SLA) studies in relation to implicit and explicit
learning. The chapter explores areas of pragmatics that can be addressed in the classroom,
through both hands-on corpus tasks as well as through the curation of teacher-led activi-
ties and materials. The chapter includes a wider range of samples of classroom materials
based around the teaching of politeness and spoken grammar, including pragmatic mark-
ers (discourse markers, response tokens, etc.), as well as vagueness and stance markers.
Developments in learner corpus research in terms of how it can inform language teaching
are showcased through learner corpus-based resources that are differentiated by level of
proficiency. Throughout the chapter, there is an emphasis on modelling corpus tasks based
on existing research findings, as it is argued that this offers a means for bringing focus to
pragmatic competence within curriculum, syllabus and materials design.
Each chapter in this book contains an annotated further reading section intended to
guide the reader to texts that expand upon key topics discussed in each chapter. The book
also includes, as this chapter has demonstrated, tasks which are embedded within key
topics. Some of these tasks involve reader interaction with specific corpus interfaces and
specially designed corpus software. Our goal is to show the reader how the interfaces and
software can be used in corpus pragmatic research – should more information about the
specifics of using these tools be required, there are some very helpful textbooks available
(see e.g. O’Keeffe and McCarthy, 2010; Weisser, 2016a; Anderson and Corbett, 2017). We
do not assume that a reader will undertake all of these activities, but we hope that they offer
an instructive application of the core concepts that we are discussing.
Clancy, B. and A. O’Keeffe, 2015. Pragmatics. In D. Biber and R. Reppen (eds), The
Cambridge Handbook of English Corpus Linguistics. Cambridge: Cambridge University
Press, pp. 235–251.
This chapter showcases the potential of corpus pragmatics to bring insights through
research into forms, their patterns and pragmatic functions in large corpora. It cov-
ers aspects of deixis, pragmatic markers, language and power; discourse organisation,
and provides a case study on the use of a corpus to explore vocative forms and their
functions.
20 INTRODUCTION
Jucker, A. H., 2012. Pragmatics in the history of linguistic thought. In K. Allan and K.M.
Jaszczolt (eds), The Cambridge Handbook of Pragmatics. Cambridge: Cambridge Uni-
versity Press, pp. 495–512.
This chapter offers a wide-ranging overview of the emergence of pragmatic thought and
how it developed across two schools, the Anglo-American and Continental European tra-
ditions. By reading this chapter, a student of pragmatics will gain greater insight into why
there are very different approaches within the field, ranging from introspective to empirical
in terms of research method.
Schneider, K.P. and A. Barron (eds), 2014. Pragmatics of Discourse. Berlin: Mouton de
Gruyter.
This edited volume brings together 21 chapters addressing approaches to the analysis
of discourse pragmatics, including discourse markers, stance, speech act sequences as
well as an overview of work across different contexts, including legal, medical, media and
classroom discourse domains. In terms of approaches to discourse analysis within which
pragmatics can be viewed, it includes work on conversation analysis, systemic-functional
linguistics, genre analysis, critical discourse analysis, corpus linguistics and multimodal
pragmatics.
NOTES
Researching pragmatics
While this chapter focuses on ways of collecting and analysing data in pragmatics research,
we acknowledge that it does not represent the totality of research in the area of prag-
matics as a whole. As discussed in Chapter 1, pragmatics has its origin in the philoso-
phy of language and there remains an important ongoing debate on the core concepts
of pragmatics; this also helps to drive and shape the frameworks that we apply within the
empirical side of pragmatics research. Such intellectual endeavour is as important to prag-
matics (and linguistics in general) as it is to philosophy. The non-empirical side of prag-
matics research involves discussion, critique and philosophical debate, but it also draws
upon intuitively derived instances of language use. Jucker (2018) reminds us that intuition
was an important source of data for early philosophers of language and pragmaticists. For
example, Austin and Searle relied on their intuition in their seminal work on speech acts:
‘[t]heir data consisted of their own intuition about the use of language’ (Jucker, 2018: 5).
We note that the term intuitive knowledge better represents what philosophers of language
bring to their theorising process. The term introspection is also sometimes used, though this
is prone to confusion as it has different coinage in other fields (see Clark, 2018; Schneider,
2018). For instance, in experimental cognitive research, introspection can refer to think-
aloud protocols, verbal reports and other means of elicitation (Clark, 2018). Distinguish-
ing between intuitive knowledge and introspection is useful in pragmatics, Jucker (2018)
notes, because when we talk about theorising about language use we draw upon intuitive
knowledge, whereas within elicitation practices (e.g. discourse completion tasks, roleplays),
participants engage in introspection to come up with the language use they consider to
be typical for a given situation. On the other hand, intuitive knowledge, within pragmatics
research, refers to the knowledge a researcher brings to the task of investigating language.
As we illustrated in Chapter 1, the iterative process of analysing language using corpus
data constantly moves between results and intuition as the researcher follows up on hypoth-
eses which frequency results trigger.
Pragmatics offers a framework for understanding language. There are many means
of arriving at a pragmatic understanding of language, ranging from the analysis of texts
to philosophical debate, as discussed. Within the empirical side of pragmatics research,
there are a number of methodologies in use to gather language data. In this chapter we will
22 RESEARCHING PRAGMATICS
look at some of the main approaches. First of all, let us begin with a conclusion: there is no
one right way to gather data in pragmatics research. Pragmatics, as a framework for the
study of intended meaning in social context, has proved to be methodologically inclusive.
When we consider language within pragmatics, we can do so through reflection or intuitive
knowledge or we can examine language from outside by collecting and analysing it (within
a range of data-collection instruments). This chapter focuses on the main options available
to a researcher when gathering language data ‘from outside’ (see Figure 2.1).
LANGUAGE INSIDE
Language as intuive knowledge
Figure 2.1 Language inside as intuitive knowledge and language outside as empirical data
Collecting language can take many forms. We can set up controlled tasks so as to
expedite the gathering of data that is most relevant to our research question. For example,
if we want to examine what speakers typically say in response to a compliment, we can
design a discourse completion task or a roleplay which focuses solely on this situation.
This task can ask participants to introspect on what they would typically say next (after the
prompt provided). This type of task is high in terms of the control and precision which the
researcher has over the data being collected but it is also high in terms of the degree of
interference on the part of the researcher (Jucker, 2018). Alternatively, we could use data
from a large general corpus of spoken language and find many instances where speakers
give compliments, and, from these, instances of interlocutor responses could be recalled
by manually sifting through examples. This type of approach gives the researcher far less
control over the data elicitation process and is more challenging and more time-consuming
(see Chapter 3), but this is offset by a low level of interference because the researcher did
not mediate the interaction by means of a controlled task to elicit a compliment. This distinc-
tion between degree of control and degree of interference is very useful when it is plotted
across different types of data. Let us consider this in terms of approaches to gathering
spoken data based on a scale adapted from Jucker (2018: 23) in Figure 2.2.
As we look in greater detail at some of the main methods of eliciting data in pragmatics
research in this chapter, we constantly keep these dimensions of control and interference in mind.
The most basic way of gathering language data is by noting it down once you have heard
it. This is referred to as attested data and it is not uncommon to see such examples in
RESEARCHING PRAGMATICS 23
Figure 2.2 Researcher interference versus research control (adapted from Jucker, 2018: 23)
pragmatics. Attested data is language which the researcher has said, heard, written or read
incidentally and then recorded (usually by noting it down). Such instances are similar to field
notes that anyone interested in language may collect on a daily basis, as they observe uses
of language around them. The method is used more systematically within ethnographic
approaches to data gathering (see below). The word attest means to show something to
be true, and when a researcher uses attested data, the reader trusts its veracity as part of
our common code of research integrity. Attested data is usually used to illustrate and add
to discussion on pragmatic phenomenon, especially speech acts. Rühlemann (2019: 33)
offers this instance of attested data as a precise example in his discussion of politeness
and indirect speech acts (see Chapter 6):
(2.1)
Normally, attested data is a short snippet (as in extract 2.1) and this aligns with a
researcher’s ability to recall, with accuracy, only a short instance of language. Ad hoc attes-
tation of language used in both broadcast and social media is much easier now because
of audio- and text-capturing facilities, and therefore it is possible to retrieve recordings and
texts of longer stretches of attested media discourse (O’Keeffe, 2012).
In order to gather multiple instances of language data in a more systematic way, there
are two main options beyond attestation. Each option has advantages and disadvantages in
terms of degree of research control and interference, as discussed above (see Figure 2.2):
1) Elicitation of language
Tasks that elicit or ‘draw out’ language are designed to focus precisely on the language
the researcher wishes to examine from a predetermined sample group of participants.
Typical tasks include discourse completion tasks, roleplays and interviews, all of which
are pre-designed with structured prompts (with varying degrees of control).
24 RESEARCHING PRAGMATICS
2) Recording of language
Ideally, a researcher would like to use language that occurs naturally in a given
context and to record it as it is happening in real time. Ethnographic methods
allow for this through situational recording, noting of attested language and field
notes made by the researcher who is embedded within a given context for an
extended period. Language corpora also give the researcher access to large sam-
ples of spoken or written language which the researcher can search for a particu-
lar language item; researchers can also build their own corpora in a more tailored
way which may also involve the researcher being present when the recordings
are made.
1) Give some specific details on an elicitation method (i.e. what type of task would you
use?) and note your thoughts on the main advantages and disadvantages with this
approach.
2) Give specific details on a corpus approach to finding examples of responses to
apologies (e.g. Which corpus would you use? Would you build one?).
Method
Advantage
Disadvantage
In this section we have seen that there are a number of ways of eliciting data and
recording language, as Figure 2.2 illustrates. Each has strengths and weaknesses which we
will consider in detail in Section 2.2, as we describe the main methodological tools. Let us
first address issues of ethics, integrity and documentation management.
Before looking at methods, there are some important issues to consider relating to research
ethics, integrity and data management. As is the norm, higher education institutions will
have rigorous protocols in place to ensure ethical approval, research integrity, and data
storage and usage. In the first instance, always seek advice from your institution on the
procedures that are in place. You will normally need, at a minimum, to prepare the following
types of documents for your study.
RESEARCHING PRAGMATICS 25
This is a short document, written in plain English, explaining what your research is about. It
will include information on how the data will be recorded, transcribed, stored and used. It is
essential to explain to the participant how their data will be used and who will have access
to it. The participant also needs to be assured that they have the right to withdraw from
the research at any point. In addition, a guarantee of anonymity for participants must be
provided by assuring them that their real names or references to any identifying information
(e.g. name of a shop, town, school or business) will be changed or redacted during transcrip-
tion. The briefing document needs to provide the name of the institution and contact details
of the principal investigator or research supervisor whom the participant may contact should
they wish to have further information in relation to the research.
Consent form
A separate form is needed for consent from each participant. It must include the following:
This form will gather the demographic details about each participant. If a detail is not rel-
evant in your study, do not ask for it. Typically, certain information is asked for: age range;
gender; first and other language information; nationality and geographical origin. In cor-
pus pragmatic or ethnographic studies especially, speaker roles and relationships will be
required.
26 RESEARCHING PRAGMATICS
By gathering consent forms and speaker information (metadata) and eliciting language
data from participants in a research study, a researcher is taking on the responsibility of hon-
ouring a commitment to protecting (1) these data through secure storage, and (2) the identity
of the participants. This needs to be a key consideration when physically storing these data in
compliance with data protection laws and in line with the ethical procedures of your institution.
DCT 1
You are a university student. Your end-of-term assignment is one day late so far. You
knock on your tutor’s door.
You are applying for a position with a multinational company. The interview committee
has requested that you have your professors send letters of recommendation directly
to the company. When you call the interview committee to check the status of your
application, you are told that one of the recommendation letters has not arrived. You
are concerned because you asked your professor for the letter over a month ago.
You stop by your professor’s office to find out what has happened.
RESEARCHING PRAGMATICS 27
DCTs have been widely and successfully used in the study of speech acts (see Chap-
ter 6) and speech events (e.g. asking the time). They are particularly favoured as a meth-
odology in the study of second-language pragmatic competence (Blum-Kulka et al., 1989;
Sasaki, 1998; Billmyer and Varghese, 2000; Ogiermann, 2018) (see Chapter 9). DCTs can
be used in dialect studies where they focus on very specific structures (see Barron, 2005;
Schneider and Barron, 2008a).
DCTs can be written tasks (as in Task 2.2 above) or they can be presented orally (and
recorded). Schneider (2018) notes that written DCTs have the advantage of allowing for
the collection of samples of language from a large number of informants in a short time;
for example, by administering a written DCT to a large lecture hall, or via email, social media
or crowdsourcing platforms. The other advantage is that large-scale gathering of DCT data
potentially offers the researcher rich metadata for comparative work (e.g. across gender,
nationality, first language, age, etc.).
The degree to which DCTs are controlled can vary. Consider the scenario in DCT 1 in
Task 2.2. The informant has a choice, albeit rather limited, as to the speech act which they opt
for. DCTs can be even more controlled when they are presented in a turn-based format. The
example in DCT 2 in Task 2.2 illustrates how a scenario is set up and then a turn is provided.
This is followed by ‘write-on lines’ (a blank turn) which the informant must complete, and so on.
A further example is taken from Beebe and Zhang Waring (2004). Here the authors
set out to investigate the pragmatic tone (Figure 2.3).
You go to a tourist bookstore where the books are kept behind a counter. You ask
to see a book on display. The lady behind the counter says, ‘If you want to browse,
go to a library.’
Figure 2.3 An example of a freer DCT (adapted from Beebe and Zhang Waring, 2004: 245)
An additional attraction of DCTs is their ‘discreteness’. The researcher has a lot of con-
trol over the language they want to elicit. The focus can be limited to a very specific context
of use, as the examples of the above tasks illustrate. Boxer and Cohen (2004) note that
DCTs are used particularly when:
28 RESEARCHING PRAGMATICS
As Schauer and Adolphs (2006: 120) put it, ‘the aim of discourse completion task research
is to investigate a linguistic act within highly predefined parameters.’
The level of control in DCTs has come in for criticism over the years (see Beebe and
Cummings, 1996; Schauer and Adolphs, 2006; Schneider, 2018). It is argued by some that
DCTs cannot be used to appraise pragmatic competence in the study of foreign-language
users because apart from providing too little context they cannot constrain discourse
options without contaminating the response, for example, where the language of the task
leads the informants to produce certain language (Yoon and Kellogg, 2002). In an effort
to redress this, Yoon and Kellogg (2002) used a cartoon DCT so as to provide a pictorial
context to constrain the response while allowing freedom to elaborate language. Not sur-
prisingly, the criticism that DCT data lacks interactional and prosodic features (in the case
of written DCTs) is frequently reported (see Ogiermann, 2018). Cultural considerations are
also at play. Rose (1994) conducted a study of requests in Japanese and American English
using both DCTs and questionnaires (see more on questionnaires below) and concluded
that DCTs may not be culturally appropriate for Japan.
Another fundamental criticism levelled at DCTs is that the language they elicit is
‘unnatural’ when compared to naturally occurring data. Quite a number of interesting
studies have resulted from the quest to test whether this is the case (e.g. Hartford and
Bardovi-Harlig, 1992; Beebe and Cummings, 1996; Bou Franch and Lorenzo-Dus, 2008;
Maíz-Arévalo, 2015). Ogiermann (2018) offers a succinct overview of these studies and
notes that, overall, they confirm that DCTs and naturally occurring data contain ‘similar
semantic formulae’ (Ogiermann, 2018: 243). As we discuss in greater detail in Chapter
3, Schauer and Adolphs (2006), in a study of expressions of gratitude in DCT data com-
pared with corpus data, illustrate that because DCT data is normally based around single
utterances, the overall reality of a speech act is distorted because the typical extended turn
negotiation and development is lacking. Some studies found the use of predictable for-
mulaic language (Maíz-Arévalo, 2015; Ogiermann, 2018) while others reported that DCTs
produce more direct (less polite) instances. However, Economidou-Kogetsidis’ (2013)
study, which used real requests for information made to an airline reservation centre as
the basis for a DCT scenario which were then administered to 86 people, concluded that
DCT and naturally occurring data were similar in terms of both degree of directness and
lexical modification across turns (see also Beebe and Cummings, 1996; Golato, 2003;
Ogiermann, 2018).
Speech acts related to conflict and disagreement are particularly elusive (essen-
tially because informants are reluctant to be recorded in such situations) and DCTs have
been used as a means of gathering these and many contrastive speech act studies have
resulted. For instance, Liang and Han (2005) looked at disagreement strategies between
American English and Mandarin Chinese. They based their study on five scenarios for
RESEARCHING PRAGMATICS 29
disagreement within a college context. The scenarios vary in the power relationships, rang-
ing from higher to lower status, including peer–peer interactions. Three of the scenarios
are shown in Figure 2.4.
Thank you very much for your time and help. Five scenarios are described below in
which you are expected to disagree with the speaker on different occasions. How
would you respond? Please write out what you are to SAY in real-life scenarios.
1. Your supervisor questions the originality of the term paper you submit. S/he
says to you, ‘I’m sorry, but I don’t think these ideas are yours.’ However, they are
yours. In response, you will say:
‘ ................’
2. Your friend makes the following comment on your thesis: ‘I think you should
supply more data to support your arguments. You know, your conclusion is a
little bit weak.’ However, you think that there has been enough evidence and the
problem is how to give a better explanation of the data. In response, you will say:
‘ ................’
Figure 2.4 Scenarios from DCT presented to American and Mandarin students (in Liang and
Han, 2005)
Among other findings, Liang and Han (2005) tell us that Chinese students employ more
politeness strategies when disagreeing with higher status interactants. Both the Chinese
and the American students showed fewer politeness strategies when disagreeing with
peers. As Ogiermann (2018: 247) concludes, DCTs have their value and their limitations,
and as long as we are aware of what they can and cannot provide and of what other alter-
native methods we can use in their place, then ‘the DCT has its place in pragmatic research’.
DCTs can be designed as multiple-choice task (MCT) questionnaires that include scaled
response tasks where participants assess situational contexts and speech acts according
to certain variables. Respondents are not required to produce language. They are not put in
30 RESEARCHING PRAGMATICS
the position of having to comprehend utterances in real time. However, Kasper and Roever
(2005) point out that they do impose a certain cognitive load upon respondents. Essentially,
they demand recognition while scaled response rating means respondents have to perform
judgement tasks. Using MCT questionnaires has the advantage of being quick to administer
and analyse compared to other elicitation methods. However, they have to be carefully designed
to suit the individual research context. The design of multiple-choice questionnaires is particu-
larly demanding (see Kasper and Roever, 2005). All of the response options must be plausible
(unlike multiple-choice options in an assessment context where only one response is plausible
and all others are implausible distractors). For advice on response design, see Schneider (2018).
Schneider (2018) notes that from his direct experience, Asian students find DCTs very
challenging because they struggle to come up with what to write and they find MCTs much
more accessible as a task. By way of caveat, Kasper (2000) and Kasper and Roever (2005)
note that multiple-choice questionnaires in pragmatics research have varying degrees of
success depending on their purpose. Multiple-choice studies which probe situational rou-
tines and implicatures were found to have a satisfactory degree of consistency (see Kasper,
2008), while those studies which look at speech act realisation strategies ‘tend to achieve
notoriously poor reliability scores’ (Kasper and Roever, 2005: 328).
Roleplays
Roleplays are particularly common as a means of eliciting oral data in the study of cross-cul-
tural and inter-language pragmatics, especially when comparing native (NS) and non-native
speaker (NNS) language use in relation to the same task (NS – NS versus NNS – NNS) or
RESEARCHING PRAGMATICS 31
within the same task (NSs – NNSs) (see Félix-Brasdefer (2018) and Schneider (2018) for
detailed coverage). A roleplay usually involves a prompt which is given to speakers. These
vary in terms of control. The participants can also be given contextual information, including
a description of the situation and details about the status of each role (e.g. you are the
employee/employer; student/teacher; friend/friend, etc.). Figure 2.5 is a typical roleplay
prompt, based on Félix-Brasdefer (2018: 310). In this case, it is being used to elicit exam-
ples of refusals to invitations but does not provide much contextual or situational detail.
There are many variations and variables in terms of enhancing the prompt and controlling the
task (see Félix-Brasdefer, 2018). This will be discussed further below (see Task 2.4). The
language generated in the roleplay can then be recorded and analysed.
A friend of yours is leaving to return to China after an academic year abroad. He has
invited you to his farewell party. Unfortunately, you can’t make it.
The term roleplay can sometimes be used as a superordinate for a number of variations
(see Table 2.1). Roleplay and role enactment involve taking on social roles (varying in how
related or familiar they are to the participants), while simulated tasks assign discourse roles
to participants who respond as themselves in these discourse roles.
Table 2.1 Terminological differences relating to roleplays (based on McDonough, 1986; Félix-Brasdefer,
2018)
Roleplay Taking on a role that is not yourself and responding to the situation that is
prompted in that role (rather than as yourself). For example, you are a doctor;
you are a professor; you are a train driver.
Role enactment Taking on a role designed to fit with your known experience and responding
to a familiar situation within that role. For example, [someone who is a
postgraduate student]: you are a postgraduate student and your assignment is
late; [someone who is a waiter]: you are a waiter and you are late for work.
Simulation Taking on a discourse role and responding as yourself to a situation. For
example, (role: direction giver) you are stopped by someone looking for
directions on the university campus. What would you say?
Boxer and Cohen (2004: 17) point out that in certain contexts roleplay data are similar
to spontaneous spoken data, ‘with the caveat that the researcher is able to set up a con-
text for studying speaking’. Demeter (2007) sees roleplay as a method which brings the
researcher closest to authentic data in the study of the production of speech acts. This is
supported by Rosendale (1989), who used roleplays to elicit data for a study on the speech
act of invitation. Based at a Romanian university, Demeter (2007) used roleplay as a means
of data gathering for a study of apologies. One of the explicit aims of his study was to
demonstrate that the use of roleplays is a valid and effective method of collecting data for
the analysis of apologies. To support this, he compared data collected through roleplays with
32 RESEARCHING PRAGMATICS
data collected via a DCT. His study involved 19 university students majoring in English. They
were asked to roleplay an apology in six situations adapted from the US television show
Friends which ran from 1994 to 2004. Some of these situations included the following:
• You did not have time to change before going to the wedding of your best friend, and
therefore you are wearing sports clothes;
• You had promised your spouse that you would stop smoking. However, you started
again, and your wife can tell that you were smoking again;
• You arrive late to your friend’s birthday dinner;
• You took your friend’s blue jeans without telling him or her. Now your friend has found
out and you admit to taking them.
Roleplays have also come in for criticism for being artificial, but Félix-Brasdefer (2018)
notes that they are a powerful tool for the investigation of interactional aspects of communi-
cation (such as turn organisation, turn taking and overlap, etc.), as well as prosodic features
and cues (such as intonations, tone and stress). In addition, they can allow for the explora-
tion of macro-social dimensions (such as gender, age, ethnicity and socioeconomic status),
within and across languages (see Márquez Reiter, 2000; Félix-Brasdefer, 2009; Cohen,
2012). Félix-Brasdefer (2018) offers extensive detail on roleplays as a methodological tool
in pragmatics. Based on Félix-Brasdefer (2018), we summarise some the main broad types
of roleplays:
Closed roleplays: These are essentially oral discourse completion tasks and seek to elicit
one-turn responses. The oral response to the prompt is recorded. Com-
pared to written DCTs, these are argued to produce longer responses,
as well as features, verbal and non-verbal, associated with spoken dis-
course, such as hesitation, backchannels, gestures, etc. that a written
DCT will not capture (see Félix-Brasdefer (2018: 308) for examples).
Open roleplays: An open roleplay is typified by being dyadic. Participants normally read
the situational prompt and are then asked to roleplay how they would
respond. The interaction is recorded for later analysis.
The degree of detail provided in the prompt (degree of formality of setting, relationship and
status of interlocutors) can have an impact on the quality and length of the data it produces
(see Billmyer and Varghese, 2000; Félix-Brasdefer, 2010). A good approach is to pilot your
prompt and then enhance it based on the pilot. Task 2.4 explores the piloting process based
on Figure 2.5.
A friend of yours is leaving to return to China after an academic year abroad. He has
invited you to his farewell party. Unfortunately, you can’t make it. What do you say?
Félix-Brasdefer (2018: 325ff.) offers extensive methodological and ethical advice which is
essential reading if you would like to use this instrument.
Interviews
Interviews have proved useful in pragmatics research. These interviews are sometimes
sourced from other fields, such as sociolinguistics, and repurposed for pragmatics
research goals (see Kasper, 2008; Schneider, 2007, 2018). For example, Schneider
(2007) compared responses to thanks in closing sequences of interviews. The interviews
were originally conducted in relation to attitudes to regional dialects in England. Others
have used interviews to elicit first-order conceptualisations of politeness, rudeness or
speech acts. For example, interviewees could be asked directly for their definitions of
small talk, banter and gossip, or to share their understanding of insults, threats, and so on
(see Schneider, 2018).
Interviews have also been used in pragmatics research to elicit interviewee examples
of particular speech acts. For example, an interviewee could be asked to detail the last
time they received a compliment or an insult. This information can be used as the basis for
designing DCTs, MCTs or roleplays, as we explore in Task 2.5.
Ask someone about the last time they experienced one of the following and use the
details as the basis for a DCT or roleplay prompt:
• received a compliment
• made an apology
• refused an invitation
• congratulated someone
• made a complaint
• insulted someone
• apologised for a mistake they had made at work
Barron (2003) and Woodfield (2012) offer interesting examples of how inter-
views can be used in conjunction with a roleplay or DCT task to gain more follow-up
insight into why informants said something within a task. For example, informants may
be asked, ‘What was going through your mind when you said this?; What made you
Another random document with
no related content on Scribd:
(This concludes the first lesson, during which the words and
significations of Variability, Variate, Array, and Median will have been
learnt.)
The second lesson is intended to give more precision to the idea
of an Array. The variates in any one of these strung loosely on a
cord, should be disposed at equal distances apart in front of an
equal number of compartments, like horses in the front of a row of
stalls (Fig. 4), and their tops joined. There will be one more side to
the row of stalls than there are horses, otherwise a side of one of
the extreme stalls would be wanting. Thus there are two ways of
indicating the position of a particular variate, either by its serial
number as ‘first,’ ‘second,’ ‘third,’ or so on, or by degrees like those
of a thermometer. In the latter case the sides of the stalls serve as
degrees, counting the first of them as 0°, making one more
graduation than the number of objects, as it should be. The
difference between these two methods has to be made clear, and
that while the serial position of the Median object is always the same
in any two Arrays whatever be the number of variates, the serial
position of their subdivisions cannot be the same, the ignored half
interval at either end varying in width according to the number of
variates, and becoming considerable when that number is small.
Lines of proportionate length will then be drawn on a blackboard,
and the limits of the Array will be also drawn, at a half interval from
either of its ends. The base is then to be divided centesimally.
Next join the tops of the lines with a smooth curve, and wipe out
everything except the curve, the Limit at either side, and the
Centesimally divided Base (Fig. 5). This figure forms a Scheme of
Distribution of Variates. Explain clearly that its shape is independent
of the number of Variates, so long as they are sufficiently numerous
to secure statistical constancy.
Show numerous schemes of variates of different kinds, and
remark on the prevalent family likeness between the bounding
curves. (Words and meanings learnt—Schemes of Distribution,
Centesimal graduation of base.)
The third lesson passes from Variates, measured upwards from
the base, to Deviates measured upwards or downwards from the
Median, and treated as positive or negative values accordingly (Fig.
6).
Draw a Scheme of Variates on the blackboard, and show that it
consists of two parts; the median which represents a constant, and
the curve which represents the variations from it. Draw a horizontal
line from limit to limit, through the top of the Median to serve as
Axis to the Curve. Divide the Axis centesimally, and wipe out
everything except Curve, Axis, and Limits. This forms a Scheme of
Distribution of Deviates. Draw ordinates from the axis to the curve at
the 25th and 75th divisions. These are the ‘Quartile’ deviates.
At this stage the Genesis of the theoretical Normal curve might be
briefly explained and the generality of its application; also some of
its beautiful properties of reproduction. Many of the diagrams
already shown would be again employed to show the prevalence of
approximately normal distributions. Exceptions of strongly marked
Skew curves would be exhibited and their genesis briefly described.
It will then be explained that while the ordinate at any specified
centesimal division in two normal curves of deviation measures their
relative variability, the Quartile is commonly employed as the unit of
variability under the almost grotesque name of ‘Probable Error,’
which is intended to signify that the length of any Deviate in the
system is as likely as not to exceed or to fall short of it. This, by
construction, is the case of either Quartile.
(New words and meanings—Scheme of Distribution of Deviates,
Axis, Normal, Skew, Quartile, and Probable Error.)
In the fourth lesson it has to be explained that the Curve of
Normal Distribution is not a direct result of calculation, neither does
the formula that expresses it lend itself so freely to further
calculation, as the curve of Frequency. Their shapes differ; the first
is an Ogive, the second (Fig. 7) is Bell-shaped. In the curve of
Frequency the Deviations are reckoned from the Mean of all the
Variates, and not from the Median. Mean and Median are the same
in Normal Curves, but may differ much in others. Either of these
normal curves can be transformed into the other, as is best
exemplified by using a Polygon (Fig. 8) instead of the Curve,
consisting of a series of rectangles differing in height by the same
amounts, but having widths respectively representative of the
frequencies of 1, 3, 3, 1. (This is one of those known as a Binomial
series, whose genesis might be briefly explained.) If these rectangles
are arrayed in order of their widths, side by side, they become the
equivalents of the ogival curve of Distribution. Now if each of these
latter rectangles be slid parallel to itself up to either limit, their bases
will overlap and they become equivalent to the bell-shaped curve of
Frequency with its base vertical.
The curve of Frequency contains no easily perceived unit of
variability like the Quartile of the Curve of Distribution. It is therefore
not suited for and was not used as a first illustration, but the formula
that expresses it is by far the more suitable of the two for
calculation. Its unit of variability is what is called the ‘Standard
Deviation,’ whose genesis will admit of illustration. How the
calculations are made for finding its value is far beyond the reach of
the present lessons. The calculated ordinates of the normal curve
must be accepted by the learner much as the time of day by his
watch, though he be ignorant of the principles of its construction.
Much further beyond his reach are the formulae used to express
quasi-normal and skew curves. They require a previous knowledge
of rather advanced mathematics.
(New words and ideas—Curve of Frequency, Standard Deviation,
Mean, Binomial Series).
The fifth and last lesson deals with the measurement of
Correlation, that is, with the closeness of the relation between any
two systems whose variations are due partly to causes common to
both, and partly to causes special to each. It applies to nearly every
social relation, as to environment and health, social position and
fertility, the kinship of parent to child, of uncle to nephew, &c. It may
be mechanically illustrated by the movements of two pulleys with
weights attached, suspended from a cord held by one of the hands
of three different persons, 1, 2, and 3. No. 2 holds the middle of the
cord, one half of which then passes round one of the pulleys up to
the hand of No. 1; the other half similarly round the other pulley up
to the hand of No. 3. The hands of Nos. 1, 2, and 3 move up and
down quite independently, but as the movements of both weights
are simultaneously controlled in part by No. 2, they become
‘correlated.’
The formation of a table of correlations on paper ruled in squares,
is easily explained on the blackboard (Fig. 9). The pairs of correlated
values A and B have to be expressed in units of their respective
variabilities. They are then sorted into the squares of the paper,—
vertically according to the magnitudes of A, horizontally according to
those of B—, and the Mean of each partial array of B values,
corresponding to each grade of A, has to be determined. It is found
theoretically that where variability is normal, the Means of B lie
practically in a straight line on the face of the Table, and observation
shows they do so in most other cases. It follows that the average
deviation of a B value bears a constant ratio to the deviation of the
corresponding A value. This ratio is called the ‘Index of Correlation,’
and is expressed by a single figure. For example: if the thigh-bone of
many persons deviate ‘very much’ from the usual length of the
thigh-bones of their race, the average of the lengths of the
corresponding arm-bones will differ ‘much,’ but not ‘very much,’ from
the usual length of arm-bones, and the ratio between this ‘very
much’ and ‘much’ is constant and in the same direction, whatever be
the numerical value attached to the word ‘very much.’ Lastly, the
trustworthiness of the Index of Correlation, when applied to
individual cases, is readily calculable. When the closeness of
correlation is absolute, it is expressed by the number 1·0; and by
0·0, when the correlation is nil.
(New words and ideas—Correlation and Index of Correlation.)
This concludes what I have to say on these suggested Object
lessons. It will have been tedious to follow in its necessarily much
compressed form,—but will serve, I trust, to convey its main purpose
of showing that a very brief course of lessons, copiously illustrated
by diagrams and objects to handle, would give an acceptable
introduction to the newer methods employed in Biometry and in
Eugenics. Further, that when read leisurely by experts in its printed
form, it would give them sufficient guidance for elaborating details.
6. The Herbert Spencer Lecture delivered before the University at Oxford, June
5th, 1907.
7. Dent’s “Everyman’s Library,” price One Shilling.
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside
the United States, check the laws of your country in addition to
the terms of this agreement before downloading, copying,
displaying, performing, distributing or creating derivative works
based on this work or any other Project Gutenberg™ work. The
Foundation makes no representations concerning the copyright
status of any work in any country other than the United States.
1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if
you provide access to or distribute copies of a Project
Gutenberg™ work in a format other than “Plain Vanilla ASCII” or
other format used in the official version posted on the official
Project Gutenberg™ website (www.gutenberg.org), you must,
at no additional cost, fee or expense to the user, provide a copy,
a means of exporting a copy, or a means of obtaining a copy
upon request, of the work in its original “Plain Vanilla ASCII” or
other form. Any alternate format must include the full Project
Gutenberg™ License as specified in paragraph 1.E.1.
• You pay a royalty fee of 20% of the gross profits you derive
from the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.F.