Text Messaging and Literacy – The

Evidence

As children are given mobile phones at increasingly younger ages, there is consider-
able media coverage of claims that mobile phones, and text messaging in particular,
are responsible for declining levels of literacy in children and young people. Such
claims are often adopted wholesale by teachers and parents, despite the fact that
there is an empirical literature which has failed to find a basis to these claims, and
to the contrary has found that text messaging is supporting children’s literacy skills.
Written by leading international researchers, Text Messaging and Literacy –
The Evidence presents an overview and discussion of the academic evidence for
and against the use of text messaging and mobile phones in supporting literate
activity, and discusses what conclusions we can and should draw about the impact
of mobile phones and their potential role in education. Areas covered include:

x the rise of texting and media reactions;

x children’s reading, spelling and texting;
x using mobile phones for literacy development;
x texting and literacy skills in adolescents and adults;
x spelling and grammar in texting and beyond; and
x the future of texting.

In challenging existing assumptions, the authors present the cutting edge of inter-
national research, highlighting their own studies involving children of all ages,
adolescents and adults. This ground-breaking book is essential reading for both
researchers and students in education, educational psychology, literacy and new
media and its impact on learning.

Clare Wood is Professor of Psychology in Education at Coventry University, UK.

She is Editor-in-Chief of the Journal of Research in Reading and is the Director
of the Centre for Applied Research in Psychology.

Nenagh Kemp is Senior Lecturer in the School of Psychology, University of Tasmania

(UTAS), Australia, and is also Associate Editor for the Journal of Research in Reading.

Beverly Plester is an Honorary Research Fellow in the Department of Psychology

and Behavioural Sciences, Coventry University, UK.
Routledge Psychology in Education
Edited by Karen Littleton

The new Routledge Psychology in Education series is interdisciplinary in nature,

publishing cutting-edge research in educational psychology and education-based
research from related areas, including cognition, neuropsychology and social
psychology. Titles will take a broad and innovative approach to topical areas of
research, and will address the needs of both researchers and advanced students
(Masters and Ph.D.) within both psychology and education programmes and
related areas. Titles in the series will:

x review the field to provide an interesting and critical introduction to the

student;
x explore contemporary research perspectives, issues and challenges; and
x signpost future directions and trends.

Other titles in the series include:

Theories of Learning for the Workplace

Building blocks for training and professional development programs
Filip Dochy, David Gijbels, Mien Segers and Piet Van den Bossche

Interdisciplinary Perspectives on Learning to Read

Culture, cognition and pedagogy
Kathy Hall, Usha Goswami, Colin Harrison, Sue Ellis and Janet Soler
Text Messaging and
Literacy – The Evidence

Clare Wood, Nenagh Kemp

and Beverly Plester
First published 2014
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Simultaneously published in the USA and Canada
by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2014 Clare Wood, Nenagh Kemp and Beverly Plester
The right of Clare Wood, Nenagh Kemp and Beverly Plester 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. The purchase of this copyright material confers the
right on the purchasing institution to photocopy pages which bear the
photocopy icon and copyright line at the bottom of the page. No other
parts 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.
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 for this book has been requested

ISBN: 978-0-415-68715-7 (hbk)

ISBN: 978-0-415-68716-4 (pbk)
ISBN: 978-0-203-69336-0 (ebk)

Typeset in Galliard
by Book Now Ltd, London
Contents

List of illustrations ix
About the authors xi
Preface xiii
Acknowledgements xv

1 Mobile phone use and the rise of texting 1

The rise of the telephone 1
Mobile telephones arrive 3
Text messages arrive 4
‘Perpetual contact’, ‘always on’ 5
Texting versus talking 7

2 The media furore 12

Text language and the media 12
Text language and the texters 18
Frequencies of textisms of various types from the elicited and
spontaneous text corpus 19
Code switching and heteroglossia 21

3 The links between children’s spelling, reading and

texting 23
Does exposure to misspellings in general harm
children’s literacy? 23
Texting and academic achievement 26
The impact of input method: how does predictive text
use contribute to literacy? 29
Textism use and reading difficulties 30
The story so far 32
vi Contents

4 Does mobile phone use facilitate literacy development? 34

A longitudinal analysis of textism use 34
Using mobile phones as a technological intervention 36
The million dollar question: should we buy our children
mobile phones? 39
Persistent negative perceptions 39

5 Texting and literacy skills in adolescents and young

adults 41
Adolescents’ and young adults’ use of texting and textisms 41
Textisms in languages other than English 43
Relationships between textism use and literacy skill:
self-report 44
Relationships between textism use and literacy skill:
experiments 45
Relationships between textism use and literacy skill:
naturalistic studies 46
Limitations 48
Alternative explanations 49

6 Understanding children’s mobile phone behaviours

in relation to written language abilities 53
The research participants 53
The assessment of written language skills and mobile
phone behaviours 54
Typical patterns of texting and text exposure via phones 55
Types of technology and predictive text 56
Levels of ‘addiction’ 57
Enjoyment and motivation 60
How do primary and secondary school children’s mobile
phone behaviours compare? 61
The relationship between text messaging behaviour and
literacy outcomes 62
So what have we learned? 65

7 Texting and grammar 68

Punctuation 68
Capitalisation 72
Omission of words 73
Ungrammatical word forms 74
Do errors mean ignorance? 74
Conclusions 77
 Contents vii

8 Methodology matters: Issues in the collection

and coding of textisms 79
Self-report of mobile phone-related behaviours 79
Self report: number of messages and textism use 80
Message translation 81
Message elicitation 82
Message production 82
Message collection 83
Naturalistic messages 84
Comparison across methods 85
Counting and categorising textisms 88
Sex differences 89
Comprehension of textisms 90

9 Lessons learned and the future of texting 92

What can we say? 92
Outstanding questions 93
Methodological points for attention 96
Texting as an educational tool? 98
Future gazing and concluding comments 99

Appendix A: Mobile phone use questionnaire 103

Appendix B: Coding textisms 107
Appendix C: The coding of grammatical errors 109
References 112
Index 122
This page intentionally left bank
Illustrations

Tables

3.1 Summary of the various forms of textism used, based

on typology from Thurlow and Brown 28
6.1 Children’s mean scores on written language tasks by
predictive text group 57
6.2 Literacy scores of children who reported enjoying
versus not enjoying using text abbreviations when
sending text messages 64
6.3 Literacy scores of children who reported enjoying
versus not enjoying creating text abbreviations 65

Figures

2.1 Textisms used in elicited texts by type of textism 20

2.2 Textisms used in spontaneous texts by type of textism 20
6.1 Responses to ‘Do you always understand the textspeak
that other people use?’ (Question 33) 56
6.2 Responses to ‘How often do you carry your phone
with you?’ (Question 7) 58
6.3 Responses to ‘How important is it to keep your
phone charged?’ (Question 8) 58
6.4 Responses to ‘Do you spend more time using your
phone than…?’ (Questions 13–17) 59
6.5 Responses to ‘How do you feel when you leave your
phone at home?’ (Question 32) 60
6.6 Responses to ‘Do you enjoy…?’ (Questions 28–30) 61
6.7 Analysis of things children reported enjoying about
their phone (Question 33) 62
This page intentionally left bank
About the authors

Clare Wood is Professor of Psychology in Education at Coventry University,

and directs the Centre for Applied Research in Psychology there. She has a
long-standing interest in children’s reading development as well as interests
in children’s use of new technology and the ways in which it can support
learning and impact on attainment. She has a slight phobia of telephones,
however, and only recently acquired her first smart phone, which she mainly
uses for emails and text messaging. And using the internet to buy books. A
lot of books. The most useful application of a smart phone to child develop-
ment that she has found to date is that it makes an excellent teether. Her
most annoying habit is that she keeps her mobile phone switched off until
she wants to send a text, and frequently forgets to keep her phone charged.
(It’s not deliberate, honest. Well, most of the time it’s not deliberate…)
Nenagh Kemp is Senior Lecturer in Psychology in the School of Psychology at
the University of Tasmania. Another reading researcher, she is particularly
interested in adults’ and children’s spelling and grammatical understanding,
and embarrasses her friends by taking photographs of signs with bad gram-
mar and incorrect punctuation when she is out and about. It is a miracle she
has not yet been arrested, but at least she resists the temptation to correct the
errors. She has learned more new words than she ever expected while con-
ducting research into the way Australians abbreviate words in speech (e.g.
coming for a barbie this arvo?). Clare has confirmed that Nenagh is a very,
very good house guest when she visits the United Kingdom, and this is the
main reason why she wanted to write a book with her. Nenagh’s other skills
are that she can fit all her essential worldly possessions into a backpack and
salsa dance. Although not necessarily at the same time.
Beverly Plester is a retired Honorary Research Fellow in the Psychology
Department at Coventry University, which is where she met Clare, and even-
tually Nenagh. A developmental psychologist who hails originally from the
US, Bev has an impressive enthusiasm for all things texting, and it was Bev
who supervised the original research work by Victoria Bell which looked at
texting and academic ability. She is known for supporting and inspiring the
xii About the authors

next generation of researchers interested in this area, and is a probably best

described as a knowledge junkie/eternal student, especially when it comes to
language. Her grandchildren are currently teaching her how to use her smart
phone, and of the three authors she writes the most textism-heavy text mes-
sages. She constantly impresses Clare and Nenagh with her knowledge of and
ability to pronounce Finnish words. One day she will teach Clare how to roll
her rs. She also makes excellent bread.
Preface

This book is concerned with the impact that the technological phenomenon of
text messaging has had on the literacy skills of children and adults, and how the
empirical evidence on this relates to the popular perceptions and portrayal of
this matter in newspapers, internet discussion forums and blogs. The summary
of this book is that public understanding of how texting relates to literacy skills
does not necessarily reflect the reality of the research evidence on this topic. This
is understandable in the face of speculative media coverage which has promoted
discussion of how declining literacy standards amongst younger people must be
linked to their increased use of, and addiction to, new technologies and techno-
logical practices which make some others feel disenfranchised. As academics, we
consumed these newspaper narratives and were struck by how strongly asserted
they were in the absence of systematic empirical work in the area. Our work and
that of others has now explored the actual evidence for these popular accounts
and this book is our attempt to organise and summarise the resulting evidence on
the topic in a way that makes sense of the data, and aims to clarify what can and
cannot be said about the positive and negative contributions of texting to rising
or falling standards in literacy.
As you may be able to tell, this is a book born partly out of frustration with
the persistent, stereotypical views of technologically-literate children and young
people, but our approach to the research (both the conduct of our own, and the
review of others’ work) was open-minded, and continues to be so. Some areas pre-
sent the reader with consistency and coherence, whereas others are more equivocal
and nuanced in their messages. Crucially, we also include a critical evaluation of
the methods that have been used to examine the impact of text messaging on lit-
eracy skills. We felt that this was an essential component of any discussion of what
can and cannot be said about texting and educational outcomes.
In Chapter 6 we also present new data which examines the nature of chil-
dren’s use of their mobile phones and relates this to their performance on
measures of written language processing. This study was conducted in an
attempt to understand whether some of the popular characterisations of chil-
dren and their mobile phone use were valid, and if they were, to examine the
extent to which they were linked to academic performance. These new data,
xiv Preface

like so much of the work in this book, challenge the popular portrayal of
children as technologically dependent.
We hope that this book is useful in drawing together a range of work in this
area, such that the reader can draw his or her own conclusions about the perils
and pitfalls of texting in the twenty-first century. As is so often the case, more
work needs to be conducted in this area, but there is an evidence base out there
which parents, journalists and educators can draw on when debating the impact
of children’s technological participation in mobile phone culture.
Acknowledgements

The authors would like to thank all the colleagues who have supported both their
research work in this area and their work on this book. In particular, we would
like to recognise the contribution and assistance of the following colleagues, stu-
dents and friends: Victoria Bell (who fired our interest in all of this in the first
place), Samantha Lowry (née Bowyer), Puja Joshi, Emma Jackson, Sam Waldron,
Lucy Hart, Roy Bhakta, Claire Pillinger, Abbie Grace, Sarah De Jonge, Cathy
Bushnell, Damon Binning, Jen Clayton, and L, H, O, E, J & G (who cn txt BP
thru blu sky).
We also gratefully acknowledge the financial support of the following agencies/
organisations which have funded some of the research detailed in this book:

x Becta (who funded the intervention study described in Chapter 6);

x The British Academy (the Wood et al. (2011) longitudinal study described
in Chapter 4, which is based on project number SG-46555; the study
described in Chapter 6 is based on project number SG090928); and
x The Nuffield Foundation (who funded the ongoing research conducted by
Wood and Kemp into Grammar and Texting described in Chapter 7, which
was funded by project number EDU/38640).

And finally, 4 all r yng txtrs, thanx!

This page intentionally left bank
Chapter 1

Mobile phone use and the rise of

texting

Before we look at the rise of mobile phones, and on to texting more specifi-
cally, we look briefly at the historical development of technologically-mediated
communication, to contextualise the current use of mobile phones by adults
and children worldwide. We pay particular attention to the increased popularity
of and interest in text messaging as a function of these devices, and reflect on
the nature of text versus talk in this very specific context.

The rise of the telephone

Before 1876, when Alexander Graham Bell received the patent for a device
that could transmit and receive speech electrically (which we know as the
telephone), the only way real-time spoken communication could occur was
between speakers and listeners who were in more or less the same place. The
notion had been discussed for a few decades previously since electricity had
been harnessed for use. Bell was not the only inventor to create a function-
ing device; Elisha Gray had independently invented one, and Thomas Edison
was in competition, but the patent and the fame went to Bell, although the
name ‘telephone’ was Edison’s choice (Baron, 2008). Any communication
with someone at a distance required time between issuing and receiving a
message – sometimes a fairly lengthy time – depending on the means available
to transport a message, and another delay before a reply could be received.
The idea of real-time communication with someone at a distance allowed a
remarkable new way of thinking about communicating.
Before the telephone, two people who could speak with one another were
also likely to be able to experience the same environmental context through
other senses. People who could not perceive the same environment were not
just geographically distant, they were in some ways cognitively and emotionally
distant, because their current knowledge of themselves and their context, and
their knowledge and feeling about the other, was not able to be shared imme-
diately, and could only be imagined. We sometimes use the term ‘distant’ to
characterise a person – even if physically present – who is unwilling or unable
to share his or her state of mind with others. Writing would require describing
2 Mobile phone use and the rise of texting

something of the context from which one was writing, so that the recipient
would have other cues through which to appreciate the nuance of the words
used. The writer would also be unable to know clearly the context in which the
recipient would read the words, not only because of the distance that required
writing, but also because the recipient would read the words in a future at least
partially unknowable to the writer, by which time the context of the writer
would have changed in ways equally unknowable to the recipient.
The telephone minimised the elements of unknowable intervening circum-
stances, but – at least in the days before video-conferencing, computers and
Skype, and mobile phones with cameras – it did not allow the distant parties to
see the contexts the speakers were in at that moment. There was still a perceptual
distance that could mitigate against clear understanding, as might be appreciated
today if we compare a conversation with a person in the passenger seat while we
are driving, with a conversation with the same person through a mobile phone
while we are driving – even with a hands-free set. The passenger can see when the
vehicle in front brakes suddenly and swerves, but the person on the other end of
the phone can only imagine why the conversation suddenly ends, as the driver
attends to a more pressing situation.
However, responses in a real-time distant conversation could be nearly
immediate for the first time in history, and the distant person could be known
more clearly as present. The absent other could be held more closely in the
social circle of the present, and common, shared understanding more reliably
affirmed. The social circle of the present expanded considerably with the advent
of the telephone, and sharing the moment in conversation allowed access to a
more informal type of communication because of that shared understanding.
The telephone became domesticated – a process Silverstone and Haddon
(1996) described as the way a new item becomes a part of ordinary,
everyday living. This process developed over the next several decades after
its invention, and some readers may remember when a telephone was
first installed for their use. Landlines spread across the United States,
across Britain, Europe, under the seas, to much of the inhabited world,
although, even today, some may be waiting for a telephone landline to
reach them. Some of those, however, have been able to bypass the wait,
because satellite-relayed cell phone coverage has reached them before lan-
dlines. We touch on this situation later with regard to keeping minority
languages alive. During the 1950s the telephone became commonly
thought of not just as an information device, but as a device for social
purposes. However, the novelty of distant real-time communication has
long ceased to astonish. The telephone in some form is an essential part
of life in the modern world, and it is difficult to imagine life without it.
As part of its domestication, the name of the device itself became short-
ened through familiarity and common use to phone in common parlance,
and the noun forms telephone or phone came to have verbal forms, to telephone
or to phone, which have, through more familiarity, come to be represented
 Mobile phone use and the rise of texting 3

as to call or to ring, and have colloquial versions, such as I’ll give you a
buzz, or, for a text message, I’ll ping you. The same technological neolo-
gisms exist in other languages, for example Finnish, where to send a text
message is tekstata, a linguistic analogue to kuklata, to google.

Mobile telephones arrive

Until the 1970s, the mobility of a telephone was tied to the length of cord
connecting the handset to the body of the phone. Cordless technology freed
conversation from a small locale, allowing a conversation to carry on while
the speakers were multitasking (engaged in other activities as well), as long
as they were within reach of the radio signals that carried their voices. How-
ever, cordless handsets are only partially mobile, as long as they remain tied
electronically to the telephone land line and its transmission range. Cordless
phones were at first cumbersome devices with visible antennae but, in parallel
with miniaturisation of electronic technology generally, they have become
smaller and lighter. They began using analogue transmission (the way FM
radio is transmitted) and later switched to digital transmission, where the
analogue voice signal is converted into binary code and sent as a burst of 1s
and 0s, corresponding to the signal being on and off.
From 1947, the idea of cellular telephones was developed at Bell Telephone
Laboratories (Gertner, 2012) with the idea of the car phone, although the tech-
nology was not widely available until the 1970s. In cellular transmission, the
radio bandwidth employed to send the signal was assigned to hexagonal geo-
graphical cells with antennae at the corners. This enabled multiple users’ phones
to send signals at the same time in different cells, but using the same bandwidth,
and phones were designed to use this cellular system. If a call connected to
another phone outside the cell, the signal was relayed by the antennae through
successive cells to the cell in which the recipient could receive the signal. The
band being used for a call may also be divided on a time-share basis to allow
multiple calls using the same band.
The Global System for Mobile Communications (GSM) was formed in
1982, first as Groupe Spéciale Mobile (a European consortium), and since
1996 has also been available in parts of the United States. GSM uses a wider
band and more, shorter time bites than other systems, and allows email, internet
access and other functions that may not be supported by other systems. Over
200 countries have agreed to use the GSM system as a compatible network for
transmission, allowing mobile phone links between countries around the world
(Baron, 2008). The United States has not, at the time of writing, opted to use
a single, compatible transmission system with global coverage, although GSM-
compatible phones are available there.
The phones were called first cellular phones, then cell phones, then cells
in the United States, and called mobile phones, then mobiles in the United
Kingdom (we keep abbreviating!), and many users now just refer to their
4 Mobile phone use and the rise of texting

phones, assuming that ‘mobile’ or ‘cellular’ is understood. If they mean a lan-

dline phone, they may specify that. The assumption was originally landline,
the exception mobile, and this expectation has been widely reversed.
The spread and saturation of the market by mobile phones was rapid once
pricing policies became friendly to the user. In the first decade of the twenty-
first century, several countries, including Finland, Sweden, Italy, Hong
Kong and the United Kingdom, reached the point where there were at least
as many registered mobile phones as there were people. In our own research,
we have found that the age of obtaining a mobile phone for one’s own use
has fallen year on year, with children reporting ownership of their first phone
as young as five or six (Plester, Wood & Joshi, 2009). New research reported
in Chapter 6 demonstrates that same trend, discusses implications, and pre-
sents data relevant to the following earlier findings. Other research (Ofcom,
2010) shows that only 6 per cent of children aged five to seven years used a
mobile phone in 2009, but among children who were eight to ten years that
rose to 32 per cent, both figures a slight rise over 2007. In both age groups,
texting was more often reported than talking. Two-thirds of children up to
15 years had their own phone by ten years of age. Over three years, pre-teen
children’s confidence in using their mobile effectively has also increased year
on year, with confidence reaching over 90 per cent in 2009, and this reached
97 per cent by age 13, even as the proportion of children using their phones
for other activities than talk or text also increased with age. As smart phones
grow in popularity, it is sometimes effective with a new one, to ask a seven-
year-old how to use it!
Among 44 UK children aged between two-and-a-half and four years (whose
experience with techno-literacy was studied by Jackie Marsh), 26 had toy mobile
phones, and had been given them on average at 12 months. Where they did not
text themselves yet, they knew that their parents’ phones were used for both
talk and text (Marsh, 2004). The study we report in Chapter 6 shows similar
domestication of mobile phones with pre-schoolers. Today’s Digikids (Marsh,
2005) will enter school with a formidable amount of experience in technological
literacy, some of which will come through their mobile phones or their parents’
phones. One two-year-old known to us cheerfully and confidently packed his
father’s iPad in his backpack for his first day at pre-school, because he knows
how to access things he wants to learn about. The mobile phone – as with other
communication technology – is now as domesticated among pre-schoolers as
among adults. The astonishment that accompanied the first real-time spatially-
separated conversations has been superseded by our expectation to carry on
these conversations more or less anywhere and any time.

Text messages arrive

Letters on the traditional telephone 12-key pad and the dials that preceded
them were designed to help users remember numbers, and the first telephone
 Mobile phone use and the rise of texting 5

numbers gave the exchange code in letter form, for example ‘REdbank 6
1122’. But the keys marked with letters were not used to send messages con-
structed of letters (although recent advertising has occasionally returned to
the idea of remembering a phone number by its word equivalent, for example
‘phone CLEAR66’).
The first text messages were sent in Finland in the 1980s by Nokia engi-
neers exploring their potential, using up bandwidth not being used by talk
transmission, and a text facility was originally offered free of charge to phone
customers. The original 160 character limit per text was put in place in
1986 by Friedhelm Hillebrand, chairman of the nonvoice services commit-
tee within the Global System for Mobile Communications (GSM) on the
basis that most postcards and most Telex messages fit within that limit (Los
Angeles Times, 2009). Text messages were first sent by the general public
in Finland during 1995, and had been enthusiastically taken up by teenagers
by 1998 (Kasesniemi & Rautiainen, 2002). Because of the early develop-
ment of mobile phone technology in Finland, mobile phones rapidly became
domesticated into Finnish life, but originally were used mainly for business
information sharing purposes, as in many other places during the spread of
their popularity. Texts were often first used to notify recipients of other com-
munications, for example emails or phone messages. Possibilities for social,
phatic communication via texting grew rapidly as more phones were taken up
by young adults and, increasingly, teens and children.
Texting, rather than talking, has become the medium of choice for the
majority of European mobile phone users, and users in Italy, Japan and Korea
were particularly prone to mention texting before talk (Baron, 2010). In
the United States, take-up of texting has been slower, attributable at least
in part to longer availability and wider expertise with the personal computer
and internet access domestically. The mobile’s texting facility took longer to
become popular, but was enthusiastically embraced by the young. A set of
young US teens, speaking publicly about their digital lives in 2007, admit-
ted sometimes obsessive texting that could become a problem to them (Kids
Speak Out, 2007). Later that year, it was reported that there were, for the first
time, more texts sent than calls made in the United States on mobile phones
(Mindlin, 2008).

‘Perpetual contact’, ‘always on’

If one of the changes in perception brought about by the availability of the
telephone for distant real-time contact with others was that a wider circle of
other people could be brought into sharing one’s immediate context, there is
also a balancing discomfort. It is sometimes difficult to be alone, off-line, with
time enough for contemplative thought. One can turn off the phone, turn
off the computer, but it is harder to turn off the knowledge that we might be
missing a message. The absent others in our circles have a cognitive presence
6 Mobile phone use and the rise of texting

closer to conscious attention than they might if we had not so recently and so
often communicated with them, and we might expect a priming effect, where
we would be increasingly responsive to cues that would lead us to thoughts of
those others, and away from attention to our immediate activities.
Gergen (2002) has discussed the challenge of ‘absent presence’, where
technologies allow us to be cognitively and emotionally absent from our
immediate context in which we are physically present, essentially to be drawn
away from our immediate face-to-face circle. He grants that the mobile phone,
although it may have this potential, also has the potential we suggest here,
namely to draw absent others into our presence. Finding the balance between
those potentials is necessary to maintain the more complex communication
and relationship circles technology has afforded. Where possession of a mobile
phone may not mean that we wish to be always in contact with someone
else, we have come to expect the potential for contact as and when we wish.
We have also had to consider, for example, the implications of other time
zones, which puts functional limits on contacts, regardless of technological
availability.
Two areas in which cultural changes have been evident derive from the pos-
sibilities of a perpetual presence of absent others, and a perpetual absence of
our present selves. Where these changes are most evident in social behaviour,
they also have an impact on the use of language, which is the focus of this
book. One change is the development of a set of acceptable social standards
for use of mobile phones in public spaces. These vary from place to place
(Baron, 2008; Baron & Hård af Segerstad, 2010), from strong restriction of
talk on mobiles in public in Japan, to a more permissive standard of private
chat in public in Italy; even within countries, or between social groups, those
standards vary. Learning the local set of conventions is a requirement – or at
least desirable – for fluid social relations, as one moves from place to place,
or contacts others in places where the etiquette is different from one’s own.
In the case of children, there are locations that are defined by adults as
‘inappropriate’ contexts for mobile phone use, the most common of which is
on school sites. As we have seen, however, children may not share that defini-
tion, and this has the potential for conflict.
Another area of change concerns the extent to which one lives one’s pri-
vate life in the public gaze. When the circle of ‘present others’ extends across
the globe, many have responded by inviting a far wider circle of others to
share their personal lives. The enthusiastic embrace of social networking
sites on the internet demonstrates the ease and desirability of that increas-
ingly public stance. It is not our purpose here to discuss communication or
social relations conducted through the internet (many others have explored
these issues in depth, e.g. Baron, 2008; Crystal, 2006a; Danet & Herring
2007; Ellison, Steinfeld and Lampe, 2007; Johnson & Ensslin, 2007; Katz
& Aakhus, 2002; Rosen, 2007; Thurlow, Lengel & Tomic, 2004; Tong,
Van Der Heide, Langwell & Walther, 2008), but the blossoming of internet
 Mobile phone use and the rise of texting 7

social and personal presence is related to the form of language that is used
in ever greater swathes of individuals’ communication. Drawing such a wide
collection of others into one’s personal circle, where the register of language
can be casual and informal, as it has traditionally been in the smaller circles
of personal conversation, suggests that a greater proportion of one’s writ-
ten communication may be in a more casual, unregulated style. This casual
style also works to get across communicative intent, even among the present
absent, or it would not continue to be used.
It has been widely suggested that more people are writing than ever before
(e.g. Roschke 2008), with all the text that is created in text messages, emails,
instant messages, tweets, blogs, internet chat sites and social networking sites.
We have noted elsewhere (Veater, Plester & Wood, 2010) that children with
dyslexia text with as much enthusiasm as other children, where they have often
withdrawn from voluntary written language. The conventions of language pol-
icy established by users within those contexts have generally abandoned many
of the constraints of formal written language, because they are not required
in informal conversation, and those settings are often seen to be informal and
conversational in nature.
Baron (2008) has argued that, because there is so much written text,
because it is so easy to create that it is all around us all the time, standards
of good writing may be less falling than becoming irrelevant. Because we
are driven by the clock, and the ability to get our meaning across quickly
is important, formal written language rules have less importance than they
once had, and we take less pride in writing formally. Baron uses the term
‘whateverism’ to describe the attitude of accepting uncritically written text
that does not conform to traditional standards, even within situations where
it would seem that it should, such as publishers’ documents. Where Stanovich
and his colleagues (e.g. Cipielewski and Stanovich, 1992; Stanovich & West,
1989) have argued that exposure to the written word is a strong predictor of
literacy skills in young and older adults as well as children, the measures they
used would limit that exposure to well-crafted written language. Uncritical
immersion in informal text may not have the same power, or may have a dif-
ferent outcome, and research has addressed the question in various ways (see
Chapter 3).

Texting versus talking

The advantages of real-time mobile communication are accompanied by some
disadvantages in being always on and in perpetual contact, and the rise of text
messaging has addressed some important ones. First, a real-time conversation
is only possible if both parties are available, and texting does not require that.
One can send a message at one’s convenience, and one can receive a message at
one’s convenience. Speech and writing, however, have several fundamental differ-
ences that affect communication. Crystal (2006a, p. 28) has outlined some of the
8 Mobile phone use and the rise of texting

characteristics of communication by speech and by writing. If we apply these to

mobile communication (which has both information sharing and social aspects),
we may summarise them here, and consider whether, by these criteria, texting
might be considered at least as much conversational speech as writing.
First, speech is time bound and transient; the spoken word is gone once
spoken. The written word remains; one’s recall can be checked, a message
read again for further understanding. The permanence of the written word is
not easily dismissed. Indeed, the written word captures in flight, as it were,
spoken words, as spoken or written words may capture evanescent thought.
When texting was first adopted by the young of Finland in the late 1990s,
mobile phone companies brought out notebooks so that the texters could
save messages beyond the limit of the storage capacity of their phones
(Kasesniemi & Rautiainen, 2002).
Speech has no discernible time lag between speaking and hearing, and in
conversation there is little time for thinking out what we will say – unless we
have prepared in advance. However, we cannot always prepare for the kind of
response that will follow the first conversational turn, and ‘thinking on our
feet’ is required. Mistakes can be corrected, but the errors have already been
heard. Writing allows us time for working out the best way to express our
intent, for impression management, and to consider how it will be received.
Texting is asynchronous, but the gap may be very short, and conversational
turns can be effective, as previously noted by Kasesniemi and Rautiainen
(2002) and Thurlow and Brown (2003).
Reid and Reid (2007) explored this advantage of texting in their discus-
sion of the social effects of text messaging. Those who preferred texting to
talking reported higher social anxiety and more loneliness; they spent more
time composing their texts and wrote longer texts when compared with
those who preferred talking to texting. Texters seemed to maintain fewer and
deeper relationships through their texting than did talkers, creating a close-
knit social world of absent but present others through their texts, with whom
they could be in perpetual contact. This element could be a direct outcome
of the permanence of written words in contrast with spoken words. Reid and
Reid proposed that the overall results suggested that texting was perceived as
a safer mode to engage in intimate contact for those with greater difficulties
in face-to-face situations.
Speech often makes the assumption of shared knowledge between parties,
allowing the use of deictic expressions – those that refer to the situation of the
moment, such as this or that or here or there, or pronouns such as he or she or
they. When deictic expressions are used in writing, the referents may not be clear
at the time the message is read, so there is a greater chance for misunderstand-
ing and a time gap before correction is possible. Spoken conversation often
uses casual, informal language, complex constructions and lengthy utterances,
whereas written language may be more formal, more grammatically clear, even
within an informal relationship. Speech is useful for phatic or social functions
 Mobile phone use and the rise of texting 9

like reinforcing a relationship, using prosody and non-verbal features to express

nuances beyond the formal meaning of the words used. Written language is more
easily limited to the semantic content of the words, even used in a relationship-
building way, but for some (as above), the affordances of written language are
deemed the better mode of maintaining relationships. Emotional nuance in tex-
ting can, however, be indicated by emoticons, capitalisation or punctuation, for
example Hi!!!!! or Well.......... may... b... or How DARE you!, or x as opposed
to xxxxxxx. We discuss these uses in texting in chapters relating to texting by
children and texting by older teens and adults.
Texting seems to enjoy characteristics of both speech and writing, but that
boundary has been blurred widely in technological communication as young
people often say that they will talk to someone on Facebook, for example, or
since last we spoke in a recent email written to one of the authors by someone
with whom spoken communication has never taken place. Written talk is com-
mon for upwards of 95 per cent of eight- to fifteen-year-olds (Ofcom, 2010).
Texting seems to function in information-transfer ways, as well as relationship-
building ways. Ling (2007), Baron and Campbell (2010) and many others
have identified the former function as typically male, the latter as typically
female, although cultural differences must be considered. Texting is clearly
writing in a structural sense, but what kind of writing? Is texting language
really like casual speech, or do its characteristics go beyond that? Our tendency
to abbreviate words and phrases with shared understanding has found a fer-
tile field in texting, where messages were originally limited to 160 characters,
including spaces.
The previous sentence contains 184 characters and spaces, for comparison,
so it was in the interest of communicative intent that ways were sought to
get as much across as possible in that small space. Early pricing policies also
favoured efficient use of that space. However, when we look at text messages
actually sent by young adults with good verbal skills, they rarely use the full
space. Thurlow and Brown (2003) found that, of 544 UK young adult mes-
sages, the mean length was 14 words, and Ling and Baron (2007) found a
US sample averaged only 7.7 words. Ling (2007), investigating the use of the
predictive text function among Norwegian texters, found that those who used
predictive text used only marginally more letters per message than those who
did not, and the mean length of text overall was 29 letters long – about six
words. In the sentence of 184 characters above there are 27 words, and the
160 character space includes half of the 25th word. So why abbreviate at all if
messages do not use the full space available?
There is more than a spatial constraint operating in a text message, because
saving time is as important in communication as it is many other areas of
modern life. A spoken comment of 14 words or less would take little time,
and if we view a text as sharing attributes with a spoken comment – a conver-
sational turn – speed would be important to maintain the sense of an ongoing
conversation. Using the 12-key design of the earlier mobile phones, most
10 Mobile phone use and the rise of texting

characters take at least two keystrokes, punctuation as much as six or more;

so, to keep time and effort down, abbreviation where possible without loss
of meaning is an efficient strategy. Some texters may wish to add further effi-
ciency by using predictive text to cut keystrokes, but they seem not to write
longer texts if they do, and it is not clear whether predictive text saves time
for adults (Ling, 2007), although there is evidence that it may for children
(Kemp & Bushnell, 2011). As more texters use QWERTY keypads, it will
be interesting to see if the abbreviation conventions continue as they have
become established; some research we discuss later suggests they will. Con-
ventions may outlive one aspect of usefulness but be retained simply because
they are conventions and are useful in other ways.
Texting combines aspects of speech with aspects of writing in ways that
are seen largely as advantageous to texters. Texters frequently report that they
do not consider texting as writing at all (e.g. Pew Internet and American
Life Project, 2008; Roschke, 2008). First, the release from contemporane-
ous mobile communication allows us greater control over when and with
whom our communication will take place. A message remains for reading
later. Baron (2008) makes this point: that the technological developments
that have enabled us to communicate so widely can also give us the option of
turning down the volume of our communication, setting limits to our perpetual
contact with others. It seems ironic that we have embraced with enthusiasm
a medium with an option that is analogous to a return to the pre-1876 age,
but it reflects the mixed opportunities and obligations that the telephone
offers. Exerting this control allows us thinking time for managing the impres-
sion our message may give. Those who have lost jobs, friends or partners by
text message – or indeed the senders of such messages – might appreciate the
extent to which the sender can remain insulated from the immediate out-
come of such messages.
Though the time lag between sending and replying may be lengthy, it
need not be. Texting also enables the construction of shared context, which,
in turn, enables an informal style of conversation. Classmates surreptitiously
texting one another during a lesson (or in any space where both parties are
situated) show the advantage of silent texting for private communication in
public spaces, achieved much more easily through text than talk.
Because it has elements of casual speech and need not conform to more for-
mal written language, text language can afford play with language. As a species,
we enjoy word play, from nursery rhymes to playground games, nicknames and
teasing, to puns and limericks and crossword puzzles (Crystal, 1998). Text lan-
guage is sometimes described as slang (e.g. Yule, 2007; Ling, 2010), but slang
has been described by Adams (2009) as the people’s poetry. As we will see,
many of the abbreviations (or ‘textisms’) used in text messages have an element
of play about them. Not least is the element of intentional transgression of lan-
guage conventions, only possible if one knows the conventions to begin with.
This ludic nature is likely to be another reason for continuing to use textisms,
 Mobile phone use and the rise of texting 11

even when the space or time requirements do not demand abbreviation. Word
play here rests on the difference between the pronunciation of spoken, casual
register language or spoken accent, and the pronunciation of the same words
as they are read aloud from a standard English rendering. Textisms often play
to this difference.
It is curious that when a novelist renders a dialect or accent in print so that
the sounds of the spoken words will survive when read as they are printed,
the accomplishment may be seen as skilled, excellent writing. When a child or
teen writing a text does the same thing, it may be seen as a sign of the poverty
or deterioration of the writer’s language skills, if not of the language itself. In
the following chapter, we will consider the way text language has been rep-
resented in the media – the alarm widely sounded – and set that against the
language actually used in text messages by children and young adults. We will
also look at the way some educators and other commentators have responded
more positively to the phenomenon of texting.
Chapter 2

The media furore

In the early days of texting, the use of abbreviations and non-conventional

spellings attracted a great deal of media coverage, much of it highly critical
and speculative. Carefully selected anecdotes characterised any attempt at an
evidence-based argument. More recently, coverage has been somewhat less
about text language itself, or its potential threats to literacy, or the entire
English language. However, there are still critical voices, more likely ques-
tioning the process of texting itself: dangers that arise, for example, through
sleep impoverishment or inattention crossing roads (Steinhauer & Holson,
2008). Here we will focus only on material that is related to literacy, and
explore not just the media comments but scholarly analysis of those com-
ments. We will consider both the negative and the more positive accounts,
and then look at text language itself, from samples of actual texts contributed
by children, teens and adults.

Text language and the media

Thurlow (2006) used Critical Discourse Analysis to address 101 examples
of media discourse about Computer Mediated Discourse (CMD) by young
people, including text messaging. His samples cover the years 2001 through
2005, analysing what evaluative and semantic threads are privileged in the
data. Much of what was analysed was highly critical of text language, conflat-
ing it with declining moral standards, fear of technology and apparent decline
of the English language.
The first theme distilled out was labelled Marks of Distinction: CMD as
Linguistic Revolution. Here the focus was on CMD as it differed from stand-
ard English; its novelty is emphasised, its separateness from traditional or
mainstream English. Exemplars included neologisms like netspeak or weblish.
To this we can add our own label for text abbreviations or alterations, textisms
(or even txtsms), as we seek to distinguish the elements of language used in
text messages that differ from traditional expressions. In this book, we also
use the term textspeak to refer to the general register used when writing a text
message as opposed to more conventional writing, with the acknowledgement
 The media furore 13

that, within textspeak, the use of textisms may be pervasive, sparse or non-
existent, depending on the author. The separateness and equivalence of CMD
language related to standard English are enhanced by references to CMD as ‘a
new’ or ‘a second language’. Some of these media comments can be construed
as positive, for example ‘thousands of teens … are fluent in another language’
(Thurlow, 2006, p. 673). Some of the more enthusiastic are those phrases
emanating from commercial enterprises, who stand to gain from the embrace
of the new and different, for example ‘Text messaging might one day be as
popular as talking’ from an AT&T spokesperson (p. 674). The emphasis on
difference strengthens the view that these changes are sudden, rather than part
of a lengthy evolution of language.
The second theme abstracted was described by Thurlow (2006) as Statistical
Panic: The Rise and Spread of CMD. One key tool in this presentation was the
citing of large numbers, for example ‘billions of text messages are already being
sent every day in this country’, and ‘text totals of 2,000 to 3,000 per month are
common for older teenagers’ (both p. 675), and ‘This Thursday Britain’s phone
network is expected [to] be swamped by up to 60 million Valentine messages’
(p. 676). The numbers in various media accounts may conflict and be inconsistent,
but the impact is that of persuasion. Detail – whether spurious or genuine – tends
to establish authenticity.
The third theme was described as Moral Panic: CMD, Literacy, and the Social
Order. Uses such as youthquake, take by storm, mania, spreading like wildfire
generate a sense of the unstoppable; reprehensible, jarring and abrasive add a
pejorative tone to that (p. 677). One of the strongest sub-themes in this area
was the threat to the English language, for example slow death, dumbing down,
Generation Grunt (pp. 677, 678), and, further, that it extends far beyond the
language, for example The English language is being beaten up; civilization is in
danger of crumbling (p. 678).
The fourth theme was ROTFLMAO: The Fetishization of CMD. Many of
the media articles used some acronym or initialism in the title, and the exam-
ples used give the flavour of text language as being wholly or largely made up
of these. These kinds of textisms abound in anecdotally-reported messages,
and the so-called dictionaries of text language, emphasising emoticons and
‘hieroglyphs (codes comprehensible only to initiates)’ (Thurlow, 2006, p. 667,
from Sutherland, 2002), and the mutual incomprehensibility between textese
and traditional language. Throughout the corpus of commentaries analysed by
Thurlow, there is no attempt in the media articles to give sources or empirical
verification for the presented texts.
Most scholarly commentators reviewed in media excerpts were shown
by Thurlow to be more positive about CMD, but many of these were said
not to be new media specialists, and even those may have been quoted
out of context or interpreted inaccurately in the service of the article
author’s agenda. Media articles have claimed school examination scripts
to be riddled with text language or flawed because text language has crept
14 The media furore

in (p. 677), when the examination boards’ reports mention the occa-
sional text language in the same context as other spelling errors that are
common in examinations.
Two underlying difficulties in the media furore in the early days of texting
have been that commentators have leaned heavily on anecdotal, unattributed
or possibly fabricated evidence to support their claims of disaster to the English
language or culture itself, and they have been either uninformed or simplistic
in overlooking the multiplicity of causal factors that inform language change
over time. Has media coverage altered in more recent years since the samples
in Thurlow’s 2006 analysis?
In 2007, the Irish government issued a statement that claimed ‘Ireland’s
youth are becoming increasingly poor spellers and writers, and their love of
text messaging on cell phones is a major reason why, according to the gov-
ernment’s Education Department’. (Associated Press, 2007), Text messaging
corrupts all languages, so The Economist stated boldly in 2008 and cited by
Thurlow and Bell (2009), taking up again the question of adult framing of
adolescence.
‘“The act of texting automatically removes 10 I.Q. points,” said Paul
Saffo, a technology trend forecaster in Silicon Valley’ (Steinhauer & Holson,
2008). This was from an article in the New York Times. The same article
mentioned – without value judgment – that Barack Obama had announced
his running mate by text message, an event which generally received applause
in the media. The online eschool news (2010) reported that Professors Not
ROTFL [rolling on the floor laughing] at Students’ Text Language, showing
ways that university and college professors were tightening up on their students’
language registers.
In 2007, in the Daily Mail, John Humphrys (a provocative standard
bearer for traditional English) wrote an article headed I h8 txt msgs: How
texting is wrecking our language (Humphrys 2007). Humphrys began
by taking issue with the Oxford English Dictionary’s deletion of 16,000
hyphens from compound words, and then cast the blame for this and more
at the feet of texters, claiming that ‘texters … [are] vandals who are doing
to our language what Genghis Khan did to his neighbours eight hundred
years ago. They are destroying it: pillaging our punctuation; savaging our
sentences; raping our vocabulary. And they must be stopped.’ Humphrys
had acknowledged earlier (2004, p. 1) in describing teen language, that
‘They have adapted the language to suit themselves … but it works for
them’.
One can only wonder at Humphry’s response to the decision by the Oxford
English Dictionary (OED), made public in March 2011, to include LOL and
OMG and the heart symbol in the latest update. Within a few days of this
announcement, there had been 52,700 internet postings about the decision,
mixed in their support or outrage. The New York Times editorial on 4 April
2011 led with OMG!!! OED!!! LOL!!! It went on:
 The media furore 15

It’s wonderful to experience the ongoing corruption and evolution of the

English language. Last month, OMG and LOL were inducted into the
Oxford English Dictionary, along with the heart symbol — the first time
a meaning enters our most exalted linguistic inventory via the T-shirt and
bumper sticker.
(New York Times, 2011)

Dennis Baron, on 8 April 2011, on The Web of Language, countered:

Only the [New York] Times, the newspaper of record, got the story
wrong. Lexicographers at the OED, the dictionary of record, didn’t
chant ‘Imago verbiosà!’ and turn the kitschy j symbol into a word.
What they did was update the definition of the verb to heart to reflect a
new sense referring to ‘the symbol of a heart to denote the verb ‘love.’
(Baron, 2011)

On the same day, James Morgan (2011) from BBC News exclaimed,
OMG! LOL’s in the OED. LMAO!, and continued with a discussion about
the origin of LOL and parallels in other languages, such as 555 in Thai,
spoken as ‘ha ha ha’. The j of the media seems to have turned a bit softer
in some places.
The controversy over the value of text language did not escape the
advertising media; billboards and the sides of buses have made use of text
language. During 2007 and 2008, AT&T ran a series of television adverts
in the United States (known there as advertisements or ads – both ends
of the reduction continuum) in service of an unlimited texting package.
Three mini-dramas focused on a family purportedly in linguistic trouble,
with a perpetually texting daughter who speaks in initialisms in response
to her long-suffering mother’s comments, and even a texting grandmother
who sides with the daughter. Jones and Schieffelin (2009) analysed both
the scripts and the reframing of the adverts on YouTube, citing them as
evidence of the cultural iconicity the adverts attained and the characterisa-
tion of teens by adults. The adverts raised the spectre of text abbreviations
taking over even spoken language, but the valence of the three adverts –
although making fun of texters, which can be seen as critical – was gener-
ally seen as funny. There was a clear sense of word play in them, serving at
least to contradict a headline cited by Thurlow and Bell (2009) stating that
‘Texting leaves teens speechless’.
An interesting contrast in media views can be seen. In the first, from Julie
Henry (2002), the Times Education Supplement reported ‘Delete text mes-
sage style, say examiners’, warning that this poses new challenges for GCSE
(General Certificate of Secondary Education) markers, as well as suggesting
that the texting phenomenon could undermine children’s grammar. Four
years later, in 2006, the New Zealand Herald reported that the:
16 The media furore

New Zealand Qualifications Authority (NZQA) deputy chief executive

of qualifications Bali Haque said credit would be given in that year’s
NCEA (National Certificate of Educational Achievement) exams if the
answer ‘clearly shows the required understanding’, even if text abbre-
viations were used. However, abbreviations would be penalised in some
exams, including English, in which candidates were required to show
good language use.
(NZPA & Smith, 2006)

School principals, on the other hand, were not convinced that the former
decision was wise.
Despite the widespread concern that textspeak is ‘creeping’ into writing –
especially into students’ academic work – there is little empirical evidence that it
is at all widespread. There is some self-report evidence about the use of textisms,
from both the writers and readers of student work. For example, 25 instructors
of English at a US university reported seeing textism-like spellings in students’
submissions (National Council of Teachers of English, 2003) and, in a Pew
survey, 64 per cent of US teenagers reported using informal writing styles in
their school work, including initialisms (used by 38 per cent) and emoticons
(used by 25 per cent) (Lenhart, Arafeh, Smith & Macgill, 2008). In a natural-
istic study, Shafie, Azida and Osman (2010) looked for intrusions of textisms
into the written answers of Malaysian undergraduates in English examinations.
The authors note that ‘few’ textism-like abbreviations appeared, but do not
provide any numbers. In a more systematic study, Grace, Kemp, Martin and
Parrila (submitted, a) went through more than 300 examination papers written
by 153 Australian university students in a range of disciplines, totalling well over
half a million words. The number of spellings which could be counted as textisms
added up to 117, a tiny 0.02 per cent of the total words, with over a third pro-
duced by a single student. It does not appear that these students just happened
to be non-users of textisms: in the brief emails they sent giving consent for their
exams to be included in the study, textism-like spellings made up nearly 5 per cent
of the words written (41 of 855 words in total). These results suggest that, in
contrast to anecdotal reports, university students (at least) are able to confine
their use of textisms to written contexts in which it is more acceptable, and avoid
it in formal examination answers.
In 2008, the National Literacy Trust (NLT) described a scheme by Michelle
Herriman, focusing on fathers to encourage literacy in their children, which sent
out information about their Rhyme-Time programme and reminders by text
(National Literacy Trust, 2008). A similar texting strategy was reported as suc-
cessful by Glenda Revelle and her colleagues at the Sesame Workshop in 2007
(Revelle et al., 2007), with more than 75 per cent of participants believing that
the mobile phone was an effective learning tool, making it easy to incorporate
learning activities into everyday life. In February 2011, the NLT suggested that
teachers might send texts to their students with reminders of the next reading
 The media furore 17

group session (National Literacy Trust, 2011). A month earlier, they headlined
‘research suggests Bottom of Form text messaging can boost literacy among
pupils’. Texting is no longer widely seen as the scourge of language.
Indeed, in December 2011 in the New York Times, Tina Rosenberg (2011)
made a case for texting being a crucial factor in the preservation of otherwise
endangered minority languages, such as N’Ko, the standardised writing system
for the Mande family of languages in West Africa. One Mandinko speaker rea-
soned, ‘The ability to text in their own language would give people a powerful
reason to learn to read’. Africa is currently the fastest growing market for mobile
phones, because they enable isolated people to communicate widely, and texting
is the method of choice for financial reasons. Similar projects use digital technol-
ogy to preserve indigenous languages in parts of the world from Papua New
Guinea to El Salvador.
Humans have a long and wide history of abbreviated and informal commu-
nication born of the assumption of common understanding between speaker
and listener. In fact, William Safire (2009) claimed that the young even refer
to abbreviations as abbreves. He did not, however, blame it on texting. He
continued, ‘No tradition is more time-honored than rebellion against lin-
guistic tradition’. This view would be championed by David Crystal (2008,
2006b) in his robust defence of the robust quality of the English language,
even as it changes over time. Crystal (1998) has also long spoken for the play-
ful, ludic uses of language. Ammon Shea (2010) questioned whether texting
might be the stimulus for eventual spelling change in English, bringing it to a
higher level of transparency, stating, ‘through, rough, dough, plough, hiccough
and trough all end with -ough, yet none of them sounds the same as any of the
others, [this] is the sort of thing that has been vexing poets and learners of
English for quite some time’. To this, Crystal had commented that he thought
not, that texting does not portend spelling reform, it is a stylistic choice, but
he did leave open the possibility that bottom-up spelling change might be
effective where top-down spelling reform is not.
In an earlier column Safire (2008) stated that we were in our third phase
of language compression, the first being three centuries ago when contrac-
tions came into use, like can’t or I’m, causing ‘a 19th-century lexicographer to
denounce writers “carrying contraction to such an excess as to make their writ-
ings unintelligible to all but the initiated”’. This comment sounds very much
like more recent criticisms of text language, a line from Sutherland (2002) in
particular, ‘The dialect has a few hieroglyphs (codes comprehensible only to
initiates)’; so media criticism of linguistic novelty is nothing new. The second
phase, according to Safire, was the folding together of two words into one, such
as chuckle and snort becoming chortle, breakfast and lunch becoming brunch.
Another of those has become commonplace: a web log is a blog. The third
phase of compression is, according to Safire, when pauses and ums and ahs are
removed from broadcasts before transmission so as to avoid the speaker looking
indecisive, and in recognition of increased pressure for quick communication.
18 The media furore

In this, texting has it over talking; all the pauses for thought are invisible to
the recipient, adding to the confidence with which the socially insecure can
communicate by text (Reid & Reid, 2007). A compression like this uses one
of the language characteristics typical of written language, for example its release
from constraints of contemporaneous discourse. To continue to compress the
language – as long as there is common understanding between the sender and
receiver of the message – is a process that is in line with the history of the English
language.

Text language and the texters

Casual registers of spoken language differ from traditional written language,
with ostensible transgressions permissible. In spoken language, the conven-
tions and rules of grammar of more formal written language may not be
observed, and the communicative intent of the speaker may be carried as
much by prosody as by semantics. Textisms can be graphic forms of prosody.
We often give nicknames to close associates, use contractions like can’t and
won’t, grammatical transgressions like see ya or dunno or fancy a cuppa? in infor-
mal speech, and use initialisms like BBC or CNN, asap (as soon as possible), fyi
(for your information), and invent terms like yob (a backward boy, used to refer
to objectionable youth) or slush fund when parties to the communication are
expected to know what they mean. The conversational register of language that
texting allows through private communication (even in public) combines with
space limitations and a desire to get things done quickly, to invite and encourage
abbreviated language that may require shared understanding for effectiveness.
How widely the forms are understood determines the range within which they
may be used, if genuine communication is the aim of the message. We will not
discuss here sociolinguistic uses of language in wider impression management, to
communicate, for example, status or position, group membership or lack thereof.
There are available on the internet (and in paper format for those who still
consult paper books) many glossaries and translation tools alleged to bring text
language – or ‘textisms’ – and traditional standard English together. These often
make most use of initialisms and acronyms, giving little space to other forms of
abbreviation, and giving the impression that text language is made up of these,
and that text messages are made up largely of textisms. One shortcoming of
these tools is that they often do not indicate how commonly any of the items
are used in real text messages, only that they have been used. What do texters
actually write spontaneously?
One early study (Thurlow & Brown, 2003) of 544 actual text messages con-
tributed by UK young adults demonstrated that a wide variety of abbreviations
were used, but that they constituted a relatively small proportion of each message –
about 19 per cent overall. Others have found even fewer, for example Crystal
(2008) 10 per cent, and Baron (2008) 5 per cent. Thurlow and Brown identified
six basic categories using alphabetic forms (after Shortis, 2001):
 The media furore 19

1 shortenings and contractions, with letters missing from the middle or

end of words, such as bday for birthday, or Tues for Tuesday; g-clippings
and other clippings of single letters from the end, such as goin, hav;
2 initialisms and the pronounceable subset of acronyms;
3 letter/number homophones such as c u l8r 2 for see you later too;
4 misspellings and typos;
5 non-conventional spellings, which respect standard English phoneme-
grapheme conversion rules, such as nite or sori, but are not the conventional
spellings; and
6 accent stylisations, which render in written form the non-standard pro-
nunciations of casual register or regional dialect speech, such as afta for
after, or wassup (or sup) for what’s up as a greeting.

They also listed the symbols used, including emoticons. Initialisms, acronyms,
symbols and letter/number homophones made relatively few appearances
overall, but more instances of accent stylisation and non-conventional spell-
ing occurred. Of 509 symbols, most of them were sets of !!!!! or xxxxx; only
39 were emoticons. Only 73 homophones were used overall. Thurlow and
Brown included examples of all textisms found, but give no other frequencies.
For a brief review of varieties of text language used across countries, cultures
and languages, see Thurlow and Poff (2009).

Frequencies of textisms of various types from the

elicited and spontaneous text corpus
Our own research with both elicited and spontaneous text messages written by
pre-teen children (Plester, Wood & Bell, 2008; Plester, Wood & Joshi, 2009;
Plester & Wood, 2009) has used these categories. Figures 2.1 and 2.2 below
(based on data reported in Plester & Wood, 2009) demonstrate the relative
frequency of the various types of textisms generated by these pre-teens. The
number at the top of each bar is the mean number of that kind used over a
set of ten elicited text messages. Over several studies, the younger texters have
tended to use a higher proportion of textisms, between 30–40 per cent, sug-
gesting that they knew what seemed to be expected of text language. Their
most frequent types of textism have also been generally phonetically based:
contractions, homophones, non-conventional spellings and accent stylisations.
However, in the children’s own spontaneous texts, the only textisms used fre-
quently were homophones – a very different result from that found by Thur-
low and Brown (2003) with texters a few years older. Similar findings were
reported by Coe and Oakhill (2011) in elicited text tasks, with homophones
and accent stylisations most likely. We now know that the patterns of textisms
observed across studies differ not only with the age of the texters, but also with
the time/country of testing and the technology available at that time/place,
as well as the text collection method used (Grace, Kemp, Martin & Parrila,
20 The media furore

2012). Conclusions about the use of textisms thus need to be made in the
context in which they were gathered.

1.4
1.19 1.18
1.2
Mean number used

1.0
0.83
0.8
0.64
0.6
0.4
0.26 0.25
0.18 0.19
0.2 0.12 0.12 0.09
0
s

ns

ls

es

gs

ns

gs

es
ng

ng

ng

m
bo
io

on

llin

io

lin

ph
is
ni

pi

pi
ct

at
al

el
ph

tro
pe
te

lip

ip
tra

Sy

is
iti

sp
cl
or

os
-c

ls

yl
In
on

is
Sh

st
er
G

na

ap
M
o
C

th

nt
H

io

ng
O

ce
nt

si
Ac
ve

is
on

M
nc
U

Type of textism

Figure 2.1 Textisms used in elicited texts by type of textism.

4.0
3.54
3.5
Mean number used

3.0
2.5
2.0
1.5
1.0 0.78 0.83
0.67
0.44 0.41 0.48
0.5 0.19 0.22 0.3
0.11
0
s

ns

gs

ls

gs

ns

gs

es
ng

ng

ne
bo
io

lin

io

lin

ph
is
ni

pi

pi

ho
ct

at
al

el

el

tro
te

ip

lip
tra

Sy

is
op
iti

sp

sp
l
or

os
-c

l
In

ty
on

om

is
Sh

er

l
G

na

ap
M
C

th

nt
H

io

ng
O

ce
t
en

si
Ac

is
v
on

M
nc
U

Type of textism

Figure 2.2 Textisms used in spontaneous texts by type of textism.

 The media furore 21

Another important factor in textism use is the language in which the texts are
written. This is particularly clear in a study of Finnish pre-teen text language
(Plester, Lerkkanen, Linjama, Rasku-Puttonen & Littleton, 2011), which
showed a different pattern of textism use, with a very large proportion of tex-
tisms being accent stylisations. Part of the explanation is that the Finnish language
does not lend itself to the creation of homophones as English does. Part of
the explanation may also be that the Finnish language has an abbreviated
form of spoken language used widely among people of all levels of education,
but which is rarely written, except by the poorly educated, or when a dialect
or a dialogue is being represented in writing. Finnish is an agglutinative lan-
guage, where grammatical inflections are added on to stems and, in spoken
Finnish, some of the inflections are omitted when shared understanding is
assumed. Mun kans is the spoken form of minun kanssani, which means with
me. Eiks niin? is spoken, eikö niin? is written, meaning isn’t it so? with the kö
indicating the question, and the ks a marker of spoken language. This register
of language was ideal for texting, and Finnish pre-teens have readily adopted
it. The case of Finnish highlights how children who speak languages other
than English generate textisms in ways that reflect the features (constraints
and affordances) of that language (and adults do too, as we shall see in Chap-
ter 5). However, as we shall discuss next, in this medium, children also code
switch in order to communicate effectively.

Code switching and heteroglossia

With young people around the world using the internet for information and
also for other purposes such as gaming, many young people become com-
fortable with the use of words from other languages as supplements to any
second or third language they study formally. Play using language – and play
with language itself – are important parts of the increased engagement that
young people have with written language because of technological oppor-
tunities. We have seen the way English speaking young texters code switch
between standard English and textisms, many of which have a clear element
of language play.
Some illustrations of textisms from other languages will demonstrate the often
ludic nature of digital language among a wider set of texters. Leppänen and
Piirainen-Marsh (2009) have detailed the way young Finnish people incorporate
English – and even Japanese – into their online gaming conversations. Following
from this, we found a small amount of English in the Finnish texts we analysed,
including sori, spelled as Finns would, but an English word nonetheless. The
Finnish would be anteeksi (Plester et al., 2011). Letter/number homophones are
not found in Finnish textisms, but they are found in other languages, sometimes
indicating heteroglossia as well. One example (Crystal, 2008) is the use of 39 in a
Japanese text, which would be pronounced as san kyu, and is used for thank you.
3Q serves the same purpose in Chinese, and a Chinese 88, pronounced baibai is
22 The media furore

used for bye bye. 555 in Chinese would be pronounced wu-wu-wu and is used as
a whimper, where 555 in Thai would be pronounced ha-ha-ha and serves as the
equivalent of LOL in an English language text. A Dutch texter might type b&,
pronounced ben, meaning am. A French texter might type @+, pronounced à
plus, meaning the same as an English texter would when writing c u l8r. Texters
have fun with words.
As a coda to this discussion of the text language that texters actually do use
(as against media descriptions), we might look again at the spectre of text lan-
guage replacing traditional language in an article by Kate Kelland (2008) in The
Telegraph. Users of the predictive text function (where the pressing of the first
keys brings up suggestions for what the rest of the word might be, based on the
likelihood of using those keys) have created some interesting alternative words
by choosing the first alternative that comes up as a suggestion. These have been
taken up in light hearted play, where cool becomes book, awake becomes cycle,
and barmaid becomes carnage, eat becomes fat, and woohoo becomes zonino.
Whether this is another instance of adults characterising young people, with
which Thurlow and Bell (2009) have taken issue strongly, or a description of
language actually used by young texters, remains to be seen. We have seen no
instances like these in texts collected in our research. Letting the texters speak for
themselves has underpinned our programme of research.
In the following chapters, we address the questions raised by the alarm in
adult media framings of the language of the young. What do the actual data
tell us, when we look at text language used by texters from pre-teens to adults?
Does their text language herald deficiency in their traditional literacy skills? Or
is literacy safe in their texting hands?
Chapter 3

The links between children’s spelling,

reading and texting

So far we have seen that there has been a good deal of concern expressed
about the potentially negative impact of text messaging on children’s lit-
eracy development. However, these arguments have been made – and largely
accepted – without recourse to empirical work which has the potential to
shed light on these issues. This chapter therefore begins with a summary of
experimental work, which has been concerned with the more general issue
of whether exposure to misspelled words can impact negatively on adults’
and children’s literacy performance. This is followed by a review of stud-
ies which have more specifically examined whether there is any evidence of
a negative association between the ability to perform well on standardised
tasks of conventional literacy outcomes, and the nature and extent of text
messaging behaviours.

Does exposure to misspellings in general harm

children’s literacy?
The previous chapters have outlined the rise in mobile phone ownership in
general and of children and young people specifically, and the widely-reported
concerns to do with the impact that text messaging in particular must be
having on children’s developing understanding of conventional spelling and
related written language skills. This debate has led to the development of a
clear thesis: that children’s texting behaviour and their use of text messag-
ing abbreviations or ‘textisms’ are causally linked to literacy, and that any
observed associations between these two domains will be negative. Woronoff
(2007) proposes that exposure to textisms will inevitably affect children’s
memory for the correct form:

There is no problem among older people because their spelling skills are
more established. Children are more prone to commit errors because
they have read less, and prefer to play games, or watch TV, etc. Much of
their time is influenced by what is going on in their environment. So we
have to be watchful that they not look stupid because they cannot spell
24 Children’s spelling, reading, texting

simple common words. Texting has come along with a flourish, making a
big impact among them. This habit forming menace can influence kids to
spell incorrectly or get confused about the correct usage. We should not
tolerate these activities, else it might endanger their progress. Many com-
mon daily words have been shortened by SMS. It is likely that it might
affect much of their ability to spell, since their minds are in the formation
stage. Can we find means to minimize their use or remind them that
texting dulls spelling?

This quote neatly summarises popular assumptions about how texting and
use of text abbreviations must be affecting children’s literacy development
and their spelling development in particular. However, such beliefs were held
(1) in the absence of specific empirical data which examined the associations
between texting and literacy, and (2) without reference to the empirical work
which had examined the impact of exposure to misspellings on children’s and
adults’ understanding of conventional spelling.
Let us take Woronoff’s claim about there being ‘no problem’ for adult
texters, because their understanding of spelling is more established. If we con-
ceptualise textisms as forms of misspelling, then there is in fact a consistent
literature which has demonstrated that, when adult participants are exposed to
a misspelled word, this can result in a decline in spelling performance for the
correct form of that word, although exposure to the correct form can enhance
spelling performance (e.g. Brown, 1988; Dixon & Kaminska, 1997; Jacoby &
Hollingshead, 1990; Katz & Frost, 2001). So the assumption that adults are
immune to the potentially negative effects of textism use is not supported by
the systematic research in this area, which suggests that implicit (unconscious)
memory processes can impact on established orthographic representations.
However, when we examine the research which has been conducted
with children, this negative orthographic exposure effect is not observed.
For example, Bradley and King (1992) gave 11-year-old children sentences
to proofread, which contained underlined words which were either spelled
correctly or incorrectly. Children were asked to indicate if the spelling was
correct. The design of the study was such that the children were allocated to
one of three groups: they either always saw correctly spelled items, or they saw
items that were misspelled half of the time, or they saw items that were consist-
ently misspelled. Bradley and King found that exposure to misspellings did not
significantly affect the children’s spelling of those words, but that exposure to
the correct spellings did enhance their spelling accuracy, even when they only
saw the items correctly spelled half of the time. Dixon and Kaminska (2007)
also conducted a study which examined the orthographic exposure effect in
10-year-old children, which paralleled the work conducted with adult samples
more directly. They also found that there was evidence that exposure to the
correct form of a word could enhance spelling ability, but there was no
evidence of any detrimental impact of exposure to misspellings.
 Children’s spelling, reading, texting 25

In more recent work, Powell and Dixon (2011) extended their orthographic
exposure work to examine the potential effects of textism-like misspellings on
adults. They assessed adults’ spelling of a series of often-misspelled words,
both before and after they had been presented with those same words writ-
ten: correctly (e.g. tomorrow); as textism style misspellings (e.g. 2moro); or as
well as more conventional misspellings (e.g. tommorrow). In line with previous
adult studies, participants’ spelling of the target words was worse after expo-
sure to conventional misspellings, and better after exposure to correct forms.
Unexpectedly, however, exposure to textspeak-style misspellings actually led to
slightly better spelling of the target words. Powell and Dixon suggest that tex-
tisms provide a partial cue to the spelling of the word they represent, and thus
‘prime’ the writer to spell the word correctly. The authors acknowledge that
more conventional misspellings also provide partial spelling cues, but suggest
that because textisms often transgress conventional spelling rules (for example
by including numbers or omitting vowels), they do not interfere with people’s
stored memories for word spellings.
We have some research currently in progress with Damon Binning and
Hélène Deacon (Binning, 2012) that uses a similar paradigm to assess the effects
of textism-like misspellings on adults’ and children’s memory for word spellings.
We looked at the effects of two kinds of textspeak misspellings (phonetically
spelled forms and vowels-omitted forms), and employed newly-learned words,
rather than familiar words. Specifically, we asked participants to read aloud a
series of novel words (e.g. dreager) embedded in written passages, and then re-
showed these novel words in a variety of forms: spelled as before (dreager); or
conventionally misspelled (dreagar); or respelled in one of two textism styles,
with vowels omitted (drgr) or spelled phonetically (dreega). We then re-showed
these spellings either on paper (where errors are typically not tolerated) or on a
mobile phone screen (where misspellings are tolerated, even expected). Finally,
participants were shown the four possible respellings of each word, and asked
to identify the spelling that they had originally read. We found that partici-
pants were generally not confused by having seen the respelled versions: both
adults and children were significantly more likely to correctly identify the orig-
inal spellings than to erroneously choose alternative spellings. Further, when
they did pick a wrong spelling, it was significantly more likely to be a more
conventional-looking misspelling than a phonetic form, and least likely of all to be a
vowels-omitted version. In contrast to our expectation, there was no difference at
all between novel words that had been presented on a piece of paper compared
with those presented on a phone. More data will be necessary to confirm these
findings, but for now it seems that children’s (as well as adults’) ability to spell
even newly-learned words is not immediately ruined by exposure to alternative
spellings of words, whether new or familiar.
However, it is worth remembering that, in the case of text messaging, it is
not just that children are exposed to ‘incorrect’ forms of words, but also that
they are active in creating their own alternative spellings. It is possible that it
26 Children’s spelling, reading, texting

is the creation of these alternative spellings which is linked with spelling dif-
ficulties for children. Again, there is existing experimental work which has
relevance to this issue. Ehri, Gibbs and Underwood (1988) asked children of
primary school age to create spellings for pseudo-words which had been read
aloud to them. The children were then told that the version that they had cre-
ated was not the correct form, and they were shown the corrected spellings for
these items, which they were then required to learn. Their ability to learn the
correct forms of these new words was not affected by their previous experience
in creating incorrect spellings, and this pattern was also observed for a sample
of adults. This suggests that the act of creating misspellings also does not
impact on the ability to learn the correct spellings.
Initially, these findings may seem surprising, but on reflection they also
make practical sense. That is, in the course of learning to read and write, a
child will come to misspell words and to read both their own misspellings
and those of other children. If exposure to the incorrect form of a word
impacted negatively on children’s ability to learn how to spell the correct
form, then children would find learning to spell an almost impossible task.
What these studies suggest is that the way that children and adults process
text is different in important ways, and that children’s still-developing
cognitive systems are more flexible in some ways than those of adults.
However, both adults and children benefit from exposure to the correct
forms of words, and both adults and children are not harmed by creating
their own (incorrect) spellings for new words.

Texting and academic achievement

So it would seem that exposure to misspellings does not significantly inter-
fere with children’s developing understanding of conventional orthography.
However, can we assume that textisms are processed as misspellings? For
many children it would seem that texting slang represents not a series of
misspellings, but an alternative orthographic form in its own right. Given
this distinction, it could be the case that textisms are represented cognitively
in ways which may conflict with more conventional orthography, and so
empirical work which specifically examines textism use in relation to chil-
dren’s cognitive abilities as well as their competency in reproducing standard
spellings was warranted.
The first study that attempted to address this issue was developed by
Plester, Wood and Bell (2008), who conducted two simple studies that
looked for evidence of significant associations between children’s academic
abilities and their mobile phone use. In the first study reported, a comparison
was made between 11- to 12-year-old children who were classified as high
text users (children who sent three or more text messages a day – three mes-
sages a day was the median reported figure for this sample of children), low
text users (children who texted but did not send more than three messages
 Children’s spelling, reading, texting 27

a day), and children who did not text message at all. The children’s perfor-
mance on the Cognitive Abilities Test (CAT), a measure of general ability
which is used by UK schools to predict Standard Attainment Test (SAT)
scores, was obtained, and the children were also asked to translate standard
English into textspeak and a textspeak message into standard English. The
proportion of textisms that children used in their textspeak messages and the
number of errors made when translating text messages into standard English
were assessed.
With respect to volume of text messages sent, there was evidence that, as
texting increased, children’s performance on the CAT decreased. This finding
needs careful reflection, as it seems unlikely that the act of pressing buttons on
a phone keypad somehow reduces one’s general cognitive abilities, or indeed
that a person’s general level of ability somehow impacts on his or her desire to
send text messages. A more likely – although tentative – interpretation would
be to suggest that there is a third variable at work which links the two sets
of scores, such as socio-economic status. That is, at the time that this study
was conducted, often the children in this age group who did not own mobile
phones (and therefore did not text) were children from families that could be
described as ‘middle class’, or at least children who came from homes where
the parents exerted greater control over and showed more investment in their
children’s extra-curricular activities and academic achievement. The popular
perception of mobile phones having potentially negative effects may have fed
into this effect.
What was more striking about this study was that there were no significant
differences across the three texting groups (high users, low users and non
users) in either the number of errors that they made using standard English,
or the proportion of textisms to real words that they used in the translation
tasks. Moreover, it was found that there was a significant positive correlation
between performance on the verbal subscale of the CAT and the tendency
to use text abbreviations in the translation task: the children who wrote the
most densely abbreviated messages had the best verbal cognitive abilities.
This finding led Plester et al. (2008) to follow this study up with a small-
scale investigation of whether there were any links between textism use and
spelling ability specifically. As before, the children were asked to complete a
translation exercise and they were also asked to complete a standardised test
of spelling ability. Not only was there no evidence of any negative association
between spelling ability and use of text abbreviations, the relationship was
actually significantly positive.
Again, this may seem surprising at first glance, but not if you consider the
nature of text abbreviations (see Table 3.1). That is, the majority of the most
commonly used forms of textism are phonetic in nature. That is, to produce
and read these abbreviations the child would need a reasonably good level of
phonological awareness (awareness of the patterns of sound in speech) and
28 Children’s spelling, reading, texting

Table 3.1 Summary of the various forms of textism used, based on typology from Thurlow
and Browna

Textism name Examples

Shortenings bro tues

Contractions txt hmwrk
G-clippings swimmin goin
Other clippings hav wil
Missing apostrophes cant dads
Acronyms BBC UK
Initialisms ttfn tvm lol
Symbols @ - <) xxx
Homophones 2moro l8r
Unconventional spellings fone rite skool
Accent stylisations wiv elp anuva gonna

Note
a Thurlow & Brown (2003).

knowledge of how these can map on to either letters or numbers. The map-
ping of phonology on to text is more commonly known in the classroom as
‘phonics’. Both phonological awareness and the application of phonics are
essential aspects of learning to read and write, and so, by creating and reading
textisms, children are in fact rehearsing the very skills that are important for
the acquisition of conventional literacy.
The initial study by Plester et al. was very limited in its methods and scope,
and was essentially a pilot study in the area. Given the results obtained –
especially with respect to spelling – Plester, Wood and Joshi (2009) con-
ducted a more extensive study of the concurrent relationships between
knowledge of textisms and literacy performance in a sample of 88 10- to
12-year-old children. The purpose of this study was to replicate the pattern
of results found previously, and to try to understand the factors contributing
to the positive association between literacy and textism use. In particular, it
was expected that individual differences in phonological awareness would
explain the link between textism use and literacy.
The children completed standardised assessments of vocabulary, short term
memory capacity, word reading, non-word reading and spelling. They also
completed two assessments of phonological processing. To assess textism use,
rather than use a translation paradigm (which could be seen as encouraging the
children to use textisms when perhaps they otherwise would not), the children
were asked to imagine that they were in a series of situations and to write down
the text message that they would send to their friend. The reason for elicit-
ing the messages in this way rather than ask the children to bring their phones
to school and provide actual samples was an ethical one, as the schools typically
 Children’s spelling, reading, texting 29

banned mobile phones from campus, and so this technique was used as a proxy
for actual texting although see Chapter 8 for a discussion of the pros and cons
of text collection methods.
In this study, textism use was found to be most strongly associated with
reading ability. A regression analysis was conducted to see whether tex-
tism use could explain any individual differences in reading ability after the
following factors had been taken into account: the age at which the chil-
dren received their first phone; short term memory; vocabulary; non-word
reading; and phonological awareness. In this highly conservative analysis,
textism use was still observed to explain a small but significant amount of
the variance in reading scores. So it would seem that phonological aware-
ness and the other variables were able to explain much of the relationship
between texting and reading ability, but somehow textism use was also
independently associated with reading ability.
What could the nature of this independent contribution be? There were
a number of possibilities. The first was that perhaps reading benefitted
from textism use, because of the underlying exposure to print that chil-
dren who owned mobile phones would inevitably have. That is, mobile
phones can be seen as devices that present their owners with daily practice
at both reading and spelling through the text message function. This prac-
tice may be additional to the exposure to print that the children receive
from other sources, rather than a substitute for it. If this is the case, then
this extra print exposure should benefit the children’s reading scores (e.g.
Cunningham & Stanovich, 1990) and also their spelling (e.g. Stanovich
& Cunningham, 1992). Another possibility was that the creative experi-
ence of playing with language in the way demonstrated when texting and
creating textisms was motivational, or prompted the children to reflect on
the nature of orthography in some way.

The impact of input method: how does predictive

text use contribute to literacy?
A common question posed about the results of these studies concerns the chil-
dren’s use of the predictive text function of their mobile phones when texting,
and whether this is linked positively or negatively to literacy outcomes. Pre-
dictive text refers to the software feature of most mobile phones which, when
enabled, allows the phone to anticipate the words that you might be trying to
type as you are typing, and suggest them to you, so that you can simply select
the correct alternative from the options available rather than finish typing the
word in full. Many children report that they keep this function turned off, as
it interferes with their ability to use textisms in their messages, although some
mobile dictionaries now include the most common textisms so that they can
also be selected through predictive text. Kemp and Bushnell (2011) looked at
30 Children’s spelling, reading, texting

the impact of input method on a sample of 10- to 12-year-old children’s use of

textisms, reading and spelling performance. They found that children who used
predictive text were faster at typing messages than children who used multi-press
entry. Children who used predictive text were also faster at reading aloud text
messages written in textspeak than were children who used multi-press entry,
although there was no significant difference in the time the two groups took
to read conventionally spelled text messages. Children who did not usually
text, but who used multi-press entry for this study, were slower than their peers
(whether they used predictive or multi-press entry) at writing text messages, but
similar or faster at reading text messages. Despite these findings, there were no
significant differences between the three groups of children (predictive texters,
multi-press texters and non-texters) on any of the literacy measures. However,
there were significant positive correlations between measures which required the
processing of textisms and performance on the measures of reading, non-word
reading and spelling ability. Specifically, the time that the children took to read
aloud messages written in textspeak was able to predict individual differences
in all three literacy measures, after taking into account both age and the time it
took to read aloud messages written in conventional English. Overall this study
further supports the argument that the ability to process textisms is linked to
positive literacy outcomes, but the children’s use of predictive text does not
seem to impact on their performance on standardised tests of literacy.

Textism use and reading difficulties

This pattern of concurrent relationships suggests that the associations
between textism use and literacy are positive, but is this likely to be the case
for all readers? It will be recalled that phonological abilities are important
in contributing to both reading and textism use. However, children with
literacy difficulties such as dyslexia are characterised by deficits in phono-
logical processing (Ramus 2001, 2003; Ramus et al. 2003) and so, for these
children, the nature of the association between texting and literacy may look
very different.
Veater, Plester and Wood (2011) compared 13 children with developmen-
tal dyslexia (mean age 11.8 years) to 26 children with the same reading ability
(reading age matched controls) and 26 children of the same age (chrono-
logical age matched controls). All the children completed an assessment of
verbal IQ, an assessment of reading and non-word reading ability, a measure
of phonological awareness and provided a sample of text messages that they
had actually sent, which were then analysed and coded to indicate the type of
textism that they used when texting. The results of this analysis showed that,
although there were no significant differences between the three groups in
the overall proportion of text abbreviations that the children tended to use
when texting, the children with dyslexia tended to use initialisms and symbols
more frequently than the non-dyslexic children, and appeared to use fewer of
 Children’s spelling, reading, texting 31

the more common phonetic abbreviations, such as homophones and accent

stylisations. As expected, for the children with developmental dyslexia, there
was no significant correlation between textism use and phonological aware-
ness, unlike the two control groups. Overall, this pattern of results suggests
that children with dyslexia avoid the use of text abbreviations which require
phonological processing to decode or create them and, as a result, there is
little evidence of a link between phonological awareness and use of text
abbreviations for these children. Whether or not texting may offer a safe envi-
ronment for children with dyslexia to practise phonic work is open to debate,
as these children’s phonological processing deficits would need to be sup-
ported to enable them to fully engage with this approach.
Similar work was conducted by Coe and Oakhill (2011), who recruited
41 10- and 11-year-olds and split them into good and poor readers, based
on their Key Stage 2 SAT results for English. Similar to the results of Plester
et al. (2008), Coe and Oakhill found that the poorer readers reported using
their mobile phones more (for calls and texting) than the better readers did.
However, the better readers used significantly more textisms when asked to
write a text message, and could read a message written in textspeak faster than
the poorer readers. This fits with previous findings, that better reading ability
is linked to greater proficiency around the use of text abbreviations. How-
ever, in this study, it was the poorer ability children who reported the greatest
use of their phones. This finding thus suggests that the link between textism use
and reading ability cannot be explained simply in terms of the increased print
exposure that comes with increased use of texting.
Another possibility emerges in a study of similarly-aged Chinese children
with dyslexia reported by Hsu (2013). Chinese is a non-alphabetic script,
morphosyllabic, with each character usually containing a semantic and a
phonetic cue. Morphological awareness (knowing the rules of Chinese word
formation from its constituent strokes) is required for reading as well as pho-
nological awareness. Hsu found – as Veater et al. (2011) had – that children
with dyslexia tended not to choose phonologically-based textisms, and that
they were significantly poorer on phonological awareness measures than read-
ing age matched peers. In measures of morphological awareness, however,
the Chinese children with dyslexia were not significantly different from their
reading age matched peers. Hsu also found that they did tend to use mor-
phologically-based textisms, and that their morphological awareness scores
were related to the textisms chosen far more than they were for the reading
age matched control group. Hsu also raised the question of whether the
safe environment of texting (or posting Facebook messages, as these chil-
dren did) allowed dyslexic children to feel able to write more freely. Further
research might explore the morphological/phonological skills questions in
other non-alphabetic scripts. Further research might also investigate the role
of morphological awareness in alphabetic texting, with dyslexic texters as well
as competently reading texters.
32 Children’s spelling, reading, texting

One further study to consider the use of texting in individuals with language
or literacy difficulties is by Durkin, Conti-Ramsden and Walker (2011), who
examined the use of textisms in British 17-year-olds with and without a specific
language impairment (SLI). As in previous research, there were significant posi-
tive links between the use of textisms and literacy scores, in adolescents both
with and without SLI. However, the 47 adolescents with SLI reported sending
fewer texts than their 47 typically developing peers, and were also less likely to
respond to a text message sent to them by the experimenter, especially if they
had poorer reading ability for their group. The adolescents with SLI who did
respond to the text message wrote shorter messages – and used fewer textisms –
than their typically developing peers. As Durkin et al. (2011) point out, children
and adolescents with SLI are at particular risk of social marginalisation (Durkin
& Conti-Ramsden, 2007), and thus less likely than their peers to take up new
digital methods of communication (Bryant, Sanders-Jackson & Smallwood,
2006). These results suggest that young people with language impairment may
benefit both linguistically and socially from support in becoming more fluent in
producing and reading text messages with their peers.

The story so far

The empirical work reviewed above can be summarised as follows:

x There is some evidence to suggest that if adults are exposed to misspelled

forms of words this can impact negatively on their own spelling of those
words.
x There is no evidence to suggest that if children are exposed to misspelled
forms of words this affects the accuracy of their spelling of those words in
future.
x When children and adults are exposed to correctly-spelled words, their
spelling performance for those words is enhanced.
x When children and adults are asked to create their own spellings for
pseudo-words and these spellings are later revealed to be incorrect, this
does not affect their ability to learn the correct spellings.
x Children who demonstrate the greatest knowledge of text abbreviations
(textisms) also demonstrate better knowledge of conventional spellings.
x Children who tend to use the most textisms when asked to write a text
message also tend to have the best reading ability.
x Textism use appears to be contributing something positive to reading
performance above and beyond factors such as memory, vocabulary and
phonological skills.
x Children with developmental dyslexia do not seem to use phonetically-
based textisms to the same extent as non-dyslexic children, and tend to
use more of the non-phonetic forms.
 Children’s spelling, reading, texting 33

However, the studies reviewed which have examined textism use specifically
are not without limitations, and caution needs to be exercised at this point in
the book. First, many of the studies were exploratory in nature, meaning that
they were limited either in sample size or in the scale or nature of the assess-
ments used. Moreover, with the exception of Veater et al. (2011), these studies
are characterised by the use of contrived tasks which aimed to assess textism
use or knowledge without accessing samples of text messages actually sent by
the children. In fact, some of the studies involved giving the texting tasks to
children who either did not text normally or did not have access to a mobile
phone. Finally, the most serious limitation of these studies was that the data
collected were concurrent: that is, the data were collected at the same point in
time, thereby presenting a snapshot of the children’s ability in literacy, rather
than presenting a picture of the children’s literacy development. That is not to
say that concurrent data are irrelevant; more that care needs to be taken in the
interpretation of findings from such work, especially those reporting correla-
tions and regression analyses. This is because it is impossible from such data to
be confident about the direction of any observed associations. Take, for exam-
ple, the results of Plester et al. (2009) on pre-teens’ elicited text messages. In
this study it is observed that textism use can account for a significant amount
of unique variance in reading ability after taking a number of control variables
into account. However, it is impossible to know whether it is use of textisms
which is contributing to reading ability, or being a good reader which results
in the tendency to use textisms. Knowing the direction of these associations is
essential for making sense of the data summarised so far. To determine this,
there are two types of study which need to be conducted: longitudinal studies,
which track the progress of learners over time; and intervention studies, which
also track the progress of learners over time, but where contrasts are built into
the study to enable us to compare the impact of one educational intervention
or treatment over that of another. Work which has done exactly this is reviewed
in the next chapter.
Chapter 4

Does mobile phone use facilitate

literacy development?

As summarised at the end of the last chapter, the empirical work conducted
into both exposure to misspellings and use of textisms in text-messaging-based
tasks suggests that there is no reason to be concerned about children’s use of
text abbreviations and the impact that it may be having on children’s devel-
oping literacy abilities. In fact, if the work is suggestive of anything, it is that
the associations are positive rather than negative. However, despite the pat-
tern of results from these studies being consistent in their message, the studies
themselves are limited in what they can tell us about the actual development
of children’s literacy abilities. That is, the literature reported so far has relied
on concurrent data; data collected from one time point, thereby offering the
researcher a snapshot of children’s literacy skills and text-messaging perfor-
mance. However, just because there is a relationship between two variables
at one point in time, it does not necessarily follow that there is a relationship
between textism use and the emergence or development of literacy skills. Also,
correlations between variables collected concurrently do not inform discussions
of cause and effect. There are, however, two studies (one longitudinal study
and one intervention study) which have been published which do offer insight
into the development of written language skills in relation to children’s actual
use of text-message abbreviations and enable us to examine the issue of causal-
ity. These are reviewed in the sections that follow.

A longitudinal analysis of textism use

Mindful of the methodological limitations of previous concurrent work into
textism use and literacy, Wood, Meachem et al. (2011) conducted a longi-
tudinal assessment of 119 school children aged between eight and twelve
years. In this study (which was funded by the British Academy) the children
were assessed on their general verbal abilities (Verbal IQ), their reading,
spelling, phonological awareness and speed of phonological processing.
They were also required to provide a sample of the text messages that they
had actually sent over a two-day period. The children were asked to copy
these messages out exactly as they had written them, complete with any
 Mobile phones and literacy development 35

abbreviations or spelling mistakes, and parents or members of the research

team were asked to verify the accuracy of these transcriptions wherever pos-
sible. These messages were coded to identify the total number of textisms
used (as well as what types), and this figure was divided by the total number
of words used in the messages to provide a score which indicated the chil-
dren’s tendency to use textisms. A score of 1 would indicate a child who
had written all his or her messages entirely in text abbreviations, whereas a
score of zero indicated a child who had written his or her messages using
only conventional English.
These assessments were conducted at the beginning of one academic year,
and then they were repeated, with the exception of the Verbal IQ measure,
at the end of the academic year. These data were then used to assess the
extent to which textism use at the beginning of the year could predict growth
in literacy skills across the academic year, and whether literacy skills at the
beginning of the year could predict growth in textism use. This would finally
address the issue of the direction of association, and provide evidence as to
whether or not textism use may be contributing to improvement in literacy
skills over time.
As with other studies in this area, the UK children in this study reported
that they mainly used their phones for texting (56.3 per cent) rather than
calls (18.5 per cent), and there was evidence that textism use peaked at the
end of primary school, where textisms represented on average 43.9 per cent
of the words in the children’s text messages at the beginning of the year, and
this rose to 49.2 per cent by the end of the year for the children in Year 6.
In contrast, the Year 7 children’s tendency to use textisms across the year fell
from 42.0 per cent to 32.8 per cent. We might speculate that the reason for
this drop is that once the children enter secondary school (Year 7 is typically
the start of secondary school education) textism use is no longer perceived as
cool, and too much use of it was even seen by some children in this age group
as a little bit ‘naff’.
Textism use at the beginning of the year was significantly positively corre-
lated with reading, spelling, phonological awareness and speed of phonological
processing at the end of the year as anticipated (these associations were in the
order of 0.261 through to 0.329). There was also evidence of significant cor-
relations between spelling, phonological awareness and speed of phonological
processing at the beginning of the year, and textism use at the end of the year
(ranging from 0.274 to 0.331).
Although promising, these longitudinal correlations did not accurately
capture growth in either digital or traditional literacy. To do this, regres-
sion analyses were run, in which reading and spelling at the end of the year
were predicted by textism use at the beginning of the year, after controlling
for individual differences in the children’s initial reading or spelling ability,
as well as other factors such as age, Verbal IQ and phonological awareness.
Once this highly conservative analysis was conducted, it was found that,
36 Mobile phones and literacy development

although textism use could not account for growth in reading ability, it was
still able to account for a small but significant amount of growth in the chil-
dren’s spelling ability. However, when the regression analyses were reversed,
there was no evidence that either reading or spelling ability at the beginning
of the year could account for growth in textism use. This suggests that tex-
tism use is contributing positively to growth in the children’s spelling ability
independently of any contribution it may also be making via the development
of phonological awareness, and that this relationship is unidirectional rather
than reciprocal. Further analysis of the data suggested that this may, however,
be explained by the contribution that textism use may be making to enhanced
speed of phonological processing. That is, the act of using/creating textisms
repeatedly seems not only to place demands on the children’s phonological
awareness, but also seemed to enhance the speed of access to phonological
representations in memory.
These results were as expected in one sense, as it was anticipated that the way
in which textism use might contribute to literacy would be via phonological
awareness and phonological processing skills, and the analysis suggests that this
was in fact the case. What was more unexpected was the lack of any reciprocal
relationship between literacy ability and textism use. The data seem to suggest
that textism use contributes phonological abilities which in turn contribute
to literacy, but literacy skills do not impact on growth in the tendency to use
abbreviations.

Using mobile phones as a technological intervention

Given the results of the longitudinal study of Wood, Meachem et al. (2011),
there was an outstanding question which needed to be addressed. That is,
given the evidence that text messaging does not seem to be ‘bad’ for chil-
dren, and that textism use when texting may even help them to develop their
phonological awareness and phonological processing abilities, is there merit
in the idea of using mobile phones as a form of educational technology, either
in or outside of the classroom? In particular, should parents purchase mobile
phones for their children with the hope that these will help the children learn
to read and write? To answer this question, an intervention study was needed,
since the children in the longitudinal study were all children who already
owned and used mobile phones. The study therefore did not tell us anything
about the extent to which texting could impact on and add value to the read-
ing and spelling development of children who do not normally have access
to this technology.
To address this, Wood, Jackson, Hart, Plester and Wilde (2011) con-
ducted a Becta-funded randomised control study, in which 114 nine- and
ten-year-old children who did not already own a mobile phone were assessed
on their IQ, reading, spelling and phonological skills. Half of the children
at each school site (to control for school and teacher effects) were then
 Mobile phones and literacy development 37

randomly allocated to the mobile phone treatment group, who received

basic mobile phones (Nokia 1112 handsets) which were topped up with
credit (which enabled them to send text messages only) to use each weekend
for 10 weeks. The phones were distributed every Friday at the end of the
school day and collected back in first thing in the morning on the Monday
(so that they did not have access to the phones at school during the week –
an important ethical consideration at the time, considering the prevailing
view of mobile phones at the time this study was conducted). The children
were also allowed to use the phones throughout the full duration of the half-
term break (one week). The data from the children’s phones – including
samples of the children’s text messages and details of how many messages
the children sent and received – was hand transcribed from the phones each
week by the research team. Once all the data had been recorded from each
phone, it was reset and the unit charged and credit topped up, ready for the
next week of use.
The remaining children did not have access to the technology during the
study but received the same amount of contact with the research team and were
given access to the mobile phones at the end of the intervention period. All the
children (intervention and control group) were assessed on their reading and
spelling on a weekly basis as an ethical requirement, in case contact with the
technology was found to impact dramatically and negatively on their literacy
development. If this had been found to be the case, the study would have been
abandoned, but happily this was not necessary. At the end of the intervention
period the children were all reassessed on the literacy and phonological aware-
ness measures.
With respect to the intervention aspect of the research, the results indicated
that, although the children in the mobile phone group typically demonstrated
the greatest gains in reading, spelling and phonological awareness over the
10 weeks, these gains were not significantly greater than those observed in
the control group. It seems likely that the 10-week period of the intervention
was too short, and the limits placed on contact with the phones during that
period too restrictive to demonstrate statistically-significant benefits for these
children. However, it should be noted that, as with the previous studies in this
area, there was no evidence in this study of any detrimental impact of mobile
phone use on the children’s literacy. In fact we would anticipate that, had the
study continued for longer or the children had been able to use the phones
during the week as well as at weekend, there may have been greater evidence
of positive impact.
This study also necessarily included a longitudinal element to its design,
and this enabled the researchers to examine the literacy development of the
56 children who were given phones in the study, in a way similar to that pre-
viously conducted by Wood, Meachem et al. (2011). Moreover, this study
also included comprehensive data on the levels of text messages sent and
received week-by-week in a way not previously reported in the literature;
38 Mobile phones and literacy development

previous studies tended to rely on self-report measures which were based

on estimated rather than actual levels of use. This enabled the researchers to
examine directly the ‘exposure to print’ explanation of why texting may be
contributing to literacy development over and above that of phonological
awareness.
With respect to the volume of text message ‘traffic’ on the phones, there
was very little in the way of significant associations with the various literacy
measures. In fact, the only significant associations found were between the
number of messages sent and received by the children in Week 10 and per-
formance on the measures of phonological fluency. The phonological fluency
tasks required the children to generate as many different words as they could
think of in 30 seconds which corresponded to a given rule, such as ‘words that
rhyme with “whip”’ or started with a given letter. Taken together, these tasks
can be seen as measures of phonological access and retrieval. So the most that
can be said is that the more the children sent and received texts, the better they
were at generating or locating words in memory, which is logical and consist-
ent with the results found by Wood, Meachem et al. (2011). Moreover, this
relationship emerged over the course of the study – it was not apparent in the
traffic data obtained from the phones at either Week 1 or Week 5. But what it
does show is that the additional exposure to print specifically afforded by the
children’s use of the phones was not linked at any point in the study to either
their reading or spelling outcomes. This would suggest that either the level
of print exposure that the children received was not sufficient to benefit their
developing reading and spelling (again perhaps because of the restricted access
to the phones which the children had), or that for exposure to print to benefit
the children’s reading and spelling, the print needs to be of a more extended
prose style (or contain more conventional English spellings) than is typically
found in text messages.
With respect to analysing the mobile phone group’s data longitudinally,
once again a very conservative analysis was applied. That is, as in the Wood,
Meachem et al. (2011) study, growth in literacy skills was assessed after con-
trolling for individual differences in IQ and the children’s reading or spelling
ability at the start of the study. Given the short duration of the intervention
period (10 weeks), it seemed unlikely that any variable would be able to pre-
dict growth in literacy across this time period over and above these control
variables as, for example, reading at Time 1 tends to be so highly related to
reading at Time 2 if the gap between the two points is very short, that other
variables are able to explain very little additional variance. However, as in the
previous study, Wood, Jackson et al. found that average levels of textism use
by the children over the 10 week period could explain an additional 8.6 per
cent of variance in the children’s spelling scores, which is a statistically sig-
nificant amount. In short, for the children who had access to the technology,
there was once again evidence that textism use when texting was enhancing
the children’s ability to spell over time.
 Mobile phones and literacy development 39

The million dollar question: should we buy our

children mobile phones?
So, the literature reviewed in this chapter emphasises the same central mes-
sage. That is, textism use by children when sending text messages is unlikely
to be harming their literacy skills, and is in fact likely to enhance their spell-
ing ability via the rehearsal of phonological skills. Given these results, is the
evidence strong enough to recommend that children should routinely be
given access to mobile phones for text messaging where they do not already
have them? Well, the answer to that question based on the data available at
this time is likely to be ‘no’. The level of educational benefit observed in the
intervention study was not really strong enough to suggest that the purchase
of a mobile phone with its associated costs and other considerations is mer-
ited. This is especially true if we take into account the fact that most children
will also have the opportunity to practise textspeak through other technolo-
gies, such as internet messaging systems and social networking media. On the
other hand, we would emphasise the point that if your child already owns a
mobile phone, the evidence also suggests that there is no reason to be con-
cerned, from an educational point of view at least, if he or she is using it to
send text messages.

Persistent negative perceptions

So, given the consistent empirical evidence reported for school-age children,
there appears to be no basis for suggesting a negative impact of textism use on
their developing literacy skills. In fact, if anything, the evidence would sug-
gest that for children who already own mobile phones, textism use appears
to be supporting the development of spelling skills rather than undermining
them. Given these findings, why is it that the popular assumption was that the
relationships must be negative, and why do these assumptions exist in spite of
the consistent evidence in this area?
One possible reason for this is likely to be illusory correlation (Chapman &
Chapman, 1969). That is, when two independent but highly salient events or
stimuli occur around the same time, an assumption is often made that there
must be a link between the two events, even when this is not the case. In the
case of text messaging, children’s creative use of simplified or alternative spell-
ings is often the first thing that people think of when you mention texting to
them. During the time when text abbreviation use was perceived to establish
itself as common activity amongst children, reports of declining literacy in the
United Kingdom were also common in news media (see e.g. Paton, 2007). As
a result of the relative salience of these two events, it seems likely that a false
connection may have been made between them. Once formed, these types of
stereotypical beliefs tend to persist and affect the processing of new information
on the topic, such that the original belief is maintained rather than challenged
(see Peeters, 1983 for a review).
40 Mobile phones and literacy development

What, then, can we do to change these false perceptions? In order to

change public perceptions of stereotypical views formed via illusory correla-
tion, Peeters (1983) suggests that the salience or distinctiveness of the two
phenomena need to be changed in a way that would then enable a new,
positive association to be formed. To achieve this for texting, the media rep-
resentations of both the literacy levels of young people and their use of new
technologies need to be revisited, and debates around them reconceptualised
and actively promoted. In particular, Peeters argues that the presentation of
information that is highly discrepant with views that are held by individuals
may have the greatest impact in challenging individuals who hold stereotypic
views. However, this may be easier said than done, as the consistency of the
positive message on texting and literacy with children is perhaps undermined
by empirical work which has been conducted with older populations, as we
shall see next.
Chapter 5

Texting and literacy skills in

adolescents and young adults

As seen in the previous chapters, there is by now clear evidence that children’s
fluency with creating and deciphering textisms is generally associated with better –
not poorer – reading and spelling skills. However, it is less clear whether such
relationships also hold in older users of text-messaging technology. There
are many differences in the experience of texting for children compared
to adults. Today’s children are growing up learning to read and write text
messages soon after – or even during – the period in which they are devel-
oping the skills to read and write conventionally. Children’s knowledge of
sounds, spellings and word structure continues to develop into adolescence,
and so there is scope for texting to influence standard literacy, or vice versa,
throughout the school years. However, most adults of today have largely
consolidated their conventional literacy skills before being introduced to text
messaging, which has meant that the two writing systems were learned at
different times, and in quite different ways. This chapter considers the use of
textisms in adolescents and young adults, and the potential links with literacy
in this older population.

Adolescents’ and young adults’ use of texting

and textisms
Since the inception of text-messaging technology, older teenagers and adults
have been particularly early and eager users of the rapid, private communica-
tion that it allows. Initially, it was not clear whether this intense use of the
technology represented a cohort effect (that would continue on as a char-
acteristic of this ‘pioneering’ group of texters), or a life phase effect (that
would remain as a hallmark of each new group of young adults, gradually
diminishing as each group entered middle adulthood). To answer this ques-
tion, Ling (2010) examined six years of data on the text messages sent by
Norwegians over the age of 13 years. His results confirmed that the pattern
represents a life phase phenomenon: older teenagers and those in their early
twenties go through a period of intense use of texting, before ‘growing out
of it’ as their lifestyles change. It is likely that this pattern of Norwegian
42 Young adults’ texting and literacy skills

text messaging is repeated in other developed nations, and is borne out by

the pattern observed in the longitudinal study conducted in the United
Kingdom by Wood, Meachem et al. (2011), who saw textism use peak in Year
6 for two successive waves of children tested in Years 4 to 7.
As well as being enthusiastic writers of text messages, teens and young
adults are keen users of textisms, although not to the overwhelming extent
portrayed in the media. Surveys of young people’s textism use have revealed
that the steady but relatively constrained use of text-style abbreviations and
re-spellings is a feature of these messages, across countries and across chang-
ing technologies. As noted in Chapter 2, in the relatively early days of text
messaging, Thurlow and Brown (2003) invited 135 British undergraduates
(75 per cent of whom were female) to transcribe five text messages which
they had recently sent or received. The average length of these messages was
about 14 words, and overall they contained about 19 per cent abbreviations,
plus another 8–9 per cent textisms that included symbols, letter/number
homophones and emoticons. In a smaller but more intensive study, Grinter
and Eldridge (2001, 2003) asked ten 15- to 16-year-old British adolescents
to keep a log of all the texts they sent and received during one week (477
messages in all). These messages were on average about 70 characters long
and contained a variety of textisms, including 32 per cent abbreviated spell-
ings and 37 per cent spellings where the letter/number/symbol’s sound was
pronounced (e.g. gr8 for great). In a similar logging study, Ling and Baron
(2007) examined all (191) text messages sent in the past 24 hours by 22
female American undergraduates. These messages were shorter than those
of Thurlow and Brown’s sample (less than 8 words on average), and con-
tained fewer examples of textisms: about 5 per cent abbreviated spellings
and only a handful of acronyms or emoticons. Ling and Baron speculated
that these differences may stem from British users’ longer experience with
composing messages on the alphanumeric phone keypads of the time, and/or
American users’ greater practice with communicating instead via Instant Mes-
saging on a full computer keyboard (a medium in which they used virtually
no abbreviations).
More recently, Grace, Kemp, Martin and Parrila (2012) invited 86 Australian
and 150 Canadian undergraduates to provide the last five text messages that
they had sent. Both groups were more experienced users of text messaging
than those in earlier studies, having owned a phone for six to seven years on
average, and sending an average 20 to 40 messages a day. Phone technology
had also changed: only about half the Australians and a quarter of the Cana-
dians still had alphanumeric keypads, and the rest now used full QWERTY
keyboards. Participants’ messages were similar to those reported by Thurlow
and Brown: about 12 words long on average, with textisms making up about
19 per cent of the messages written by Australian students and 16 per cent
of those written by Canadians. Most students pressed each key once for each
letter (rather than using the older multi-press entry system), but the usual
 Young adults’ texting and literacy skills 43

use of predictive entry was more common for Australian students (55 per
cent) than Canadians (34 per cent). In a similar study, Drouin and Driver
(2012) collected five recently sent text messages from 183 American under-
graduates. These students wrote about 60 messages a day, and 62 per cent
used a full QWERTY keyboard. Their textism density was relatively high, at
28 per cent of the words in their messages. This cannot be attributed simply
to their entry system, as the percentages were close to those of the Australians
described above; specifically, 59 per cent of these American students used
predictive entry most of the time. Drouin and Driver’s participants had been
texting for fewer years (about four, on average) than Grace et al.’s partici-
pants, and it is possible that the novelty or enjoyment of using textisms had
not yet worn off to the same extent.

Textisms in languages other than English

Although our focus is on writers of text messages in English, it should be
noted that young adults who write messages in other languages also use a
variety of textisms. As noted in Chapter 2, researchers have described the
use of textisms in a range of languages, from Finnish, to Chinese, to Thai.
Rather more research has been published on the use of textisms in European
languages, although much of this is more qualitative than quantitative. For
example, Alonso and Perea (2008) describe of a range of textism types used
in Spanish, similar to those reported in English messages. These authors note
examples of vowel omission and phonetic re-spelling (e.g. ksa for casa, house),
letter/number homophones (e.g. to2 for todos, everyone/all), the excessive
use of punctuation for effect (e.g. no!!!!!) and the use of emoticons.
Anis (2007) analysed a corpus of 750 messages written and sent by French
teenagers and young adults, and listed a range of linguistic changes made
to words, although no numbers are given. Textisms in this corpus included
phonetic respellings (e.g. k for que, that), vowel omissions (e.g. vs for vous,
you), letter/number homophones (e.g. + for plus, more), and truncations
(e.g. la dèf for La Défense). A set of 800 text messages written in Italian to an
interactive television programme was analysed by Herring and Zelenkauskaite
(2009). They observed textisms that included (what they called) clippings
(e.g. cmq for comunque, anyway), letter/number homophones (e.g. c for ci,
us), phonetic spellings (e.g. perke for perché, why) and initialisms (e.g. TVB
for ti voglio bene, I love you).
Bieswanger (2007) compared a set of English and a set of German text
messages, each containing several hundred messages, written by young
adults. Both sets of text messages contained various textism types, including
initialisms (e.g. HDL for hab dich lieb, love you), clippings (e.g. Antw for
Antwort, answer), contractions (e.g. hab’s for hab es, have it) and phonetic
spellings (e.g. leida for leider, unfortunately), although no letter/number
homophones were observed in the German messages. Most strikingly,
44 Young adults’ texting and literacy skills

although the average length of the two sets of messages was similar (around
93 characters), the English texts contained six times as many textisms
(about 5.6 textisms per message), as the German texts (only about 0.86 per
message). Thus, the patterns of textism use might vary quite a lot between
languages, largely because of the different affordances of individual lan-
guages. For example, the large number of homophones in English means
that letter/number homophone textisms (e.g. c for see, 4 for for) are much
more common than, for instance, in German, and the frequent abbrevia-
tions of spoken Finnish lead to a large proportion of similarly abbreviated
written forms in texting (as noted in Chapter 2).

Relationships between textism use and

literacy skill: self-report
It is clear that the use of textisms remains an important part of the text mes-
sage writing of adolescents and young adults. However, as noted earlier, a
crucial question is whether this use is linked to young people’s more tra-
ditional literacy skills. Some researchers have considered this question by
investigating young adults’ self-reported use of texting and textisms. In
one of the earliest studies on this question, Massengill Shaw, Carlson and
Waxman (2007) tested whether more frequent text messaging (assumed to
mean more frequent exposure to textisms) was associated with poorer spell-
ing proficiency. However, they saw no significant association between the
self-reported text-message-sending frequency of 86 American undergradu-
ates and either their estimated or actual spelling skill. More recently, Drouin
(2011) found contrasting results with 152 American undergraduates: the fre-
quency with which students reported sending text messages turned out to
be positively related to their scores on tests of spelling and reading fluency,
although not reading accuracy. The rapid growth of texting in the United
States during the years between these two studies might help to explain the
discrepancy in these results. Of course, the number of text messages that one
typically sends is at most a very indirect measure of exposure to textisms.
Thus, Drouin also looked at participants’ self-reported use of textisms and
their literacy scores, but found no significant links between these measures.
In a similar study, Drouin and Davis (2009) compared the literacy skills of
34 American undergraduates who said that they used textisms in their text
messages, and 46 who said that they wrote their messages in standard English.
The two groups were found to have virtually identical scores on measures of
spelling, word reading and reading fluency.
One further study on the self-reported use of texting and textisms was by
Rosen, Chang, Erwin, Carrier and Cheever (2010), who collected data in
an online survey from over 700 young people in the United States aged 18
to 25 years. Participants answered questions about their textism use (rating
their use of several textism types on a five-point scale from ‘never’ to ‘very
 Young adults’ texting and literacy skills 45

often’), and also produced short formal and informal writing samples. Those
who reported sending more messages produced better informal writing
samples, although for those with some university education, more message-
sending was related to poorer formal writing. Rosen et al. observed some
negative correlations between self-reported textism use and formal writing
quality, and some positive correlations between textism use and informal
writing quality. However, these relationships were not straightforward: they
varied with textism type and participants’ level of education. For example,
for respondents with no tertiary education, the more they reported missing
apostrophes in their text messages, the better their informal writing, whereas
for respondents with some tertiary education, the more they reported using
i for I, the worse their informal writing. In sum, the self-report studies
reviewed here provide no clear answer about whether texting frequently,
or using many textisms, is associated with better or poorer literacy skills in
young adults.

Relationships between textism use and

literacy skill: experiments
Other researchers have compared participants’ literacy skills with their use of
textisms in an experimental setting. The relative advantages and disadvan-
tages of both self-report and experimental studies are discussed in Chapter 7,
but experimental studies of textism use can offer further information on the
links (if any) between textism use and literacy skill. Kemp (2010) asked 61
Australian undergraduates to read aloud, and to compose (to dictation) text
messages in both standard English and textspeak. Kemp also gave participants
several language and literacy tasks, testing their general spelling and reading,
and their awareness of words’ phonology (sound structure) and morphology
(meaning structure). As discussed in Chapters 3 and 4, being good at analys-
ing words’ sounds may assist with the creation and deciphering of the many
textism spellings based on sound, such as letter or number sounds (u for you,
2 for to) or sound-based re-spellings (e.g. skool for school). Similarly, being
adept at analysing meaning structure (morphology) might increase one’s
facility with reading and writing textisms based on units of meaning, such as
bf for boyfriend, or sum1 for someone.
Kemp found no significant differences on any of the language task
scores between students who sent text messages more frequently (an aver-
age of more than five messages per day) and less frequently (five or fewer).
However, when it came to creating and deciphering textisms, proficiency
was positively related to literacy skill, even after controlling for the usual
number of messages sent per day. Being quicker at composing both textese
and standard English messages was positively related to standardised spell-
ing and reading scores. Furthermore, making fewer errors when reading
aloud both textspeak and standard English messages was positively related
46 Young adults’ texting and literacy skills

with standardised reading score. Finally, higher morphological awareness

scores were positively correlated with fewer errors in deciphering textspeak
messages.
At first glance, these generally positive links contrast with those from the
self-report studies discussed above. However, the rather artificial situation
created by the careful experimental manipulation may have meant that gener-
alisations to real-world texting behaviour cannot easily be made. To keep the
task equal for everyone, all participants entered their messages using one par-
ticular phone type, with the alphanumeric keypad of the time (circa 2008).
Messages were composed using the multi-press entry system to allow unlimited
scope for textism use. In this sample, however, less than half the students
reported regularly using multi-press entry, and the rest usually used a more
modern predictive entry system. Thus, students found themselves creating
and deciphering text messages under time pressure, on an unfamiliar phone,
via an entry system that not all of them still used regularly. If the task seemed
more like a timed language test than a reflection of normal texting behaviour,
this may explain why we observed positive correlations between texting task
performance and literacy skills.
In an attempt to create a more realistic situation, De Jonge and Kemp (2012)
asked 52 teenagers (13–15 years) and 53 young adults (18–24 years) to translate
a set of written messages into ‘how they would text them to a friend’ on their
own phones with their usual, predictive entry system. Participants also com-
pleted tasks of spelling, real and non-word reading, morphological awareness,
and orthographic awareness (sensitivity to the frequency of certain letters and
letter combinations in a given writing system). The two age groups performed
so similarly that their responses were combined for the analyses. In contrast to
previous findings, self-reported text-message-sending frequency was negatively
related to scores for spelling, real and non-word reading, and morphological
awareness. Also, in contrast to previous findings, participants’ use of textisms
in the dictation tasks was negatively related to these same language-task scores.
The more often these young people wrote text messages, and the more textisms
they used, the poorer were their language skills. In case the link between textism
use and literacy could be explained mainly in terms of frequency of message-
sending, De Jonge and Kemp (2012) controlled for this measure statistically,
and found that only the negative correlation between textism use and non-word
reading remained significant. This more realistic experimental study, like the
self-reported studies, thus led to another rather mixed set of conclusions about
the links between textism use and literacy skills.

Relationships between textism use and

literacy skill: naturalistic studies
Only recently have researchers begun to consider young adults’ naturalistic
use of textisms. Although it is quite time-consuming and intrusive to collect
 Young adults’ texting and literacy skills 47

participants’ actual sent messages compared to asking for their self-reports, or to

giving experimental messaging tasks, the examination of naturalistic text mes-
sages gives a more accurate reflection of people’s real-life habits. Grace, Kemp,
Martin and Parrila (submitted, b) asked 150 Canadian and 86 Australian first-
year university students to provide their last five sent text messages, as well as
asking them to complete a series of language and literacy tasks. As noted earlier,
the Canadian students produced messages with a textism density of about
16 per cent, and reported sending about 40 messages per day. They completed
a non-word reading task and a standardised spelling task. The non-word reading
task was presented in its standard, untimed version, but also in a more sensitive
timed version. Contrary to previous findings, there were no significant correla-
tions between texting frequency and literacy measures, but there was a negative
correlation between textism density and spelling scores. That is, Canadian stu-
dents who used more textisms in their naturalistic messages had poorer general
spelling than those who used fewer textisms. The Australian students in this
same study used an average of 19 per cent textisms in their text messages, and
reported sending an average of about 24 messages a day. They completed the
same literacy tasks as the Canadians, plus two more tasks to look at their abil-
ity to process phonological (sound-based) information: an additional non-word
reading task, and a sound manipulation task. Finally, the Australian students
completed a questionnaire about any difficulties they had experienced with
learning to read at school. This group’s patterns of correlation were different:
the number of messages sent daily correlated negatively with scores on the addi-
tional untimed non-word reading task. Furthermore, Australian students who
used more textisms were poorer at reading unfamiliar words under time pressure
and at manipulating words’ sounds, and were more likely to have had difficul-
ties learning to read at school. Some of the differences between the two samples
may stem from differences in the use of technology (as discussed further below).
However, this study, like previous ones, confirms that any relationship between
textism use and literacy and language skills in adults is a rather mixed one.
Drouin and Driver (2012) also asked their participants (183 American
undergraduates) to provide five of their recently sent text messages. Par-
ticipants’ overall textism density (28 per cent, as mentioned earlier) did
not relate significantly to their word reading, reading fluency, spelling or
vocabulary scores. However, Drouin and Driver wanted to explore the
idea that some textisms are negative ones: ‘lazy’ omissions or transgres-
sions against standard English (such as the omission of apostrophes or
capitals), whereas other textisms can be seen as more ‘positive’, or creative,
such as accent stylisations (e.g. wiv for with) and the insertion of symbols
(sorry / ). This might lead to a different pattern of links with literacy,
depending on the types of textism being considered. However, this kind
of negative/positive textism distinction is difficult to pursue among those
who use predictive texting: this entry system often corrects missing apos-
trophes and capitals, and thus the appearance of any ‘negative’ textisms
48 Young adults’ texting and literacy skills

might be limited more by the sophistication of the technology than its

users. In fact, in the predictive-entry-only students, Drouin and Driver saw
only one significant correlation: a positive one between the use of accent
stylisation and spelling skill, similar to the findings of Plester et al. (2008)
with children. In students who never used predictive entry, Drouin and
Driver observed that reading fluency was linked positively to the use of
initialisms but negatively to the use of letter/number homophones, while
word reading was linked negatively to the omission of apostrophes. How-
ever, caution should be exercised in drawing strong conclusions from these
few significant correlations, especially in the context of the relatively large
number that were calculated.

Limitations
Most of the data reported here have come from young people at university,
whose literacy skills are in the average to above-average range. This may have
restricted the possibility of seeing links with literacy task scores. There are
some small cues that when literacy skills are closer to average than above aver-
age the links with textism use are more likely to be negative (De Jonge &
Kemp, 2012; Grace et al., submitted, b), but this cannot be the whole story.
Future researchers would do well to follow Rosen et al., (2010) and collect
data from participants with a wider range of educational experience (prefer-
ably naturalistic data), in order to learn more about links between textism use
and more varied literacy skills.
The variety of language and literacy tasks used in the tasks discussed here
may also explain some of the differences observed in the relationships (or
lack thereof) seen with textism use. If the links with textism use are not
strong ones, they may be affected by the nature and sensitivity of the lit-
eracy tasks chosen, from more familiar tests of reading and spelling to more
unusual or challenging tests of reading non-words or thinking about word
structure. The nature of the textisms themselves is also an important con-
sideration. Although most authors have differentiated between the types of
textisms produced (contractions, emoticons, omitted capitals, etc.), textisms
have usually been lumped together when correlations are calculated with
literacy skills. One exception is the study by Rosen et al. (2010), although
it is difficult to draw clear interpretations when there were so many different
patterns of correlation across education and writing formality levels.
As noted above, Drouin and Driver (2012) also suggest that the rela-
tionships we see may depend on the nature of the textisms being examined.
Further support for this idea can be sought by running such finer-grained
analyses of existing textism data. We calculated correlations between indi-
vidual textism types and literacy task scores for the data from participants
in De Jonge and Kemp’s (2012) experimental study, who were all users of
predictive text. The clearest findings were that the use of letter/number
 Young adults’ texting and literacy skills 49

homophones (such as c for see and 2 for to) was negatively related to spelling,
real and non-word reading, and morphological awareness, and that the use
of accent stylisations (such as tomoz for tomorrow) was positively related to
spelling and real word reading. The other, less easily explained findings were
that real word reading was related positively to omitted capitals, but nega-
tively to the use of initialisms. A similar fine-grained analysis was run on the
naturalistic texting data and literacy scores from Grace et al. (submitted, b).
Following Drouin and Driver (2012), these texters were split into users and
non-users of predictive text entry. However, the patterns of significance
were again so mixed that it is impossible to draw meaning from them. For
example, in predictive texters there were positive links between emoticon
use and spelling scores, but negative links between omitted apostrophes and
non-word reading. In contrast, non-predictive texters showed positive links
between omitted apostrophes and word reading, but negative links between
emoticon use and non-word reading. Drouin and Driver make a good point
in exploring the different patterns of correlations between literacy scores and
different types of textism. However, the results across three studies are far
from clear, and no strong conclusions can be drawn at this stage about how
the relationship between literacy and textism use might differ with the type of
textisms being considered.

Alternative explanations
If differences in young adults’ use of textisms are not obviously related to
differences in their literacy skills, we must consider what else could contrib-
ute to textism use. As alluded to above, some of the differences may come
from the technology being used. In the early days of text messaging, the
small screen, laborious multi-press entry system and the 160-character limit
all laid practical constraints on the length of messages, inviting the use of
abbreviations. There is evidence that the use of multi-press entry encourages
the use of more textisms than the use of predictive entry, at least in chil-
dren (Kemp & Bushnell, 2011). Further, we see more textism use in groups
with more users of alphanumeric keypads than QWERTY keyboards (Grace
et al., 2012). From the published research to date, it is not apparent that
the use of textisms is decreasing as technology becomes more sophisticated.
However, ongoing research with PhD student Abbie Grace suggests that this
trend might be more obvious when similar cohorts are compared across time.
Grace and Kemp have been collecting five recent sent text messages from suc-
cessive cohorts of first-year Australian university students for four years now,
and have observed a gradual reduction in the overall textism density of the
messages, from 27 per cent, to 19 per cent, to 18 per cent, to 15 per cent,
from 2009 to 2012. Of interest is that within the overall textism densities,
we are seeing changes in the types of textisms used, and this is perhaps a clue
to the effect of technology. The use of textisms that abbreviate words seems
50 Young adults’ texting and literacy skills

to be diminishing across time (from 12.8, to 6.0, to 5.6, to 3.8 per cent
of all textisms, across the four years), whereas the use of more expressive
textisms, such as emoticons, extra letters (pleeeease) and extra punctuation
(yes!!!) seems to be slightly increasing (from 3.6, to 4.3, to 5.7, to 5.7
per cent of all textisms, across the four years). This will be something to
observe as the message hardware (keyboard type) and software (entry sys-
tem) continue to evolve.
One factor that has been largely overlooked is that differences in tex-
tism use might depend heavily on young adults’ motivation to compose
text messages in ‘correct’ English (or other language), regardless of their
conventional literacy skills. There is no requirement that text messages fol-
low the conventional rules of writing and, despite the existence of various
glossaries of textspeak, there is no standard way that words should be abbre-
viated. Many texters may feel that being efficient (in terms of time or effort
taken to compose a message) is more important than the use of conventional
spelling and grammar, when the main aim is communication. The desire
for efficiency may cut across differences in conventional literacy skills, lead-
ing to the rather messy picture that we have seen about any links between
textism use and literacy scores. Of course, this efficiency is two-fold: there is
no point saving time writing a message so abbreviated that the reader can-
not decipher it (see Kemp, 2010). But as long as both sender and receiver
share an understanding of the types of textisms that might be used, it makes
sense that the use of textisms might depend as much on a person’s desire to
get their message across with the minimum amount of time and effort as on
their conventional literacy skills.
People’s desire to spell text messages ‘correctly’ is unlikely to be all-or-
none. Despite the concerns of the media, there is evidence that young adults
are generally aware of when it is appropriate to use textisms and when it is
not. This sense of register is seen in studies in which young people have been
asked what they think of the use of textisms in various types of messages.
Drouin and Davis (2009) found that while 75 per cent of their undergradu-
ate participants believed it to be appropriate to use textisms in a message to
a friend, only 6 per cent thought it appropriate to use them in a mes-
sage to a university instructor. There is social meaning in these opinions;
Lewandowski and Harrington (2006) report that undergraduates were more
likely to perceive the writers of formal emails to be lacking in skill or effort
if the writers included textisms in these emails than if they did not. Grace et al.
(submitted, a) found that undergraduates thought it more appropriate to
use textisms in contexts that were informal or to recipients who were socially
close (such as a friend) than in contexts that were formal or to more socially-
distant recipients (such as a lecturer).
In a more detailed study of the perceived appropriateness of textism use
in different types of message, Clayton (2012) found that university stu-
dents rated textism-laden messages as more appropriate for friends than
 Young adults’ texting and literacy skills 51

for peers, and as more appropriate for classroom peers than for lecturers.
Conversely, they rated messages written in standard English as more appro-
priate for lecturers than for peers, and more appropriate for peers than for
friends. These patterns were replicated in the messages that participants
were invited to compose. When asked to write to particular recipients about
particular scenarios, the students included significantly more textisms in the
messages that they composed for friends than for peers, and more textisms
in the messages that they composed for peers than for lecturers. It is clear
that university students understand that the formality of a message should
fit the intended recipient. However, the participants in these studies could
not help but notice that they were being asked to write or rate messages
to people of different levels of social closeness and formality (even though
the order was carefully counterbalanced). This may have led to exaggerated
differences in textism use or ratings across recipients, and it remains to be
seen whether these differences would remain if the writing of naturalistic
messages were examined. Many university lecturers would have received
student emails that contain a range of textisms and other examples of non-
standard writing. Such examples suggest that further research is needed into
the differentiation of textism use according to the intended recipient, not
only by university students, but by adolescents and children as well.
However, in the naturalistic texting studies reviewed above, it should be
noted that the researchers did not control or record the recipients of the five
recent messages sent by study participants. Although it is statistically likely
that recipients would mainly have been socially close, it is possible that some
may have been more distant, or that others may have been close but not
appropriate recipients of textism-peppered messages. (Anecdotally, many
students note that their parents and grandparents are unable to decipher tex-
tisms, and must be written to in conventional English to ensure that they
comprehend the meaning of the message.) Presumably, participants with a
good understanding of register would have varied their textism use according
to their audience, which would be another factor contributing to differences
in the extent and type of textisms seen in naturalistic studies. Again, this is an
area for further investigation.
A final reason for differences in young adults’ use of textisms concerns
the social aspect of this unconventional form of writing. Although the social
nature of textism use is not the focus of this book, its importance should not
be ignored. Texting itself has a strong social role: it helps to create and main-
tain social cohesion (e.g. Igarashi, Takai & Youhida, 2005; Ling, Bertel &
Sundsøy, 2012). Using textisms can be fun and can help to make messages to
friends more expressive (Grinter & Eldridge, 2001); it can also be a marker of
social identity, especially for teenagers (Jones & Schieffelin, 2009). As teen-
agers grow into young adults, they may begin to find the use of textisms to be
immature (Lewis & Fabos, 2005). The age and the importance of textism use
to the social groups of the participants in the studies reviewed in this chapter
52 Young adults’ texting and literacy skills

may therefore also have had important roles to play in the use of textisms,
beyond any differences in textism use.
In conclusion, it seems that even if there are some associations between
conventional literacy skills and the use of textisms by teenagers and young
adults, these links are far from clear. Any relationships seen in the studies
reviewed in this chapter seem to vary with a host of other factors as well,
some of which have been well-studied, and others which could be the
subject of future research. Links between textism use and literacy may be
observed when the literacy tasks are more challenging for the participants,
although this may also vary with the way that textisms are collected (via
self-report, or experimental or naturalistic means). The way that young
adults enter their text messages may also affect the type of textisms they
produce: older alphanumeric keyboards may encourage abbreviation, and
predictive entry may correct omitted apostrophes or capitals, regardless of
the sender’s literacy level. Considering the nature of the textisms them-
selves will also be useful in future research: deliberately inserting symbols
or extended spellings for emotional effect probably represents a very differ-
ent process from dashing out a message in which conventional punctuation
is omitted. Textisms of ‘omission’ could represent ignorance, carelessness
or deliberate efficiency, and any of these may be moderated by the mes-
sage’s recipient: a teenager might vary her use of textisms according to
whether a message is written to a friend with whom she enjoys the banter
of extensive textisms, or her grandmother who struggles to read anything
not written in conventional English. Further research is clearly needed on
the combination of factors that are associated with the use of textisms
in young people’s messages, but our overall conclusion is that variations in
textism use cannot be easily attributed to differences in levels of conventional
literacy skill.
Chapter 6

Understanding children’s mobile

phone behaviours in relation to
written language abilities

The previous chapters have demonstrated that children’s use of text abbre-
viations (textisms) when text messaging has been positively related to their
performance on a range of literacy measures (e.g. Plester, Wood & Joshi,
2009; see Chapter 3) and is particularly closely associated with spelling
performance over time (Wood, Meachem et al., 2011; Wood, Jackson, Hart,
Plester & Wilde, 2011; see Chapter 4). The picture for adolescents is, however,
more mixed (see Chapter 5). These studies prompt further questions about
how school-aged children in particular use mobile phone technology and how
those more general behaviours might also contribute to this overall pattern of
results. Specifically, it is possible that individual differences in the children’s use
of mobile devices and the nature of the phones themselves may be contributing
to the results observed in such studies. Moreover, there is frequent discussion
about the issue of children becoming addicted to mobile technology (e.g. Lewis,
2012), and in 2010 there were reports of the first ‘rehab’ clinic for children who
had been identified as ‘technology addicts’ being opened in the United King-
dom (Hough, 2010). However, there is very little data which has attempted
to characterise children’s actual reported use of mobile phones in a way that
would enable us to see whether children are showing evidence of overreliance
on or addiction to them. This chapter therefore presents previously unpublished
data from a British Academy funded study conducted by Wood1 with primary
and secondary school-aged UK children, which surveyed them on their mobile
phone behaviours, and then considered these responses in relation to data on
their performance on standardised assessments of spelling and related underly-
ing written language processing skills.

The research participants

In September 2010, a total of 201 children were recruited from schools
in the West Midlands of the United Kingdom: 106 primary school children
and 95 secondary school children. Primary and secondary school chil-
dren were recruited into this study to enable us to compare their responses,
in case there were distinctive differences in their mobile phone behaviours.
54 Phone behaviours and written language

The primary school children were aged between 8 years and 2 months
and 11 years and 1 month (average age 9;9), and comprised 51 boys and
55 girls. The secondary school children were aged between 11 years and
5 months and 14 years and 7 months (mean age 12;10) and comprised
47 boys and 48 girls. Overall the children had a verbal IQ score of 96.4
(standard deviation = 14.0), as measured by the Wechsler Abbreviated Scales
of Intelligence test (Wechsler, 1999), and this indicated that the children
were typically of average ability for their age, and the range of scores they
demonstrated was as we would expect for the population as a whole. So
we can say that the children in this study were reasonably representative
of other children of the same age developmentally, and there was an even
gender distribution overall and within each age group.
All the children in this study owned their own mobile phones. When
asked how they got their phone, 36.3 per cent reported that they had asked
for it, but 31.8 per cent said that it was an unexpected gift; in other words, in
almost a third of cases the children were given a phone by adults when they
had not explicitly requested one. So, contrary to some expectations, a large
number of young phone users appear to become so as a result of unsolicited
adult intervention. An additional 9.5 per cent of children were told that they
needed a mobile phone, and 19.9 per cent were given a second hand phone.
The average age of the children when they were given their first phone was
9.7 years, with the youngest reportedly being just two-years-old and the
oldest being 12 years. Twenty-two of the 201 children reported receiving
their first phone at age five years or younger. Although it seems likely that
the children who reported receiving their phones at age two or three may be
misremembering or guessing the age incorrectly, it does seem that for these
children they cannot remember a time when they did not have access to one,
and the trend towards giving children mobile phones at increasingly younger
ages is continuing.

The assessment of written language skills and

mobile phone behaviours
To assess the written language skills of the children, we asked them to
complete a spelling test (Single Word Spelling Test; Sacre & Masterson,
2000), and two tasks from the Phonological Assessment Battery (PhAB;
Frederickson, Frith & Reason, 1997). The first was a spoonerisms task,
which required the children to create spoonerisms from sounds and words
that were presented to them. This was used as a measure of phonological
awareness (awareness of and the ability to isolate and manipulate speech
sounds). As noted earlier in this book, phonological awareness is included
in our research into literacy and texting because it is a skill known to underpin
the ability to learn the alphabetic principle (that letters and letter combi-
nations represent individual sounds in speech; knowledge that is taught
 Phone behaviours and written language 55

as ‘phonics’ in schools), and it is also a required skill for the processing

of phonetically-based textisms. The second task taken from the PhAB was
a rapid automatised naming (RAN) task, which required the children to
look at two grids of 50 pictures (five line drawings of common objects,
repeated randomly) and name the objects as rapidly as they could. This task
was used as an assessment of rapid phonological access and retrieval. This
is a different skill from phonological awareness, as it is possible that a child
could have good representations of speech sounds and good awareness of
sound structures in speech, but be slow or otherwise impaired in his or her
ability to retrieve those representations or articulate them in a fluent man-
ner. Rapid automatised naming is also a task which is very highly correlated
with performance on literacy measures, and difficulties in completing it are
implicated in explanations of reading difficulties, including developmental
dyslexia (e.g. see Wolf & Bowers, 1999).
Orthographic processing (the processing of information about spell-
ing conventions in a specific written system, such as in English) was also
assessed, using a task which presented the children with sequences of words
which were printed together without spaces in between (e.g. carspoonleap)
and the children had to indicate where there should be spaces to separate
the words (Wordchains; Guron, 1999). This task was completed as a timed
task, and we also asked the children to repeat the syllable ‘la’ out loud
whilst doing it (articulatory suppression) to ensure that they were only able
to process the items orthographically, rather than both phonologically and
orthographically.
To assess the children’s mobile phone behaviours, a questionnaire was
developed which the children completed with the assistance of a member of
the research team, in case there were any issues with the children’s ability to
either read or understand the questions. The questionnaire is reproduced in full
in Appendix A. In the next section we will review the children’s answers to the
questionnaire to present a snapshot of their self-reported mobile phone use.

Typical patterns of texting and text exposure

via phones
Overall the children reported that they typically sent and received between three
and five text messages a day, although the range on these estimates of use was
between zero and more than 10 in both cases. Almost 60 per cent of the children
(59.7 per cent) reported preferring to use their phone to send a text message
compared to calling someone, and the same proportion reported that they would
send text messages at times when they would otherwise not bother to contact
that person. When asked how many of their friends they texted, the typical figure
reported was just three, although there was a wide range of responses, with some
children (in both age groups) reporting that they did not text their friends at
all, and one child (at secondary school) reporting that they texted around 300
56 Phone behaviours and written language

50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
Always Mostly Sometimes Never

Figure 6.1 Responses to ‘Do you always understand the textspeak that other people use?’
(Question 33).

friends. A similar pattern emerged when the children were asked how many other
people they texted. Again the typical (median) figure reported was around three
other people, but some children reported not texting others at all, and one child
reported texting 91 other people (again, a secondary school pupil). When asked
whether the children always understood the textisms that other people used, just
under half (46.8 per cent) reported that they always did (see Figure 6.1).
In order to get a sense of how often the children used their phones in
ways that might increase their print exposure, excluding texting, we asked
‘How often do you use your phone to browse the internet?’ and ‘How
often do you use your phone to access social networking sites like bebo
or facebook?’ Most children responded that they browsed the internet on
their phone ‘now and again, but not regularly’ and used it to browse social
networking sites ‘very rarely’. However, 76.1 per cent reported using their
phone to play games, and 57.2 per cent of those games were reported to
involve text in some way. Only 14.4 per cent reported using Twitter on
their mobile phones, and 26.9 per cent used it to instant message their
friends more than text them. So, outside of text messaging, there was only
limited evidence of print exposure via mobile phones, and that primarily
originated from playing games.

Types of technology and predictive text

In terms of the types of phones owned by the children in this study, 46.8
per cent of the children surveyed reported owning a smart phone, and 33.3
 Phone behaviours and written language 57

Table 6.1 Children’s mean scores on written language tasks by predictive text group

Yes No Sometimes

Mean Std dev Mean Std dev Mean Std dev

Spelling 96.7 11.2 100.0 14.5 97.1 12.5

Phonological awareness 20.5 6.9 21.7 5.7 20.8 5.7
Phonological processing 83.8 13.0 88.0 22.3 87.9 21.2
Orthographic processing 51.9 26.4 46.5 21.9 53.1 22.2

per cent said that their phone had a QWERTY keyboard, 23.9 per cent had
an alphabetic keyboard and 40.8 per cent used a more traditional number
pad on their phone. With respect to predictive text use, only 18.4 per cent
of the children reported that they did use predictive text, with an additional
21.9 per cent reporting that they sometimes used it. This relatively low level
of predictive text use amongst the children is perhaps explained by the fact
that predictive text can interfere with the user’s ability to construct their own
spellings and text abbreviations, and so many children keep this function
switched off for this reason. However, it should equally be noted that it is
possible to programme text abbreviations into the dictionary of the phone,
and in fact many phones now include the most commonly used textisms in
their dictionaries.
A question that we are frequently asked is ‘Is predictive text use associated
with literacy skills’? We examined this question in this study by comparing the
children who reported never, always and sometimes using predictive text on their
performance on the written language tasks. No significant differences between
these groups of children were found on any of the tasks (see Table 6.1).

Levels of ‘addiction’
As mentioned earlier, one of the things that we were keen to explore in this
study was whether there was evidence of the children appearing to be highly
dependent on – or even addicted to – their mobile phones. When asked how
often the children carried their phones with them, 43.8 per cent reported
‘all of the time’ and 52.2 per cent reported that it was ‘very important’ to
keep their phone charged (see Figures 6.2 and 6.3).
We also asked the children whether they spent more time using their
mobile phones than other devices or engaging in other technologically-
mediated or reading activities, and their responses are summarised in
Figure 6.4. This showed that there was a tendency for the children to use
their phones more than reading books (49.3 per cent) but generally other
activities were completed with the same or greater frequency as phone use.
However, it is still of some concern that in addition to the responses in
58 Phone behaviours and written language

50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
All the time Most of Only when Rarely Never
the day I think
I will need it

Figure 6.2 Responses to ‘How often do you carry your phone with you?’ (Question 7).

60%

50%

40%

30%

20%

10%

0%
Very Quite Not that Not important
important important important at all

Figure 6.3 Responses to ‘How important is it to keep your phone charged?’ (Question 8).

relation to book reading, 38.3 per cent of the children reported spending
longer on their phones than doing homework.
The average number of hours that the children spent using their mobile
phone each week was reported to be a quite modest 18.6 per cent, although
 Phone behaviours and written language 59

70%

60%

50%

40%

30%

20%

10%

0%
Yes No About the I don’t have
same one / do that

Your computer TV Video games

Homework Reading books

Figure 6.4 Responses to ‘Do you spend more time using your phone than…?’ (Questions
13–17).

one child reported 160 hours of phone use per week, which suggests some
degree of exaggeration (we hope) in the case of this child, as this would average
out as almost 23 hours per day of phone use. So the typical pattern of response
indicated that the majority of children did not use their phones in a way that
could be considered excessive, although clearly there is evidence of some more
extreme reports within the sample.
In the UK-based research studies we have conducted, we have typi-
cally found that the children were not permitted to bring their phones to
school. When we asked the children in this study whether bringing phones
to school was allowed, 46.3 per cent said that it was, but 48.3 per cent
reported that they actually did bring their phones, which included 15 chil-
dren out of the 201 who went to schools where this was not permitted.
When asked whether they believed that they needed their phone every day,
62.2 per cent reported that they did. However, when we asked them how
they felt when they left their phone at home, 47.3 per cent stated that it
did not bother them (see Figure 6.5).
Overall, these responses do suggest some high levels of usage and indicate
that the children viewed the technology as important to them, but at the
60 Phone behaviours and written language

50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
I usually It doesn’t I am frustrated, I am anxious,
don’t realise bother me as I would as I need it
have used it

Figure 6.5 Responses to ‘How do you feel when you leave your phone at home?’ (Question
32).

same time there was little consistent indication that the children should be
characterised as being over-dependent on mobile phones.

Enjoyment and motivation

Another area that we were keen to explore with the questionnaire was
the extent to which the children enjoyed using their mobile phones, and
whether that enjoyment may be motivational for the children in important
ways, such as encouraging them to engage with text or rehearse phonic
knowledge by inventing new spellings. This enjoyment factor was some-
thing that we (Plester et al., 2009) had previously speculated might be
able to explain the valued-added contribution of textism use to spelling
development. The children’s responses to the questions about how much
they enjoyed phone use are summarised in Figure 6.6. The overwhelming
majority of children reported that they enjoyed using their phones (75.6
per cent), using textisms (65.2 per cent) and creating their own textisms
(57.2 per cent).
Finally, we asked the children to indicate all the things that they enjoyed
about their phones, and their responses are indicated in Figure 6.7. As can
be seen, all of the features we highlighted to the children were enjoyed by at
least 70 per cent of the children in the study. The most enjoyable aspect was
reported to be social; that of being able to contact friends (95.5 per cent) and
not having to worry about spelling was valued by over three quarters (75.6 per
cent) of the children.
 Phone behaviours and written language 61

80%

70%

60%

50%

40%

30%

20%

10%

0%
Yes A little I neither No, not No, not at
enjoy or really all
dislike it

Using your phone

Using smiley faces, alternative spellings etc. when texting
Creating your own alternative spellings and other types of abbreviation when texting

Figure 6.6 Responses to ‘Do you enjoy…?’ (Questions 28–30).

How do primary and secondary school children’s

mobile phone behaviours compare?
The recruitment of separate primary and secondary school samples enabled
us to examine whether or not there were any differences between the two
groups of children in their responses to the questionnaire. The number of
text messages that the children reported sending and receiving did not dif-
fer substantially between the primary and secondary school children, with
both reporting sending and receiving around three to five messages a day.
The secondary school children used their phones to access the internet and
for social networking more frequently than the primary school children did
(‘very rarely’ compared to ‘now and again, but not regularly’). The primary
school children reported carrying their phones with them less often than the
older children did (‘most of the day’ compared to ‘all the time’). However,
the secondary school children reported using their phones for significantly
62 Phone behaviours and written language

Other
Not worry about spelling
Have a nice phone
Personalise phone
Listen to music
Make videos
Take photos
Access internet
Play games
Make calls
Send texts
Contact friends

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 6.7 Analysis of things children reported enjoying about their phone (Question 33).

longer periods each week2 (ten hours compared to four hours; p = 0.0013).
The primary school children reported receiving their phones at a significantly
younger age than the secondary school children did (p < 0.001), typically at
eight years of age, rather than at age nine years. There were no other signifi-
cant differences in mobile phone behaviours reported by the children.

The relationship between text messaging behaviour

and literacy outcomes
The children who appeared to be highly-dependent phone users also tended
to be the children with strong text processing skills and rapid phonological
retrieval abilities. For example, perhaps as we might expect given previous work
in the area, there was a significant tendency for the children’s performance on
the rapid picture naming task and orthographic processing test to improve as
the frequency of the children carrying their mobile phones with them increased
(Jonckheere-Terpstra Test, 2.098, p = 0.036 and 2.045, p = 0.041 respectively).
Similarly, children who said that they took their phone to school (n = 97) had
significantly faster picture naming speeds (phonological retrieval) than children
who did not (n = 102, U = 6761, p < 0.001) and significantly better ortho-
graphic processing scores (U = 4109.5, p = 0.039).
However, it was not the case that high levels of phone dependency
were consistently associated with positive outcomes. With respect to
the children’s standardised spelling scores, these seemed to decrease as
 Phone behaviours and written language 63

reported levels of phone carrying increased (Jonckheere-Terpstra Test, 2.349,

p = 0.019). Similarly, spoonerism performance (phonological awareness)
was found to decline as responses to ‘How important is it to keep your
phone charged?’ indicated increased importance (Jonckheere-Terpstra Test,
1.992, p = 0.046). This suggests that whilst high levels of phone use may
have some benefits for the speeded processing of text (most probably as a
consequence of texting exposure), there is also evidence that overreliance on
mobile phones could be linked to poorer performance on spelling-related
tasks. This may be the result of too much exposure to text messages, which
might limit the children’s exposure to correct orthographic forms, which (as
we noted in Chapter 3) are observed to enhance children’s spelling perfor-
mance (Dixon & Kaminska, 2007). However, we would emphasise the point
made earlier in Chapter 3 that there is no evidence to suggest that exposure
to misspellings or alternative spellings can harm or compromise children’s
understanding of conventional spelling.
The idea that too much use of mobile phones could be problematic for
some literacy skills is also borne out to some extent by correlational analyses
of the data from this study. After controlling for the influence of verbal IQ,
the age at which the children received their first mobile phone was positively
related to their standardised spelling scores (pr = 0.218, p = 0.003) and perfor-
mance on the orthographic processing task (pr = 0.195, p = 0.008), suggesting
that children who were older when they received their first phone tended to
fare better on these tasks.
Similarly, children who reported texting a smaller number of friends tended
to score better on the phonological awareness (spoonerisms) task than the
children who texted a larger number of their friends (pr = 0.195, p = 0.008).
These findings are fascinating, not just because they suggest again that exces-
sive levels of phone use, even for texting, could be a problem, but because
any negative effects of exposure to textisms may be mitigated if children’s
exposure to textisms occurs within a limited circle where there is likely to be
greater consistency in the ways in which textisms are developed and used.
Exposure to too wide a range of alternative spellings could be disorientating
and problematic for the children’s developing phonological and orthographic
skills. We did find that children who reported that they only sometimes or
never understood the textisms that others used in the messages sent to them
(n = 44) scored significantly more poorly on the test of orthographic process-
ing than the other children (n = 156; U = 2218.5, p < 0.001). It seems likely
that when texting across a wider network of friends there is an increased pos-
sibility of encountering textisms which are difficult to decode or interpret.
When the correlational analysis was repeated for each of the two school
groups individually, we found that for the primary school children there was
a significant positive partial correlation between age of phone acquisition and
orthographic processing (pr = 0.320, p = 0.001) after individual differences
in the children’s verbal IQ were taken into account. This means that, even
64 Phone behaviours and written language

taking out the effects of verbal ability, the older that children had been when
they received their first phone, the better was their current ability to process
orthographic information. This relationship was not observed for the sec-
ondary school children. So it seems that for the primary school sample in
particular, it was better for them to receive their first phone later rather than
earlier. However, it should be noted that this apparent relationship does not
necessarily imply causation. That is, there may be a third variable at work
here which can account for this association. For example, perhaps the par-
ents of children who restricted their children’s access to mobile phones until
later were also parents with greater involvement in or engagement with their
children’s education. Similarly, there may be an association between parental
income levels and the age at which parents are willing to purchase a phone
for their offspring. The nature of this association therefore needs careful
consideration before forming firm conclusions.
For the secondary school children, the most salient issue was to do with actual
phone use. That is, there was a positive association between phonological aware-
ness scores and how important the children stated that keeping their phone
charged was (pr = 0.322, p = 0.004) and a negative relationship between pho-
nological awareness and how many friends they texted (pr = 0.301, p = 0.008),
again taking into account differences in verbal IQ. This suggests that, for the
older children, the development of phonological awareness may be influenced
by their level of phone dependence and they need to exchange text messages
within a smaller network of friends.
Finally, in terms of the children’s self-reported enjoyment of using text
abbreviations, there were no statistically significant differences in the lit-
eracy scores of the children who reported that they did enjoy using them
(n = 170) as compared to those who did not (n = 25). The mean scores
of these two groups, plus those of the children who neither enjoyed or
disliked it, are presented in Table 6.2. This pattern was repeated for the
children who reported enjoying creating their own textisms (n = 156)
compared to those children who did not (n = 34; see Table 6.3).

Table 6.2 Literacy scores of children who reported enjoying versus not enjoying using
text abbreviations when sending text messages

Yes No, not really I neither enjoy

A little No, not at all or dislike it
(n = 170) (n = 25) (n = 6)

Mean Std dev Mean Std dev Mean Std dev

Spelling 98.2 13.1 101.1 14.1 108.2 20.8

Phonological awareness 21.0 5.9 21.4 5.8 26.7 3.7
Phonological processing 86.9 20.5 90.6 20.9 80.0 18.8
Orthographic processing 49.3 23.0 48.2 25.9 61.7 31.8
 Phone behaviours and written language 65

Table 6.3 Literacy scores of the children who reported enjoying versus not enjoying
creating text abbreviations

Yes No, not really I neither enjoy

A little No, not at all or dislike it
(n = 170) (n = 25) (n = 6)

Mean Std dev Mean Std dev Mean Std dev

Spelling 98.4 13.1 101.0 13.7 101.1 8.3

Phonological awareness 21.0 6.1 22.1 5.0 24.2 3.6
Phonological processing 87.2 20.4 89.9 22.7 80.9 11.7
Orthographic processing 49.9 24.3 49.3 23.5 44.1 11.1

So what have we learned?

One of the questions we were interested in was whether there was any evi-
dence of the children being ‘addicted’ to mobile phone technology. The
data presented here suggest that there is very little evidence to support this
popular characterisation of children and young people as, although there was
some evidence of high levels of phone use, there was little evidence of chil-
dren having been overly-dependent on phones. We argue that the pervasive
negative stereotype of the “child as technology addict” needs to be directly
challenged. Moreover, if we as adults have a problem with the idea that
children and young people have access to this technology, then we need to
recognise that we appear to be the ones encouraging and supporting this;
around 40 per cent of the children in this sample either received their phone
as an unsolicited gift or they were told that they ‘needed’ to have a mobile
phone. Only around a third of the children in this study actually asked for
a phone.
In a similar way, the stereotypical view of children continually text mes-
saging on their phones is also not entirely supported by the data here. That
is, whilst more children did prefer to text rather than call their friends, it
should be noted that text messaging was not the main activity that the
children reported undertaking: the most popular mobile phone activity was
that of playing games, which over three quarters of the children in the
study engaged in. So it seems that when we see children and young people
staring intently at their handsets, pressing keys with a look of fixed concen-
tration, this is more likely to be because they are playing the latest mobile
game rather than ignoring the immediate social context in order to text
their friends who are not there. Although excessive game play may have its
own issues, it is worth remembering that many of these games involved the
presentation of text on the screen and therefore also offered the children
print exposure.
66 Phone behaviours and written language

We were also interested in the children’s enjoyment of mobile phone tech-

nology, and we did find evidence of high levels of enjoyment; this included
enjoying the use and creation of alternative spellings. However, there was lit-
tle evidence that self-reported levels of enjoyment in creating and using text
abbreviations was linked to the participants’ performance on the measures
of written language ability. However, it should be noted that the data from
this study are concurrent and therefore only offer a snapshot of the children’s
abilities at one point in time – it is possible that a longitudinal analysis of
similar data could show evidence of an association between enjoyment in the
use and creation of these textual forms and the development of these skills
over time.
Perhaps surprisingly, there were not too many significant differences
between the primary and secondary school children in their responses to
the questionnaire; the two most noteworthy were age of first phone acqui-
sition (the primary school children received their phones on average a year
younger than the secondary school children did) and the number of other
children that they sent text messages to. We have reiterated these findings
here because these two variables were also found to be linked to the chil-
dren’s scores on the written language tasks. Age of first phone acquisition
was positively linked to orthographic processing ability, perhaps suggest-
ing that it was better for the children to be older rather than younger when
they received their first phone. This may be because a level of metalinguis-
tic awareness or print exposure ideally needs be in place before the children
are most able to benefit from exposure to and use of textisms. However, as
noted earlier, this finding needs to be interpreted with caution, as there are
other factors which could explain the nature of this association. The size
of the children’s texting network was negatively related to phonological
awareness, and we have interpreted this finding as perhaps indicating that
textism exposure is most beneficial if it is ‘managed’ by gradual exposure
to more diverse forms: smaller networks of friends are likely to have less
diversity in the range of textisms they use collectively for the same words.
It seems plausible that this is important in enabling the children to man-
age their understanding and application of letter (number) sound rules, at
least initially.
Overall, it is striking just how little relationship there was between
mobile phone behaviours and written language skills, which underscores
the earlier findings that what seems to be important is the specific use of
text abbreviations when texting rather than texting per se. Previous chap-
ters led us to argue that, whilst there is no reason to deliberately purchase
a mobile phone in order to develop a child’s developing literacy abilities, if
they already have one, there is no need to be concerned. Taken together,
the analysis of the data from this new study has further suggested that there
is no significant impact of phone type on literacy outcomes, and there is
also no need to be concerned as to whether the children use predictive text.
 Phone behaviours and written language 67

What is important is the age at which the children receive their first phone
(older is better than younger) and the number of other children they text
(smaller is better than larger).

Notes
1 This study was conducted with the assistance of Lucy Hart, Sam Waldron and Roy
Bhakta.
2 Where significant differences between groups are reported, a Mann–Whitney U
test has been performed to determine this.
3 ‘p’ is used to indicate statistical significance where it is possible to test this. ‘p’ refers
to the probability of the observed result being spurious rather than reflecting a
genuine difference or relationship. For example p < 0.05 represents a less than 5%
chance that the result highlighted is not genuine, p < 0.01 = less than a 1% chance,
and so on. The smaller the p value, the better.
Chapter 7

Texting and grammar

Much of the research that has been conducted so far has focused on the
way that individual words are spelled in text messages. However, as well as
changing the spelling of particular words, texters often also ignore or bend
the constraints imposed on standard writing by grammatical rules and con-
ventions. Although some of these constraints also overlap with orthographic
rules, in this chapter we discuss the range of conventions at the level of the
word, phrase and sentence that are governed by grammatical or syntactic rules,
that many people would consider to constitute a knowledge of grammar. For
example, we consider word-level transgressions, such as omitting the capital
letter from a personal name (Hi ben), or the apostrophe from a contraction
(Lets go), as well as the use of verb forms (how is you?) and word combina-
tions (wanna, hafta) that are not unique to texting, but that do not represent
conventional (written) language. We also consider sentence-level grammati-
cal transgressions. It is common to see text messages with no sentence-initial
capitals, with minimal punctuation, or with symbols or emoticons used as
punctuation marks or discourse markers (hi - how are u). Writers of text mes-
sages may also omit words or construct sentences in unconventional ways.
Such unconventional ways of writing may be used deliberately: to save time,
effort or characters, or to add a more casual or friendly tone to a message.
Alternatively, the frequent use of unconventional grammar may signal that a
young writer has not learned the appropriate conventions of written language,
or that an older writer is forgetting or ignoring the conventions learned at
school. This chapter reviews some of the research on the nature and extent of
the use of unconventional grammar in text messages, and its relationship with
knowledge of conventional written language.

Punctuation
English, like other languages, is governed by grammar-based rules and con-
ventions, some of which apply to spoken language and some only to written
language. At the level of the phrase or sentence, some conventions concern
the use of punctuation. Spoken language has pauses between phrases and
 Texting and grammar 69

sentences, and these are often represented in writing by punctuation marks,

including commas, semi-colons and full stops. Sentences that indicate strong
feeling can be ended with an exclamation mark and interrogative sentences
with a question mark. Direct speech is indicated by quotation marks, and
possession and contraction by apostrophes. Children are not always taught
punctuation in the same systematic way as they are taught the letters and
numbers. Further, many punctuation marks represent concepts that take
some time to acquire. For example, a child might find it difficult to use
question marks correctly before she has a strong grasp of what makes a sen-
tence a question, or find it hard to end sentences consistently with a full stop
before he has fully understood what a sentence is. Perhaps for these reasons,
beginning writers sometimes omit punctuation marks in their conventional
writing in the primary years (Wagner et al., 2011; Wilde, 1988) and even
into high school (Lee & Gavine, 2003).
Literate adolescents and adults generally use the more frequent punctuation
marks appropriately most of the time, but anyone who has marked student work
even at the university level will have noted transgressions of varying frequency.
One type that seems especially difficult to learn is the apostrophe, whether it is
used to signal possession (the boy’s shoe, the boys’ game) or contraction (the boy’s
leaving). Although there are improvements across grade level, even in conven-
tional writing, children in primary school have been shown to use apostrophes
correctly only about 15–50 per cent of the time (Bryant, Nunes & Bindman,
2000; Leong, 2009; Stuart, Dixon & Masterson, 2004), and even university
students use apostrophes little more than half of the time for singular possessive
nouns (Hokanson & Kemp, 2012).
Text messaging is a medium in which standard grammatical conventions
are not always followed, and it seems as though the conventions regarding
punctuation are some of the most regularly flouted. There are some quantita-
tive data on the use of punctuation in text messages. Ling and Baron (2007)
examined a 24-hour block of text messages sent by 22 female college students
in the United States; 191 messages in all, totalling 1473 words. The authors
also compared the features of these messages with a similar sample of com-
munications made by Instant Messaging (IM), where the whole QWERTY
keyboard and a larger screen were available, compared to the more limited
alphanumeric keypads and small screens of phones at the time. The relative
lack of punctuation in mobile phone messages was obvious. Sentences that
occurred at the end of a message received sentence-final punctuation only
29 per cent of the time (compared to 35 per cent in IM), although
non-message-final sentences were punctuated 54 per cent of the time (78
per cent in IM), presumably to help the reader distinguish sentences. Stu-
dents appeared to be more careful about signalling questions: overall, ques-
tion marks were used at the ends of 73 per cent of questions in texts (100
per cent in IM), but other types of sentence-final punctuation were used in
only 30 per cent of sentences in texts (41 per cent in IM). Finally, contractive
70 Texting and grammar

apostrophes were used in 32 per cent of the places that they were required
in texts, but in 94 per cent of required contexts in IM. The differences seen
between texting and IM suggest that at least some of the reason for omit-
ting punctuation can be attributed simply to the relative difficulty of inserting
marks and characters when using an older mobile phone, compared to using a
computer keyboard and screen. However, it is also likely that punctuation
marks are left out deliberately to save space or time. For example, Herring
and Zelenkauskaite (2009) examined a subset of 800 text messages sent from
viewers of an Italian interactive television programme. These Italian text-
writers were presumably more familiar with texting by mobile phone than
participants in the US study by Ling and Baron published two years earlier.
Herring and Zelenkauskaite’s writers omitted conventional punctuation in
an average of 26 per cent of the messages examined, and omitted inter-word
spaces in nearly 2 per cent of messages, both of which appeared to be deliber-
ate devices for saving screen space and/or texting time.
One type of punctuation mark whose use in text messages has been reported
on more widely is the apostrophe. The methods of reporting have varied, but
they do show converging evidence for a tendency to omit apostrophes from the
language of texting. Plester, Wood and Joshi (2009) elicited ten text messages
from 88 British children of 10 to 12 years, and found that 61 of these children
omitted apostrophes where they were required, an average of 3.23 times each.
De Jonge and Kemp (2012) asked Australian teenagers and university students
to rewrite a series of conventional English sentences into text messages, and
reported that, of all the textisms produced, 17 per cent consisted of missing
apostrophes. In a similar analysis of the textisms seen in a set of naturalistic mes-
sages sent by young American adults (Drouin and Driver, 2012), 11 per cent
of these textisms were omitted apostrophes. The participants in De Jonge and
Kemp’s study all used predictive text entry, whereas this was the case for only
59 per cent of Drouin and Driver’s participants. However, predictive entry can-
not reliably correct apostrophe use, as there are a multitude of words in which
both versions constitute real words (e.g. its/it’s, as well as most nouns, such as
cars/car’s) and most phones do not yet have sufficiently sophisticated gram-
mar checkers to identify the correct form in a consistent manner. Finally, in an
online survey of American adults of varying educational backgrounds, Rosen,
Chang, Erwin, Carrier and Cheever (2010) found that respondents rated their
mean likelihood of omitting apostrophes from text messages at 3.31, on a five-
point Likert scale from 1 (never) to 5 (very often). It is difficult to compare
these statistics with those concerning the omission of apostrophes in formal
writing. However, it does seem that although some of these punctuation marks
are probably omitted from text messages through ignorance, in other cases the
rules might be known but flouted through a lack of care, or a desire to save
time and effort when creating a message.
When they do use standard punctuation marks, text messagers may use
them in unconventional ways, to bring a sense of casualness, urgency or
 Texting and grammar 71

fun to their communication. Writers may end their sentences with multiple
exclamation marks and/or question marks (did u see that???!!!), or sepa-
rate phrases with ellipses rather than with standard commas and full stops
(just writing to say hi... im bored... are u?). The use of multiple punctuation
marks may not be widespread, at least in English-language text messages: for
example, Thurlow and Brown (2003) noted just 0.68 per cent in the adults’
naturalistic messages they analysed. However, Herring and Zelenkauskaite
(2009) noted more numerous uses of multiple punctuation in Italian text
messages to an interactive television programme, with an average of 1.2
examples per message. Multiple exclamation and question marks might be
more the domain of younger texters: adolescents interviewed for the 2010
Pew Internet Survey (Lenhart, Ling, Campbell & Purcell, 2010) com-
mented on their own use of excessive punctuation, with boys anecdotally
observing that girls seemed to use it much more than boys did.
A much-discussed characteristic of text messaging, compared to conven-
tional writing, is the use not only of standard letters and numbers, but of
symbols (@, +), also including hugs and kisses (xxox) and emoticons (-). It
is a widespread assumption, common in media headlines on the dangers of
texting, that text messages are riddled with these kinds of symbols, but their
overall use is in fact relatively limited. In text messages translated by British
children, Neville (2003) found symbols to be used by only 2 per cent of
participants, and Plester, Wood and Bell (2008, Study 2) found that sym-
bols made up only 1 per cent of all textisms produced. Plester et al. (2009)
showed that across ten elicited text messages, 33 of their 88 child participants
used symbols (including emoticons), a mean of 4.30 times. Adults’ use of
symbols seems even more limited. Thurlow and Brown (2003) reported that,
of all the words in their naturalistic corpus, only about 7 per cent were sym-
bols and less than 1 per cent were emoticons, and Ling and Baron (2007)
observed just 2 emoticons in their 1473-word text message corpus. Similarly,
of all textisms produced, only 3–4 per cent represented symbols, in both the
translated messages elicited by De Jonge and Kemp (2012) and the natural-
istic messages gathered by Drouin and Driver (2012).
Most researchers have simply reported the number of symbols used in the
messages examined, as described above. However, a look at the raw data of
text messages suggests that, when symbols are used, their appearance is not
random, but highly predictable, and concentrated at the beginnings and ends
of phrases and sentences. Thus, text messagers are much more likely to write
- Just to say hi or r u there - than to interrupt a phrase with an emoticon or
other symbol. Provine, Spencer and Mandell (2007) confirmed this tendency
in a study of instant messaging. These authors examined 849 statements (all
containing one or more emoticons) posted by 226 users of several web-based
message boards, and found that, in 99 per cent of cases, emoticons appeared
before or after sentences or at mid-sentence phrase breaks. For many texters,
then, symbols – especially emoticons – are being used in place of conventional
72 Texting and grammar

punctuation. This is clearly a novel way of addressing sentence-level gram-

matical constraints. It plays the dual role of separating phrases and sentences,
and imbuing social information. The use of such unconventional punctuation
seems to represent a deliberate way of playing with language, rather than
ignorance of the conventional punctuation marks of written English.

Capitalisation
In many alphabetic orthographies, there are both sentence- and word-level con-
ventions about the use of capital letters. In English, the first word of a sentence
should start with a capital, regardless of its grammatical status, and proper nouns
always need a capital, regardless of their place in the sentence. Proper nouns
include the names of people, places, days and months, languages, brands and
institutes. To be able to use capitalisation consistently correctly, a writer must
learn the relevant underlying grammatical conventions, and some of these dis-
tinctions can take children some years to master (e.g. Wagner et al., 2011). A
lack of conventional capitalisation is a common characteristic of text messages,
probably because of the relative effort involved in switching between upper- and
lower-case letters on at least some phones, and for at least some words. Rosen
et al.’s (2010) adult respondents reported that on a scale of 1 (never) to 5 (very
often), their mean likelihood of using i for I was 3.43. Thurlow and Brown
(2003) also noted the generally minimal use of capitalisation in the naturalistic
text messages that they collected from British undergraduates.
However, other researchers have not reported the number of times that
capital letters have been incorrectly omitted (or included). This is often
because it is difficult to determine whether the text writer has deliberately
used a capital, or whether the capitalisation has been inserted automati-
cally by phone’s predictive-entry software, at least for sentence-initial words
and pre-programmed proper nouns. For example, in De Jonge and Kemp’s
(2012) study of Australian high school and university students who all used
predictive entry, nearly one-quarter (24 per cent) of the textisms produced in
a message translation task represented omitted capitals. In contrast, in Drouin
and Driver’s (2012) study of the naturalistic text messages sent by US under-
graduates, only 59 per cent of whom used predictive text entry at least some of
the time, nearly 40 per cent of the textisms consisted of omitted capitals. This
does suggest that the potential role of predictive text entry should be taken
into account when considering the use of capitalisation.
As with other examples of unconventional grammar common to texting, fre-
quent exposure to uncapitalised proper nouns and sentence-initial words may
make these spellings appear acceptable, both to children still learning the conven-
tions of capitalisation and to adults who may begin to ignore or forget them. The
discomfort that many people feel at seeing their own name written all in lower
case, or at receiving emails with the pronominal I written as i, does not seem
to be shared by everyone, especially not by younger people who have grown
 Texting and grammar 73

up with digital communication. The growing number of printed advertisements

that feature lower-case pronominal i (which presumably were no quicker to
design or print than an advertisement with the conventional I) suggests that this
is one capital that may fade from use in the future.

Omission of words
In casual spoken English, it is possible to omit auxiliary verbs (you want some
lunch?) and/or pronouns (want some lunch?) without losing meaning. These
types of whole-word omissions are also common in text messaging. Further,
other types of words, including articles, personal pronouns and prepositions,
are often left out of text messages, leaving just the essentials of the message, as
in an old-fashioned telegraph (front tyre gone flat; what time home?). It seems
clear that these types of omissions represent the writer’s attempt to save time,
effort or characters, rather than a misunderstanding or forgetting of the gram-
matical requirements of spoken or written English. It may require a relatively
sophisticated understanding of language to be able to make messages more
concise like this. It seems likely that younger children might find it difficult to
focus in on the essential words and omit the extraneous ones in order to create
concise messages in this way. However, there is little research on this aspect of
children’s texting, which has focused more on the much more prevalent ten-
dency of children to respell words in different ways.
Bodomo (2010) suggests that, when attempting to shorten a sentence to
be texted, writers omit functional categories such as tense and aspect, but leave
lexical categories such as nouns and verbs. He gives interesting examples of
the ways that writers may achieve such shortenings, but overall in the literature
there seems to be relatively limited quantitative data on these kinds of omis-
sions. One source of information is a corpus analysis by Tagg (2009), who
examined the omissions of pronouns in a corpus of over 11,000 text messages
written by British adults. Tagg noted that the indefinite article a(n) was missing
in 16 per cent of the contexts in which it would be required in formal writing,
and the definite article the in 31 per cent of such contexts. Personal pronouns,
especially I, were also often omitted. I was missing before the verb am in 53
per cent of cases (am on my way), and before was/were in 13 per cent of cases
(Sorry was at the grocers). Other verbs also occurred without a subject pronoun:
will in 71 per cent of occurrences, and have in 10 per cent of occurrences. Some
of these omissions correspond to what one may hear in casual speech. How-
ever, Tagg also notes other omissions that are quite uncharacteristic of spoken
language. In 2 per cent of the uses of is in her corpus, the subject was omitted
(is easy once have ingredients), and in other cases (number not reported) the
subject was retained but is (or other form of be) was omitted (hope your day
good). Again, these kinds of omissions appear to reflect the deliberate saving of
time or space, or a purposeful tone of brevity or informality, rather than text-
writers forgetting how to write in grammatical English.
74 Texting and grammar

Ungrammatical word forms

The English spelling system is basically alphabetic, but there are many words
whose spelling is determined by grammatical conventions. Some words can be
spelled in several ways on the basis of their sound, but have only one correct
representation in terms of grammar (e.g. their/they’re; its/it’s). These grammati-
cal homonyms can be challenging to learn to spell, and are very often written
incorrectly even by adults (e.g. Kemp, 2009). Other words have endings that
should be spelled in a consistent fashion to show their common grammatical
status (e.g. s for plurals; ed for regular past-tense verbs), despite differences in
pronunciation (e.g. shoes, socks; kissed, hugged, cuddled). It takes some years for
children to stop spelling such patterns phonetically (e.g. kist for kissed), and to
start using grammar-based spellings instead (Nunes, Bryant & Bindman, 1997),
and inconsistent spellings can persist even into adulthood (Kemp & Bryant,
2003). It is common in text messages to write some words just as they sound,
whether for the sake of efficiency (e.g. no for know), because of the carelessness
or uncertainty that leads to errors in formal writing (e.g. your for you’re), or in
order to play with spellings for humorous, social or ironic effect (e.g. frenz for
friends). Thus a teenager might spell you’re as your because she is dashing out
a message and can’t be bothered thinking about which is the correct form. But
she might spell thanks as thanx simply because she likes the informality of the -x
rather than because she does not know that -s is required.
Finally, text messages sometimes include word combinations common
in colloquial spoken language, such as comin, wanna, shoulda, doncha and
gimme (Carter & McCarthy, 2006). Deliberately ungrammatical forms of
verbs in messages, such as how duz u know and i is busy, also sometimes appear
in text messages. Even if the former are sometimes used to save time or screen
space, or for social reasons, it is likely that both types of ungrammatical word
forms are used most often to convey a sense of fun or casualness, especially
among younger texters.

Do errors mean ignorance?

We know from the research described in the other chapters that, in general,
the use of textisms is related to better conventional literacy skills in children,
whereas in adults, the relationship is not so clear. However, previous studies have
looked at the creation and decoding of textisms in general, many of which rep-
resent new spellings of individual words. We know much less about the use of
unconventional grammar in text messaging, and its relationship – if any – with
conventional grammatical knowledge. As already noted, some violations of the
conventions of grammar seen in text messages could represent attempts to save
time, effort or the number of characters typed, and others could represent delib-
erate ways of including extra social or emotional information in a message. Thus,
we would not expect that such uses would reflect the absence or decline of con-
ventional grammatical knowledge, in either children or adults. However, it is also
 Texting and grammar 75

possible that in many cases children might transgress the conventions of written
language because they have not yet consolidated them properly and, further,
the more such transgressions children see, the more difficult they might find
it to create strong mental representations of formal grammatical conventions.
Adults who once knew the formal conventions might end up being exposed to
unconventional grammar in text messages so often that this informal writing
style might affect their memory of the conventions, or might mean that they no
longer take such care to get them right, even when writing in standard English.
Baron (2008) proposed that, with so much informal text around us, traditional
standards of English may not be declining so much as becoming irrelevant.
There are some indications that there are links between the awareness of
word structure, or morphological awareness, and the use of textisms in gen-
eral. Kemp (2010) and De Jonge and Kemp (2012) both used an oddity task
of morphological awareness, asking participants to pick the odd-word-out of
triplets with a shared ending, such as meanest, smartest, honest. For each tri-
plet, the ending of two of the words constituted a separate morpheme, or unit
of meaning (e.g. mean + est, smart + est) and the ending of the odd word was
simply part of the whole word (e.g. honest). Kemp (2010) found a positive
correlation between university students’ scores on this task and their accu-
racy at deciphering textisms in a message. De Jonge and Kemp (2012) found
a negative correlation between teenagers’ and young adults’ morphological
awareness scores and the proportion of textisms they used when translating
standard English sentences into text messages. However, the textisms that
participants produced did not necessarily represent unconventional use of
grammar, and the text-based tasks involved the reading and translating of mes-
sages in an experimental situation rather than a naturalistic one.
Cingel and Sundar (2012) considered the concurrent relationship between
the naturalistic use of textisms and a more general measure of grammatical
performance in a group of US adolescents aged 10 to 14 years. These authors
asked participants to look at their own last three sent and three received text
messages, and to count and categorise the textisms that they saw, according to a
five-category classification scheme. Participants also completed a 16-question
grammatical assessment that tested knowledge of verb tense and agreement,
the spelling of grammatically-determined homophones, and the use of punc-
tuation and capitalisation. Unfortunately no scores are reported for either the
texting or the grammatical task, but the authors report a significant negative
relationship between textism use (in sent messages) and grammatical score.
This was attributable to textisms based on adaptations to words: abbreviations/
initialisms (e.g. lol), letter omission (e.g. u for you, although many researchers
see this as a letter homophone) and homophones (e.g. be4)); but not to tex-
tisms based on adaptations of structure (non-conventional use of apostrophes
and other punctuation, and capitalisation). These results are interesting, but
further research is needed to determine the accuracy with which children as
young as ten can categorise their own textism use, and without mean scores or
76 Texting and grammar

reliability being reported, it is difficult to know how well the grammatical task
differentiated participants’ performance. This is particularly important, as the
questions were taken from a state assessment aimed at older children. Impor-
tantly, the results of this study, like those of the studies reported in Chapter 3
on literacy, should not be interpreted as indicative of likely patterns of cause
and effect, as longitudinal evidence is needed to address this. Moreover, stand-
ardised assessments of grammatical competence need to be included in work
of this nature where they are available, as an unstandardised 16-item assessment of
grammar is unable to indicate whether the sample in this study was typical or
atypical in terms of its grammatical competencies.
In recent research funded by the Nuffield Foundation, we have also begun
looking more specifically at the use of grammar in naturalistic text messages, and
comparing it to people’s ability to correct examples of unconventional grammar
in given text messages, as well as to their scores on standardised grammar and
literacy tests. We have examined the use of unconventional grammar in the nat-
uralistic text messages of 89 children in primary school (Years 4 and 5),
84 children in secondary school (Years 7, 8 and 10), and 70 university students.
We examined five recent text messages sent by each of these British students,
and coded the number of uses of unconventional grammar as a proportion of
the number of words texted overall. The primary and secondary school students’
messages were remarkably similar, with a proportion of grammatical transgres-
sions of about 50 per cent, whereas the proportion seen in the adult group was
about half that; 24 per cent. Children omitted punctuation marks (about 22 per
cent overall) and used unconventional grammar at the level of individual words
(about 14 per cent overall). These individual word errors included omitting
pronouns, using grammatical homonyms, such as your for you’re, using ungram-
matical word forms, as in I is good, and making colloquial combinations, such as
hafta and wanna. Adults also made individual word transgressions (about 8 per
cent) and omitted punctuation (about 7 per cent). The number of omitted capi-
tals was of course constrained by the number of words which required capitals,
but as a proportion of all words texted, it made up about 9 per cent for children
and 3 per cent for adults. Finally, the use of unconventional punctuation (kisses,
emoticons, ellipses and multiple punctuation marks in place of single standard
marks) was most popular with secondary school students (10 per cent), but
also seen in adults (7 per cent) and primary school students (4 per cent). These
results confirm the previous findings of the widespread use of unconventional
grammar in the everyday text messages of individuals of a range of ages.
However, it is still unclear whether violating the conventions of written
English in text messages necessarily reflects ignorance of those conventions,
as might be feared by parents and educators. To test this, we asked the same
participants to correct the unconventional uses of grammar in a given set of
12 text messages, so that they were transformed into standard English.1 The
main finding was that the proportion of grammatical violations left uncorrected
in the given text messages was significantly less than the proportion observed
 Texting and grammar 77

in participants’ naturalistic messages. In other words, all three age groups

were able to correct, when asked, the very type of grammatical violations
that they made in their everyday text messaging. Their corrections were not
complete, as errors still remained (especially for the two school groups), but
these findings do suggest that at least some of the unconventional uses of
grammar frequently seen in text messages are deliberate, and do not mean
that the writer is incapable of using grammatical constructions appropriately
when required.
In terms of whether the use of unconventional grammar in texting is
related to poorer performance on tests of literacy and grammatical skill, the
answer seems to be a clear ‘no’, at least for children. There were no significant
correlations between the grammatical errors in the naturalistic text messages of
children in primary school or high school, and their scores on tests of receptive
grammar, grammar-based spelling, orthographic processing, general spelling
ability or IQ.2 The only significant correlation that we observed was a posi-
tive one, between secondary school students’ tendency to use unconventional
punctuation and their IQ score. This suggests that the more able children are
more likely to add emotional and social meaning to their messages by substitut-
ing standard punctuation with emoticons and kisses, and/or by adding extra
exclamation marks and question marks. For adults, a clear relationship between
grammatical violations in texting and poor scores on literacy tasks was similarly
lacking. There was only one indication of a negative link: university students
who tended to omit more punctuation and capitalisation also scored more
poorly on a task which required them to choose the grammatically appropriate
spelling of a novel word.
When we considered the relationship between the proportion of grammatical
errors left uncorrected in the given set of text messages and the participants’ per-
formance on these literacy and grammar tasks, the picture was quite different.
For all three age groups, poorer ability to correct grammatical errors was associ-
ated with lower scores on the grammar-based spelling choice task. Furthermore,
leaving more grammatical errors uncorrected was associated with poorer gen-
eral spelling ability for both primary and secondary school children, and poorer
receptive grammar for the primary school children.

Conclusions
Numerous studies have shown that the writers of text messages frequently
flout many of the grammatical conventions of standard written English.
However, it appears that poorer literacy skills and grammatical understanding
are not clearly related to a tendency to violate the conventions of standard
written language when composing text messages. Instead, poorer literacy
and grammatical skills appear to be related to the poorer ability to correct
grammatical errors. However, the work conducted to date is concurrent,
and longitudinal work is needed to address the issue of whether there are
78 Texting and grammar

associations between text messaging violations of grammar and the devel-

opment of understanding of conventional grammar, although this work is
ongoing at the time of going to press. It is likely that in some cases the trans-
gression of word- and sentence-level conventions in text messages is indeed
attributable to ignorance or carelessness. However, in other cases, texters
seem to be deliberately saving time or effort, or playing with language. Over-
all, then, it seems that parents, educators and texters themselves need not be
overly concerned that exposure to unconventional grammar in text messag-
ing can ruin the ability to use grammar and spelling in standard ways.

Notes
1 Kemp, Wood, Waldron and Hart (submitted).
2 Wood, Kemp and Waldron (submitted).
Chapter 8

Methodology matters
Issues in the collection and coding of
textisms

The data discussed in this book come from a variety of studies, conducted
with a variety of methodologies. In this relatively new area, there are many
questions to answer, and researchers have developed a range of ways to col-
lect and code the text messages of children, adolescents and adults. These
methods must be designed with due regard for the ethical and financial
concerns that arise when asking participants (especially child participants)
to contribute text messages to a study – a procedure that can involve asking
people to spend their own money or to provide a glimpse of their personal
lives. In this chapter we describe the advantages and disadvantages of various
collection and coding methods.

Self-report of mobile phone-related behaviours

Some researchers are interested in people’s use of text messaging in the social
context: for example, they may wish to investigate the importance of mobile
phones in the lives of young people, the perceived benefits and responsibili-
ties they bring, or the way that they are used to interact with others. For
these more qualitative questions, (semi-) structured individual interviews or
discussions in small focus groups are useful (e.g. Blair & Fletcher, 2011;
Grinter & Eldridge, 2001, 2003). These methods are labour-intensive in
terms of the amount of time needed for face-to-face interviewing and the
later transcribing and interpretation of responses and themes, but they do
not require large numbers of participants and can often provide detailed
qualitative information that simple questionnaires cannot. Another method
is to use online or paper surveys (e.g. Baron & Campbell, 2012; Cingel &
Sundar, 2012). This does not allow researchers to follow up on answers
in the same way, but does provide the opportunity to gather more data
from more people. The focus of this book, however, is more on quantita-
tive measures of people’s use of texting and textisms, and it is important
to note that self-report measures can also be used to provide quantitative
data which can inform the analysis of textism use and literacy performance,
in the way reported in Chapter 6. As we demonstrated in that chapter, the
80 Issues in collecting and coding textisms

collection of numerical data via self-report methods can result in some exam-
ples of overestimation, but they also provide data that enable the researcher
to examine statistically the extent of any apparent relationships between vari-
ables, including data on the number of messages sent and received, or the
tendency to use textisms when texting. Although such data has the potential
to be captured accurately, many studies have relied on self-report (including
self-reported estimates) of these variables.

Self-report: number of messages and textism use

Numerous researchers have considered the frequency with which people
send text messages as a proxy for exposure to textisms (e.g. Massengill
Shaw, Carlson & Waxman, 2007) or to explore the concern that exposure
to informal writing (such as text messages, instant messages and blogs) is
overtaking exposure to formal writing, especially for younger people (e.g.
Clark, 2011). The usual method is to ask people to estimate their usual daily
rate of sending, or to recall the number of messages sent the day before. Like
any self-report measure, this question is vulnerable to reporting bias. As Ling
(2010) notes, the frequent sending of text messages might be perceived as
a mark of popularity by teenagers (leading to overestimates) or a sign of
immaturity by older adults (leading to underestimates). Nevertheless, as text
messaging becomes a more normal part of everyday life, regardless of age,
these potential biases should become less of a concern.
However, the increasing popularity of texting may make it more difficult
for people to estimate the number of messages they send, even on a daily
basis. In his detailed report of text message data collected from Norwegian
teenagers and adults from 2001–2007, Ling (2010) notes that the estimated
mean number of text messages sent daily for the whole sample was only
2.4 in 2001, rising to nearly 8 in 2007 (numbers were higher for teenagers
and young adults). Continuing research shows that reported means are
steadily increasing across time, from a mean of 7 per day in US under-
graduates (Massengill Shaw et al., 2007), to 18 in Australian teenagers and
young adults (De Jonge & Kemp, 2012), to 24 in Australian undergradu-
ates and 40 in Canadian undergraduates (Grace, Kemp, Martin & Parrila,
2012), to 45 in US pre- and adolescents (Cingel & Sundar, 2012), to 60 in
US undergraduates (Drouin & Driver, 2012). Younger children, too, can be
asked to estimate the number of text messages they send per day, for example
the 4–5 per day recorded by British pre-teens (Coe & Oakhill, 2011; Plester,
Wood & Bell, 2008). Accurately estimating larger numbers of daily sent mes-
sages may be especially challenging for children. For example, the Australian
pre-adolescents in Kemp and Bushnell’s (2011) study estimated that they
sent between zero and 315 messages per day. If sending rates continue to
increase, future researchers may need to consider asking participants to report
on the basis of their itemised phone bills, rather than their own estimates.
 Issues in collecting and coding textisms 81

Our main focus is people’s use of textisms in the messages that they send
and, again, a variety of methods have been developed to collect this informa-
tion. The simplest is to ask people whether they use textisms or not (e.g.
Drouin & Davis, 2009). This is useful for identifying individuals who make a
point of writing only in the conventional way. However, it makes it difficult
to distinguish between those who might drop in the occasional tmrw for
tomorrow, and those whose messages are filled with textisms. More detailed
information can be gained by asking participants to indicate their use of tex-
tisms in general, or of particular types of textisms, on a Likert scale with
various numbers of points (e.g. Bodomo, 2010; Bushnell, Kemp & Martin,
2011; Drouin & Davis, 2009; Rosen, Chang, Erwin, Carrier & Cheever,
2010). As with any use of a Likert scale, it can be hard to be sure whether
respondents share similar meanings for responses such as ‘rarely’ or ‘often’,
but we presume that they base these decisions on comparisons with other
messages that they see (which of course may themselves vary widely). Even if
participants are good at recalling and judging the extent of their use of tex-
tisms, it is also important to consider the possibility of response bias. Just as
with the number of text messages sent, participants may differ in their percep-
tions of the desirability of reporting particularly high or low levels of textisms
in their messages. Younger teenagers might think that textisms are ‘cool’ and
over-report their use; young adults – especially those who feel that they may
be being judged by a university-based researcher – might think that the use
of textisms is a sign of immaturity or poor academic skills, and under-report
them. Finally, for many people, the inclusion of textisms in a message is likely
to vary with the intended recipient, as discussed in Chapter 5.
Rating one’s overall textism use might be straightforward for those who
use a certain proportion of textisms fairly indiscriminately in their messages.
However, it could obscure the subtleties of written communication mastered
by others. For example, a young adult might write messages in standard
English to her grandmother, but include a range of common textisms when
writing to her broad social circle, and further idiosyncratic abbreviations with
her two best friends. Researchers do not always have the time or resources to
investigate each question at so many layers of detail, but there are likely to be
differences between people in terms of the way they use textisms in different
situations. These potential differences should be borne in mind when inter-
preting research findings, and when planning future studies.

Message translation
Researchers can also examine textism use in an experimental situation. This
overcomes the potential problems of self-reports, because texting behaviour
is tested directly. It also allows the precise coding and counting of textisms
used, and control over the length and content of messages produced, so that
responses are easily compared. One focused method is to provide a list of
82 Issues in collecting and coding textisms

words that are frequently re-written in text messages, and to ask people to
write them ‘as they normally would in a text message’ (e.g. Coe & Oakhill,
2011; Bushnell et al., 2011). However, translating isolated words could focus
attention on textism use more than simply writing those same words within a
message, and could lead to overestimation of textism use.
Researchers using experimental paradigms can instead ask their partici-
pants to produce whole text messages. The collection method chosen will
depend on the overall aim of the study. If the goal is to compare how (groups
of) participants construct the same messages in different ways, messages can
be presented in standard writing, and participants can be asked to ‘translate’
them into how they would write them as a text message. It may be helpful
to specify that the message should be the type that they would write ‘to a
friend’, to make it clear that the casual, non-standard register is appropri-
ate. This method has been employed with children (e.g. Kemp & Bushnell,
2011; Plester et al., 2008) and with adolescents and adults (e.g. De Jonge &
Kemp, 2012; Drouin & Davis, 2009). Translation tasks do have the potential
to lead to some overestimation of textism use. One reason is that to provide
more scope for the use of textisms, the messages created for translation may
include a preponderance of words and phrases that lend themselves to textism
creation. Another reason is that, no matter how carefully the instructions are
worded, participants might see this task as one of ‘translation into textese’
and use more textisms than they normally might.

Message elicitation
If ecological validity is more important than equal message length or con-
tent, researchers may choose instead to elicit text messages by presenting
participants with scenarios, and asking them to write the message that they
would in that situation. This method has the advantage of allowing more
natural, individual responses, even if it reduces the scope for directly compar-
ing the way that particular words or phrases are written. This method has
been successfully used to elicit messages from children (Coe & Oakhill, 2011;
Plester, Wood & Joshi, 2009; Plester, Lerkkanen, Linjama, Rasku-Puttonen &
Littleton, 2011) and adults (Clayton, 2012). Some participants may still use
more textisms than they normally would, because they are aware that they are
doing a texting task for researchers interested in text language. However, if
an experimental task is to be used, this method of message elicitation through
scenarios provides a reasonable compromise between artificiality of situation
and comparability of messages.

Message production
Whether messages are translated or elicited via scenario, researchers must
decide how the messages are to be produced. The obvious solution would be
 Issues in collecting and coding textisms 83

for participants to type messages into their phones, and to send the messages
to the researchers. This method was employed by Durkin, Conti-Ramsden
and Walker (2011), although these authors required participants to send one
only message, and even then not all of their participants replied (87 per cent
of typically-developing adolescents and 68 per cent of those with specific
language impairment). Unfortunately, however, this method is not always
a practical one. As noted in Chapter 6, when the participants are school
children, they are often not allowed to have their phones at school, even to
participate in a research study. In response to this restriction, some research-
ers have asked children to write down their messages on paper. This solution
is simple, cheap and private, and has provided useful data on the nature of
children’s text messages in both dictated and elicited text tasks (e.g. Plester
et al., 2008, 2009). However, it does raise the question of whether hand-
written messages are fully representative of children’s usual text-writing
habits, as children must imagine the spellings that they would normally cre-
ate with finger- or thumb-movements on a keyboard, and then transfer that
into handwriting.
If participants are instead asked to create messages in a more realistic way,
by typing them into a mobile phone, there are different experimental issues to
consider. If a researcher is focused specifically on how individuals differ in their
texting habits, he or she might provide a phone on which all participants are
asked to type their messages, so as to control for technology-related differences,
such as keyboard type, use of predictive texting and ease of accessing non-letter
characters (Kemp, 2010; Kemp & Bushnell, 2011). If the research focus is
instead on individual differences that also depend on individuals’ technology
use, it makes more sense to ask participants to produce messages on their own
phones, using their own usual entry method (e.g. De Jonge & Kemp, 2012).
Of course, the differences observed in the types of textisms used can depend as
much on the functions of the phone or ease of accessing them as it does on the
text-message writer.

Message collection
Once the messages are composed, they must be transferred to the research-
ers for coding and analysis. The most reliable way of doing this would be to
ask participants to type their messages into a phone and forward them to a
central number. Again, however, the best solution is not always a practical
one. Especially where school children are concerned, ethics committees – as
well as parents and school staff – often have qualms about having children
sending messages from their own phones, for reasons of privacy (in case the
child’s number is learned, although this could be avoided) and cost (in case
the child/parent has to pay for the texts, although this concern is diminishing
with modern plans with unlimited texts). These restrictions have led to some
researchers asking participants to compose their messages on their phones
84 Issues in collecting and coding textisms

and then to write them down verbatim for coding (e.g. Drouin & Driver,
2012; De Jonge & Kemp, 2012; Wood, Meachem et al., 2011). It is always
possible that some elements may be missed or extra elements added in the
translation to writing. However, it is likely that, overall, messages collected in
this way provide a reasonably true picture of their original forms. The experi-
menter checks the messages, and participants know that the whole point of
the exercise is to make sure that the copying is correct. It is even possible to
ask participants to code their own textism use (Cingel & Sundar, 2012), but
the reliability of this method would need testing.
Two of the collection methods – handwriting text messages and copying
them down from phones – were compared in a study by De Jonge and Kemp
(2012). Adolescents and adults translated sentences into text messages (‘as
they would send them to a friend’) via both methods. We found that the
density of textisms used in both message types was virtually identical: for
adolescents, 15 per cent in their handwritten messages and 16 per cent in
their texted messages; and for adults, 14 per cent in both. This suggests that
when there are restrictions on phone use, reasonably representative data can
be gained by asking even adolescent participants to handwrite the type of
messages that they would usually text.
Our participants did use a greater range of textism categories in their hand-
written than their texted messages. Thus, if a researcher’s aim is to consider
more fine-grained categories of textism use, the collection of texted messages
would be more important to achieve. To date, one of the few studies that has
managed to achieve this (and over a sustained period of time) is the interven-
tion study of Wood, Jackson et al. (2011) described in Chapter 4, in which the
children handed in phones that they had used each weekend to the research
team. The team then transcribed the messages from the handsets and copied
down other data on the number of messages sent and received, before clearing
the memory of the phones and recharging them ready for next use. Although
time consuming, this method enabled total confidence in data obtained,
and the phones were used in a naturalistic way. The data collection methods
adopted within this study made it the first study of literacy and textism use
where real, independently-verified data was obtained, not only with respect
to the textism use within messages sent, but also in providing week-by-week
data on the volume of text messaging ‘traffic’ that the children were exposed
to via their handsets.

Naturalistic messages
Of course, the most representative text messages are those that are com-
posed in real life, not those translated or elicited under experimental
conditions. Although naturalistic messages are harder to compare than
experimental messages (because they vary in length, content and recipi-
ent), they do provide a realistic picture of people’s real text-messaging
 Issues in collecting and coding textisms 85

habits. Valuable text-message corpora have been set up by researchers who

have invited people to provide their sent/received text messages for analy-
sis. Such corpora have allowed the exploration of various characteristics of
text messages in a number of languages, including English (Tagg, 2009),
French (Anis, 2007) and German (Bieswanger, 2007). An alternative col-
lection method is to access messages sent in the public domain. This is
easier to do for other forms of computer-mediated communication, such
as social media or online forums, but is also possible in some cases for
text messages. For example, Herring and Zelenkauskaite (2008, 2009) col-
lected a corpus of text messages posted to a public interactive television
programme in Italy. The focus of such studies is usually to analyse the mes-
sages themselves, rather than to link the features of the messages with any
characteristics of the people who wrote them.
In more experimental work, researchers may invite people to provide a set
of text messages (typically five) that they have recently sent (e.g. Bodomo,
2010, Drouin & Driver, 2012; Grace et al., 2012, Veater, Plester & Wood,
2010). Participants are also often asked to answer questions about their texting
behaviour, such as length of time they have been texting or the number of text
messages they typically send and receive per day. This method is more appro-
priate for researchers who are interested in comparing how the characteristics
of text-writing behaviour vary with, for example, texters’ age, experience or sex,
as well as linguistic or cognitive skills. By collecting a sample of messages that
have already been written, researchers gain a much more representative view
of textism use than they would from experimental tasks. To avoid the financial
and ethical concerns noted above, many authors have asked participants to
write down their messages from their phones, but ideally they would be for-
warded to a central number for analysis.

Comparison across methods

Some consideration is needed of how well the self-report, translation and elici-
tation methods represent the way that people normally produce text messages.
Authors of self-report studies do not usually have the opportunity to view par-
ticipants’ actual messages. An exception is a study by Grace et al. (2012) in
which undergraduates were asked to provide five examples of their recently sent
messages, but also to indicate how much they normally used textisms, on a
three-point scale. Participants who reported using textisms ‘none of the
time’ (n = 113) turned out to use an average of 13 per cent textisms in their
naturalistic messages, whereas participants who reported using textisms ‘some
of the time’ (n = 120) or ‘most of the time’ (n = 3) both produced an average of
20 per cent. This suggests that caution should be exercised in drawing strong
conclusions about participants’ own perceptions of their own textism use.
The validity of experimental text message collection techniques can also
be examined by comparing the results to those of naturalistic collection tasks.
86 Issues in collecting and coding textisms

Plester et al. (2011) asked Finnish children to bring in examples of their sent
text messages (written down from their phones), and compared these to mes-
sages elicited by scenario and written down in the classroom, the five scenarios
being based on those used by Plester et al. (2009). The pages of elicited texts
were identified only by participant number, and the sheets were collected,
shuffled and redistributed to other members of each class, ensuring no-one
received texts written by themselves or a person sitting next to them. On the
reverse side of the sheet, participants were asked to write out the text they
would send in reply if they had received each of the first five texts written. The
textism density for the naturalistic text messages was a rather large 48 per cent,
compared to the significantly smaller 33 per cent seen in the elicited messages.
However, only 16 children submitted naturalistic texts, and it may be that
they were the more enthusiastic users of text messaging and textisms.
When Plester and her colleagues compared the density of textism use by
individuals across the three types of messages written, they found that the
density of textism use in the elicited replies was not related to that found
in the writers’ natural texts, although it was to the writers’ earlier elicited
text ratios. There was, however, a strong correlation between the textism
density of the elicited replies and that of the elicited texts to which they
were replies, although written by another person. This indicates that the
texters were being sensitive to the style of the message to which they were
replying, possibly changing their style from their own natural text style.
The putative recipients for the elicited replies were not identified except
by number, so this procedure does not quite address the question of sen-
sitivity to recipient, but these were very interesting findings which call
for replication and checking. One cannot draw firm conclusions, however,
because of the small number of natural texts, and there was also a strong
correlation between textism ratio in elicited texts and in the elicited replies
written by the same children. This suggests a general task effect over the
classroom texts.
A larger comparison of different collection methods was made by Grace
et al. (2012). We asked 86 Australian and 155 Canadian university students
to compose five messages via translation from standard English and five
messages in response to scenarios, and to provide five messages that they
had recently sent. We were interested to see that textism density decreased
with the ‘naturalness’ of the task: students produced significantly greater
textism densities in the translated messages (23 per cent) than in the elic-
ited messages (20 per cent) than in the naturalistic messages (17 per cent).
Moreover, the types of textisms varied with message collection method:
students produced significantly more ‘contractive’ textisms (in which letters
were removed, e.g. wht for what) in translated (11 per cent) than in elicited
(7 per cent) than in naturalistic messages (5 per cent). In contrast, they
produced slightly more ‘expressive’ textisms (in which letters or symbols
were added for expressive effect, e.g. whaaaaat?!) in naturalistic (3 per cent)
 Issues in collecting and coding textisms 87

and elicited (3 per cent) than in translated (2 per cent) messages. Thus,
although experimental techniques such as message translation and message
elicitation are good for collecting messages that can be compared across
participants, they may lead to slight overestimations of the proportion of
textisms – especially abbreviated textisms – that are normally used.
Asking people to provide five of their recent sent messages seems a reason-
able way of collecting a representative sample of their usual message-writing.
However, it is possible that some participants may choose messages which
they think represent them in a favourable light (either with particularly many
or particularly few textisms), or that they may avoid sharing messages of a per-
sonal, intimate or rude nature. A recent study has used an innovative method
to gather all of the messages sent by its participants over an extended period
of time, rather than just a snapshot at a certain time or a set collected over a
shorter period. Underwood, Rosen, More, Ehrenreich and Gentsch (2012)
have provided BlackBerry devices to a group of 175 tenth-grade students,
and are recording all of their sent text messages, as well as emails and instant
messages. One potential concern associated with this ambitious method is
that participants, knowing that their messages were being monitored, might
not write completely naturally. However, an initial assessment of the use of
obscenities and sexual themes (Underwood et al., 2011) suggested that these
adolescents were writing in a similar way to typical unmonitored communi-
cation. Of course, providing phones to all of the participants in a study, and
collecting, coding and analysing the vast number of messages produced is
an expensive and labour-intensive procedure, not financially possible for all
research groups. Nevertheless, it does guarantee the collection of representa-
tive messages, and may become a more common technique as phones and
sending rates become cheaper, and as automated ways of examining messages
become easier to implement.
The study by Underwood et al. (2011) is an ongoing one, and will pro-
vide longitudinal data on the changing nature of text messages by American
adolescents. To date, there has been little longitudinal research in this area,
and thus little chance to draw causal conclusions about, for example, any
links between textism use and literacy skills. One exception is a study by Wood,
Meachem et al. (2011) with 119 eight- to twelve-year-old British children.
As discussed in Chapter 4, the children were assessed at the start and the end
of their school year on their reading, spelling, phonological skills and verbal
IQ, and at these time points they also wrote down the text messages that they
had sent on a given weekend. The authors found that reading and spelling
scores could not predict variance in the use of textisms. However, textism
use at the beginning of the school year predicted spelling scores at the end of
the year, even when statistical controls were made for initial spelling scores,
age, verbal IQ and phonological awareness. This relationship seems to be
mediated by children’s ability to rapidly retrieve the phonology of words on
a picture-naming task. We are currently undertaking a longitudinal study
88 Issues in collecting and coding textisms

of the relationship between textism use and the understanding of grammar

in child, adolescent and adult texters. Longitudinal research is helpful for
answering questions about causality, but it does require more time and
resources than cross-sectional research, and may not be possible for all
researchers.

Counting and categorising textisms

Once text messages have been collected, researchers must decide how to
count and categorise the textisms contained in them. One can count the
number of words of a message that have been subject to re-writing (e.g.
Plester et al., 2009), or the number of changes that have been made in
total, even if there is more than one per word (e.g. Drouin & Driver, 2012;
De Jonge & Kemp, 2012; Plester et al., 2011; Varnhagen et al., 2009).
The latter method captures the scope of textism use more fully, as it differ-
entiates textisms such as Ive (apostrophe missing) and ive (capital missing,
apostrophe missing). The difference between these two counting methods
leads to differences in textism densities of only one or two percentage points
(Grace et al., 2012), but must still be decided on. When calculating textism
densities, researchers must also consider whether to divide the number of
textisms by the number of words in the whole database (e.g. Ling & Baron,
2007; Thurlow & Brown, 2003) or to gain a more individualised score
by dividing by the number of words produced by each participant (e.g.
De Jonge & Kemp, 2012; Grace et al., 2012), or to cite both (Drouin &
Driver, 2012).
A larger question is how to categorise the textisms observed. Research-
ers will use a categorisation system that suits their research question and
the level of detail in which they want to analyse their methods. In some
cases, specific types of changes will be the focus; in others, the goal may
be to place each textism into a category. In the research published to date,
some authors have been interested in identifying how many characters are
added to or deleted from words (Herring & Zelenkauskaite, 2008), while
others have considered how specific phonemes are represented (Frehner,
2008). It is more common to categorise different types of transformations,
but there is no standard way of assigning textisms to categories, and no
standard category naming system. For example, shortened words can be cat-
egorised in various ways. The term ‘clipping’ can be used to refer to any
textism in which part of a word is deleted (e.g. Bieswanger, 2007; Herring &
Zelenkauskaite, 2008), or differentiated more finely. An abbreviation such
as Tues for Tuesday could be labelled a ‘clipping’ (Kapidzic, 2010; Shaw,
2008), a ‘shortening’ (Thurlow & Brown, 2003), or a ‘truncation’ (Anis,
2007). A spelling such as u for you could be categorised as a ‘letter/number
homophone’ (Kapizdic, 2010; Thurlow & Brown, 2003), an ‘abbreviation’
(Ling & Baron, 2007), or a ‘shortening’ (Rosen et al., 2010). Using similar
 Issues in collecting and coding textisms 89

categories makes it easier to compare results across studies, and even if the
category labels chosen are different, giving clear explanations and examples
for the categories used makes comparison easier. As a result of the impor-
tance of this matter, we have shared the textism categorisation system that
we have used in our research (e.g. Plester et al., 2009; Wood, Meachem et al.,
2011; Wood, Jackson et al., 2011) in Appendix B of this book. Like the tex-
tism coding systems used by many other researchers, this system focused on
the idea of ‘textism as alternative/incorrect spelling’. However, in our more
recent research, as mentioned in Chapter 7, we have also devised a textism
coding system which focused on coding of the various types of grammatical
violations or alternative conventions which children and young adults make
when composing text messages. For completeness, we have also included
this coding system as Appendix C.
There is also the question of what kinds of changes to count as textisms.
For example, authors need to decide whether it makes sense to count miss-
ing capitals or missing apostrophes, since some message-entry programs
correct these automatically and some do not. Similarly, there is the question
of whether spelling or typographical errors count as textisms: some could
be one-off typing mistakes, and others could be consistent spelling errors,
neither of which is unique to text messaging. Some researchers have used
published lists to decide if an abbreviation or re-spelling should be counted
as a textism (e.g. Coe & Oakhill, 2011), but this process should be consid-
ered with caution. People may come up with all kinds of spellings that do
not appear on any list, but this does not mean that they are not textisms.
If published lists are to be consulted all, they are probably best used as a
guide to identifying the meaning of textisms unknown to researchers (e.g.
Varnhagen et al., 2009). In general, researchers should be cautious about
drawing strong conclusions about how people create textisms for particular
words, especially if their data come from a relatively limited set of messages.
Words are not transformed into textisms in a consistent way (e.g. Bushnell
et al., 2011; Varnhagen et al., 2009) and that, even within individuals,
the way that words are texted is quite variable, with De Jonge and Kemp
(2012) calculating a mean of 72 per cent (standard deviation 12 per cent)
consistency across adolescent and adult participants.

Sex differences
As seen above, care must be taken in deciding how text-messaging data
are collected and coded, to ensure that the conclusions drawn are justified.
However, it is also worth considering whether the sex of the participants
writing the messages can affect the conclusions made. In the experimental
studies reported here, there has been mixed evidence about whether males
and females differ in their use of texting or textisms. Females have been
observed to produce messages that are longer and/or contain more textisms
90 Issues in collecting and coding textisms

than do males in samples of pre-adolescents in the United Kingdom (Plester

et al., 2009), adolescents in Canada (Varnhagen et al., 2009), and adults
in the United Kingdom (Thurlow & Brown, 2003), Italy (Herring &
Zelenkauskaite, 2008) and the United States (Rosen et al., 2010). However,
other studies have reported no significant sex differences in texting behav-
iour (De Jonge & Kemp, 2012; Drouin & Davis, 2009). However, the large
range of studies reviewed by Baron and Campbell (2010) suggest an overall
difference in the texting behaviours and language used by males and females
that reflect more societal-level differences in choice of conversational topic,
choice of vocabulary, conversational role and expression of emotion. The
methodological implication is that participant samples dominated by one sex
may lead to different conclusions than more evenly balanced samples, and
that care should be taken before generalising results across sexes.

Comprehension of textisms
Much of this chapter, like much of the relevant research, has focused on
people’s production of textisms. However, people do not send text messages
into the void; they address them to recipients who must decipher these mes-
sages. It is therefore also important to consider how well people understand
the textisms they read. The most common method of doing this has been
via written and spoken ‘translation’ tasks. Researchers may provide study
participants with messages written in textspeak, and assess how well they can
translate these sentences back into standard written language (e.g. Drouin
& Davis, 2009; Plester et al., 2009). Alternatively, researchers may measure
how quickly and/or accurately participants can read the messages in their
full spoken form, either aloud (Kemp & Bushnell, 2011) or silently (Perea,
Acha & Carreiras, 2009). Such studies show clearly that it takes longer to
read messages that include textisms than to read messages in standard writ-
ten language, and that mistakes of interpretation are made. However, as
discussed previously, most writers tailor their use of textisms to their recipi-
ents, at least to some extent. Thus, it is likely that the textism disadvantage is
much less in real-life text messages than in those presented in experimental
conditions. Nevertheless, it is interesting to note that any time or effort
saved in writing a textism-filled message may well be passed on to a recipient
who has to struggle to draw out the intended meaning.
Finally, more detailed information about people’s understanding of text
abbreviations can be gained through the use of laboratory-based word decision
tasks. These kinds of tasks can help to establish how textisms are represented
in our mental lexicon, or dictionary, by providing information on what hap-
pens when people encounter a textism. Do they first translate the textism to
its whole-word form, and then access that word’s meaning? Or have many
adults now developed mental representations of the textisms themselves, so
that encountering a textism can itself activate the associated meaning?
 Issues in collecting and coding textisms 91

Ganushchak, Krott and Meyer (2010) asked adult participants to look at

a series of real words, textisms (e.g. gr8) and false textisms (e.g. qr8), and
to decide whether each was a real word or not. Participants took longer to
identify textisms than false textisms as non-words. There were significant dif-
ferences in the electrical activity in participants’ brains (event-related poten-
tials, or ERPs) in the later stages of making these lexical decisions. These
results led Ganushchak et al. to conclude that some well-known textisms do
activate their own mental representations, stored in the mental dictionary, or
lexicon. Berger and Coch (2010) present ERP evidence that in young adults
who use textspeak fluently, the meaning of messages written with textisms is
processed in a way similar to any other second language. In a study in which
participants made lexical decisions after being ‘primed’ by an initial textism
or whole-word phrase, Ganushchuk, Krott and Meyer (2012) found further
evidence that some textisms do have their own unique mental representa-
tions that can rapidly activate their whole meaning. There is much scope for
further studies of this type; studies that can provide more detailed informa-
tion on the way that textisms are represented in the mental lexicon. These
representations will presumably differ according to the age at which textisms
are first encountered, and the frequency with which they are seen, both of
which will also change with time.
In this chapter we have sought to review the different techniques that
have been used to elicit, collect, categorise and interpret textisms in people’s
text messages. Each method has its strengths and weaknesses, and the final
choice will depend on the particular goal of the researchers in each study. We
hope that some of the issues discussed here will help current researchers and
readers to interpret the work already completed in this area, and help future
researchers to plan their studies in a way that allows them to draw the most
meaningful interpretations possible from the data gathered.
Chapter 9

Lessons learned and the future of

texting

As stated at the outset, the purpose of this book has been to collate the
available evidence on text messaging and a range of literacy skills, across
samples of different ages and drawn from different populations, and to
summarise what can and cannot be said about the nature of the interrela-
tionships between text messaging behaviours and literacy in its conventional
(text-related) sense. Clearly there are other ‘literacies’ that mobile phone
use has the potential to impact on (digital literacy, multimodal literacy), but
the scope of this book has been deliberately focused on the more traditional
meaning of literacy as relating to written language skills, because of the
nature of the debates which have been explored in the media.

What can we say?

The story we have described is simple in some ways, but it should not be over-
simplified or misrepresented to suit a particular purpose. So below we lay out
and clarify what can and cannot be said based on the data available so far:

x If adults are exposed to misspelled forms of words this can impact nega-
tively on their own spelling of those words.
x If children are exposed to misspelled forms of words this is unlikely to
affect the accuracy of their spelling of those words in future.
x Exposure to correctly-spelled words improves spelling performance for
adults and children.
x The act of creating incorrect ‘made-up’ spellings does not affect chil-
dren’s or adults’ ability to learn the correct spellings of new words.
x Children who demonstrate the greatest knowledge of text abbrevia-
tions (textisms) also demonstrate better knowledge of conventional
spellings.
x Children who tend to use the most textisms when asked to write a text
message also tend to have the best reading ability.
x Textism use by children who have access to mobile phones appears to
contribute to growth in the development of spelling skills over time.
 Lessons learned and future of texting 93

x Textism use appears to be associated with reading and spelling ability

because of their common links with phonological awareness and rapid
phonological processing abilities.
x Children with specific language and literacy problems do not seem to use
phonetically-based textisms to the same extent as non-dyslexic children,
and tend to use more of the non-phonetic forms.
x Use of predictive text by children does not appear to impact on their
literacy.
x The types of phone used by the children, or the keyboard types used, are
also not related to performance on literacy measures.
x There is no strong evidence to date to suggest that giving children who
do not own phones access to mobile phones for text messaging signifi-
cantly boosts their literacy skills, although the work in this area so far has
been limited in scale and scope, and needs replication.
x When giving children access to a mobile phone, it appears to be better to
do this when they are older rather than when they are younger.
x The results of studies which have looked at literacy skills and texting in
adolescent and adult samples are more mixed, and this is likely to be the
result of differences in the methodologies used across these studies and
the impact of the technology used by these samples on texting speed
and other behaviours.
x It would seem to be better for children to text a smaller rather than a
larger network of friends, if phone use is to impact on their phonological
awareness.
x Understanding of conventional grammar is not clearly associated with
the tendency to make grammar-based errors when texting.
x It is hard to know whether the spelling and punctuation errors reported in
these studies are intentional or accidental, or affected by the intended recip-
ient of the message, as different levels of care and attention are reported to
be applied depending on who may receive the message. These interrelation-
ships need further unpacking.
x There is little evidence that children are ‘addicted to’ mobile phones.
x There has been a wide variety of methodological approaches applied to the
study of texting over the course of the studies reported in this book. The
wide variation in approach contributes to a lack of clarity with respect to
adult populations, and the need for standardised tests and greater meth-
odological care should be an important consideration for future work in
this area.

Outstanding questions
As ever, there remain a number of outstanding research questions which need
to be addressed in future work in this area. The first one, noted in Chapter 7,
is that there is a need for longitudinal work to be conducted in the area of
94 Lessons learned and future of texting

grammatical understanding and development. In fact, it is fair to say that more

longitudinal work which can address questions about how literacy skills change
over time as a result of contact with mobile technology and texting would con-
tinue to make a valuable contribution to our understanding of these important
questions. At the time of going to press, we are completing our own longi-
tudinal study of texting and grammatical understanding in both children and
adults, but more focused longitudinal work which isolates specific populations
and studies participants’ engagement with technology and literacy from a very
young (pre-literate) age would make a significant contribution to this field.
To date, studies have looked at reading, spelling, phonological awareness and
grammatical understanding. These are all integral components of literacy. How-
ever, there is a danger that by defining and studying these processes separately
we are at risk of missing the sum of their parts. That is, many of the media stories
referred to in Chapter 2 talked about the impact of textism use on children’s and
young people’s academic writing. The processes and products of written com-
position draw on (proof) reading, spelling and grammar, as well as writing or
keyboarding speed and related skills. It may be the case that individually assess-
ing the relationships between texting and literacy variables is of limited value,
as the real impact is evidenced in the production of connected text of the kind
assessed by schools and universities and needed in the workplace. The analysis
of how texting impacts on the act of producing a written composition needs to
be the next phase of work in this area.
It was noted in Chapter 6 that there seems to be some evidence to suggest
that children who exchange texts with a smaller network of friends have better
phonological awareness than children who have larger networks. The reasons
for this association are unclear, but we argued that it could be that smaller
networks necessarily limit the amount of variation in the different numbers of
textisms for the same words that children are exposed to. This may prevent
the children from being overwhelmed or disorientated by too much variation
in spelling patterns or conventions, thereby affording a safe environment to
consolidate and rehearse phonological knowledge. However, this suggestion
is a speculative one, and so there is also a need for more work in this area to
test this proposal. Of particular value would be to monitor the text messages
passed between friends in the early days of their phone ownership, tracking
and systematically recording the degree of variation in spelling conventions
to see if they stabilise over time and reduce to a set of ‘agreed’ textisms to be
used within the friendship groups. We assume that some form of standardisa-
tion of spelling patterns would inevitably occur. However, this assumption
may not be the case, and it may not hold, and it may not be true for children
whose texting networks are large or dominated by other children who are less
concerned with the need to be inclusive with respect to texting practices. The
contribution of metalinguistic awareness and the ability to take the perspective
of the message recipient/audience would also be factors which could be use-
fully integrated into future work in the area.
 Lessons learned and future of texting 95

The users of textspeak also, indirectly, demonstrate their awareness

of prosody through its emphasis in the spellings created. For example,
unstressed syllables are omitted from some textisms (e.g. bout or coz, instead
of about and because) and primary stress may be emphasised through the
use of repeated vowels in the appropriate place in the word (absoloooootly).
The coding of prosody-based textisms has not been conducted to date as
these types of textism have fallen under other labels, but there is a study
too to be done based on this type of analysis. That is, sensitivity to the pro-
sodic features of speech is something that is receiving increased research
attention because individuals who have literacy difficulties appear to show
deficits in this sensitivity, and it is believed to be linked to the development
of phonological representations in early readers (e.g. see Wood, Wade-
Woolley & Holliman, 2009). It may be the case that individuals who are
‘prosodic texters’ are also likely to show greater awareness and develop-
ment of phonology, and better literacy outcomes as a result, than children
whose abbreviations lack this characterisation. It may also be the case that
encouraging children to use more prosodic textisms could promote their
awareness of speech rhythm in educationally-important ways.
We have seen that children with dyslexia (Veater, Plester & Wood,
2009) text with as much enthusiasm as other children, but use fewer
phonologically-based textisms, parallel to their frequent difficulties with
phonological processing. We have seen a relationship between texting
and other literacy skills in adolescents with Specific Language Impairment
(Durkin, Conti-Ramsden & Walker, 2011). However, we have not yet
seen any text research with profoundly deaf or seriously hearing-impaired
children and adults, whose development of language must use differ-
ent routes from those with other impairments, whether spoken or sign
language is their first language. Only a few anecdotal examples come to
mind, to the effect that yes, the seriously hearing impaired do text, but
use few textisms and depend on predictive text, but how generally true
might this be? We can ask first whether, then how those with serious hear-
ing impairment text, how their text language relates to their other written
language skills, but also how it relates to what spoken language skills they
develop, and also to their communication through sign. Following Hsu
(2013), we can look at these children’s morphological awareness skills
and relate those to their choice of textisms. To what extent will their texts
reflect written English conventions, casual spoken language conventions,
or the conventions of their particular sign language, depending on their
extent of hearing impairment?
We might also investigate the texting of visually-impaired cell phone
users and its relationship to their traditional literacy skills. Sight is integral
to successful use of the mobile phones that most people have, particularly
smart phones, because there is no tactile feedback to indicate where the
screen is touched. Technological developments were announced in 2012,
96 Lessons learned and future of texting

which could give the visually impaired effective use of smart phones (BBC
News, 2012a, 2012b). An Android operating system application has been
developed to give voice feedback to touch on the screen, enabling users
to send texts. Another ‘app’ has been developed – to be available for both
Apple and Android devices – enabling users to text in Braille on the screen.
If the visually impaired are able to use these phones, questions arise about
the texting implications of this relatively new freedom for the visually
impaired. Will textisms be used? Will they be used as freely with Braille as
with voice feedback, and will the applications support textisms easily? And
what will the educational consequences of this be?

Methodological points for attention

Writing a book of this kind inevitably prompts us to reflect more broadly on
the nature of the work that we have been conducting into this area, and see its
limits. We can see, for example, that, on a methodological point, an oversight
which is common to most work in this area to date is that when coding text
messaging for analysis, we have omitted to record the intended recipient of
each message. We have come to realise what a significant oversight this is, as
writers do take into account their intended audience and adjust their voice and
writing conventions accordingly, and this is especially true when texting, as
discussed briefly in Chapter 5.
We refer back to the Plester et al. (2011) Finnish study here, described in
Chapter 7, because the elicited reply procedure used there is one that could
easily be adapted to elicited texts with any group, especially to test whether
the lack of relationship between natural texts submitted and elicited replies
holds (as it did with the small sample in that study), and whether a relation-
ship between elicited texts and their elicited replies also holds. As Grace,
Kemp, Martin and Parrila (2012) have shown, textism density may be a little
higher in elicited texts than natural ones, so this should be kept in mind, but
the relationship between natural texts, elicited texts and elicited replies with a
larger sample would be a beginning to investigating texters’ sensitivity to the
recipients of their messages.
We need to be able to differentiate between textisms that are used indis-
criminately across contexts, and textisms that the writer deliberately varies
between different recipients to suit the nature of their relationship, the type
of information being exchanged, and awareness of the recipient’s fluency
with particular textisms. As noted in Chapter 1, digital communication allows
more and more people to be drawn into one’s communicative circle, in which
a casual register of writing is often the norm. Some writers – especially those
who do not have such a nuanced understanding of the registers of written
and spoken language – may become so accustomed to this informal style of
written language that they cease to adjust their use of textisms according
to the situation. For example, such writers might be more likely to include
 Lessons learned and future of texting 97

textisms in messages to parents, professors and employers, rather than only

in messages to friends. Perhaps it is the ability to modulate one’s use of tex-
tisms according to the intended recipient which has the potential to better
differentiate more and less literate texters, especially in adolescent and adult
samples. As noted in Chapter 5, people also make judgements about the skill
and effort of text-message writers, based on whether the use of textisms in
the message is perceived to be appropriate for the intended recipient. Thus,
future work should continue to investigate the patterns of literacy skill that
accompany the tendency to vary (or not) the use of textisms in messages to
different recipients. Such work could include examination of the interpretations
that readers make of the use of textisms in messages to different audiences, an
appreciation which may also vary with literacy skill.
Similarly, we need to record the nature of the participants’ normal phone
use and their history of access to the technology, given how rapidly things are
changing and how quickly children are given access to ever more advanced
handsets. Simply analysing data by cohort is not sensitive enough to identify
individual differences in experience with mobile technologies and the use of
the features that various types of handset can afford. As we saw in Chapter 7,
some of the differences observed in individuals’ use of textisms, in terms of
age group, sex and time-point when data were gathered, may stem simply
from the particular features of the technology that participants were using, or
the technology on which they first learned to compose text messages. More
fine-grained analyses of current and previous phone use will help researchers
to tease apart the contributions of personal preference, and the constraints or
facilitations of the technology used, in determining the style of writing people
employ in composing messages.
There is a question over whether there is merit in assessing the person-
alities of texters. For example, there is discussion of (and assumptions made
about) the idea that young people are ‘addicted’ to mobile phone technolo-
gies, and there is some work which has linked students’ tendency to text and
use Facebook during college classes to poorer levels of attainment (Junco &
Cotten, 2012). Similarly (as noted in Chapter 1), different types of phone
user appear to use their phones in response to their need for social distance
or privacy, and this has also been borne out by work which has examined
the interrelationships between linguistic variables obtained from analysis of
a sample of recent text messages. Holtgreaves (2011) found relationships
between extraversion and use of personal pronouns, neuroticism and use of
negative emotion words, and agreeableness and positive emotion words. In short,
the ways in which texts were composed reflected the personalities and the inter-
personal context of the writer. There is therefore scope to examine the extent
to which personality factors such as these may contribute to the patterns of
association between literacy and text abbreviation use, and to consider the
implications of this from an educational psychology (and even motivational)
point of view.
98 Lessons learned and future of texting

Texting as an educational tool?

Texting is a medium which has a great deal of educational potential, and
we would argue there is perhaps more potential than peril inherent in
children’s and young people’s use of texting and text abbreviations. There
is still more work to be done in tracking the development of texting prac-
tices in relation to educational outcomes, but there is also a great deal
of scope for beginning to examine its potential as a safe environment for
educational intervention for at least some young people who struggle with
literacy and the way that it may be taught at school. Given the evidence
linking texting with enhanced phonological processing, we would suggest
that teachers could reflect on how they could incorporate discussions of
and exercises around textism use into their lessons as a fun way of engag-
ing children with phonic work. This may be particularly true for children
who are at the transition between primary and secondary school, where
existing phonic interventions may be perceived as patronising by the chil-
dren who need to engage with them. Some ideas for using texting as a
basis for fruitful educational activity and discourse around phonics and
written work might include:

x Reminding children that some words in English can be spelled in differ-

ent ways but sound the same.
x Confirming that some people write text messages in a way that is not
correct or formal English, but can make it quicker or more fun to
write.
x Showing how many textism spellings represent the word with different
letters, but that make the same sound as the original.
x Practising spelling words in different ways that make the words shorter or
simpler to spell, and use this as the basis for a discussion about the transpar-
ency of English spelling.
x Practise reading words spelled in different ways, in case their friends send
them such spellings.

Mobile phones (and paper-based exercises based on them) may also afford
children a different way to engage in creative writing as a collaborative
activity, in which a narrative is constructed in real time through text-based
conversations between pupils. It represents a genre for enabling children to
become authors without the constraint of worrying about conventions that
perhaps they are less familiar with and therefore may get wrong. In the case
of children with literacy difficulties, mobile phone communication means
that they can engage in writing without worrying about their ability to
spell conventionally, as any difficulties that they might have with this are
concealed from classmates in this medium. There is a barrier of course,
which is presented if they are required to decode the texts of other children,
 Lessons learned and future of texting 99

but careful planning around the sharing of work and collaborative activities
could address this.

Future gazing and concluding comments

In this chapter, we have looked at what we do know, but we have also
looked at many things we do not yet know about the relationships between
the language used in text messages and skills with more traditional reg-
isters of language. Here we summarise the issues that need investigation,
and invite readers to consider if they might be able to contribute to this
ongoing research in any way:

x Longitudinal work on the way literacy skills such as grammatical under-

standing change through contact with mobile technology.
x Investigation of the relationships between texting and the composition
of formal connected text.
x Consideration of the recipients of text messages, and how individual tex-
ting styles may vary with intended recipients.
x In coding text messages, we need to be able to differentiate between
unintentional text errors and intentional text style, discerning typing or
spelling errors from erroneous choice of language register.
x A corpus of data is needed showing textisms used where a more formal
style would be preferable. Anecdotes claim many such intrusions, but a
clear indication of frequency is needed.
x Exploration of individual differences and experience of normal phone use
and how that changes with technological advances.
x Exploration of the relationships between personality and literacy skills
and text style.
x The mapping of how textisms evolve within texting networks of different
sizes.
x Analysis of prosody-based textisms, and their relationship with literacy
among users.
x Exploration of texting among hearing and visually-impaired children
and adults, and relationships of texting with literacy and language
conventions.
x Exploring the role of morphological awareness skills in texting choices.
x Exploring the role of texting as a positive tool in education, as below.

Will text messaging replace ‘traditional’ literacy practices? For us, it seems
unlikely. We see texting as offering a new layer to language use rather than
supplanting standard literacy conventions. The language of texting has
evolved in response to technological developments and opportunities, but
has also outlived the situation which gave rise to its development. That is,
100 Lessons learned and future of texting

it was predicted that textism use would decline as mobile phones enabled
users to send longer messages, and keyboards on smart phones are now
QWERTY-based and therefore resemble a computer keyboard rather than
phone keypad. And yet the use of textisms remains as strong as it always has.
Its social function has outlasted its practical one, and evidences children’s
and young people’s understanding of audience and voice. Its users like it
because it enables them to demonstrate sensitivity to the emotional states of
others in an economical way, for example, through the use of emoticons and
added punctuation. It has now reached a point where it is used as a written
system in its own right to communicate thoughtful messages in ways not so
elegantly achievable in conventional prose. Take for example Nick Davies, an
artist who translated Roland Barthes’ The Pleasure of the Text into textspeak
to consider the ways in which this transforms the message of the original. In
this work Davies notes that:

If Barthes were to see all the fuss surrounding text messaging, I believe he
would wonder what the problem was. In fact, I believe that he would enjoy
seeing one of his creations pushed through such a translation, being cut,
squashed and humiliated in the process. What Barthes work opens up to
me is that we are all both consumers and creators and each other alter the
world we both perceive and inhabit in all that we do. The mythologies and
hierarchies attached to certain things only serve to perpetuate the social
and cultural system that they are part of. This is a big reason why we need
to be diligent of them and also never cease in creating our own.
(Davies, 2011, p. B12)

Similarly, poet Norman Silver has written poetry in textspeak, and has
achieved thought provoking and humorous messages through this linguistic
medium, as in the case of his poem langwij:

langwij
is hi-ly infectious

children
the world ova
catch it
from parence
by word of mouth

the yung
r specially vulnerable
so care
shud b taken how langwij
is spread
 Lessons learned and future of texting 101

symptoms include acute

goo-goo
& the equally serious ga-ga

if NE child
is infected with langwij
give em
3 Tspoons of txt
b4 bedtime
& 1/2 a tablet of verse
after every meal

We hope that in this book we have clarified what we know so far, and perhaps
highlighted the potential of texting rather than reinforcing messages that
appear to have been based on stereotypical views of young, technologically-
literate people. Texting does not appear to harm children’s literacy, and the
nature of the relationship between literacy and texting in skilled readers is
likely to be mediated by a wide range of other factors that we are only just
beginning to examine. Texting is not a problem to be eradicated and textism
use is not an affliction or affectation which ‘should not be tolerated’. And
young people who text are not passive, mindless consumers of text trends
with vulnerable minds, but the architects of a new creative form of commu-
nication which continues to evolve.
This page intentionally left bank
Appendix A
Mobile phone use questionnaire

Please read the questions below and circle the answer that best applies to you, or
fill in the gaps as indicated.

1 How many text messages do you send in a normal day?

None 1 or 2 3–5 6–9 More than 10

2 How many text messages do you receive in a normal day?

None 1 or 2 3–5 6–9 More than 10

3 Do you prefer to text or call?

Text Call

4 Do you text when you are bored? Or have nothing to do?

Yes No

5 How often do you use your phone to browse the internet?

Never Very Now and Less than Every week Every day
rarely again, but once a week
not regularly

6 How often do you use your phone to access social networking sites
like bebo or facebook?
Never Very Now and Less than Every week Every day
rarely again, but once a week
not regularly

7 How often do you carry your phone with you?

All the Most of Only when Rarely Never
time the day I think I
will need it

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
 8 How important is it to you to keep your phone charged, so it is ready
when you want to use it?
Very Quite Not that Not important
important important important at all
9 Do you play games on your phone?
Yes No
10 If ‘yes’, do the games you play have words in them? E.g. do they have
instructions?
Yes No
11 What are the top three activities that you used your phone for yester-
day (or the last time you used your phone)?

12 Do you have a smart phone (i.e. iphone, htc, blackberry, something

posh)?
Yes No
13 Do you spend more time using your phone than you do using your
home computer (if you have one)?
Yes No I spend about the I don’t have a
same amount of computer
time on each
14 Do you spend more time using your phone than you do watching
television?
Yes No I spend about the I don’t watch
same amount of television
time on each
15 Do you spend more time using your phone than you do playing video
games?
Yes No I spend about the I don’t play
same amount of computer
time on each games
16 Do you spend more time using your phone than you do completing
homework?
Yes No I spend about the
same amount of
time on each

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
17 Do you spend more time using your phone than you do reading books
outside of school?
Yes No I spend about the I don’t read
same amount of books
time on each
18 How many hours a week would you say you spend using your mobile
phone?
hours per week
19a Are you able to bring your phone to school?
Yes No
19b Do you bring your phone to school?
Yes No
20 What would be a good thing about being able to bring your phone to
school?

21 What sort of keyboard type does your phone have?

QWERTY Alphabetic Number pad
22 Do you use predictive text when you send text messages? This is when
you type in the first letters of a word, and your phone guesses what
you want to say, e.g. if I typed in tom, the phone would automatically
say tomorrow.
Yes No Sometimes
23 How many of your friends do you text via mobile phone?

24 How many people (other than your friends) do you tend to send text
messages to? E.g. family, cinema tickets.

25 Do you use Twitter on your mobile phone? (Not your computer.)

Yes No
26 Do you use MSN or similar ‘chat’ software more than you send text
messages?
Yes No About the same

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
 27 Would you say that you need your mobile phone everyday?
Yes No

28 Do you enjoy using your mobile phone?

Yes A little I neither enjoy No, not really No, not at all
or dislike it

29 Do you enjoy using smiley faces, cutting down words and using other
ways of spelling?
Yes A little I neither enjoy No, not really No, not at all
or dislike it

30 Do you enjoy making up your own spellings and words in texts?

Yes A I neither enjoy No, not really No, not at all
little or dislike it

31 Do you always understand the textspeak that other people use?

Always Mostly Sometimes Never

32 How do you feel when you leave your phone at home?

I usually It doesn’t I am frustrated I am
don’t bother as I would anxious
realise me have used it as I need it

33 Tick all the things you enjoy doing on your phone:

† Being able to contact my friends and family when I need to
† Being able to send text messages
† Being able to make calls
† Being able to play games on it
† Being able to access the internet on the phone
† Being able to take photographs with my phone
† Being able to make videos with my phone
† Being able to listen to music on my phone
† Being able to personalise my phone (by downloading pictures
and ringtones etc.)
† Having a nice phone that I think looks good
† Not having to worry about my spelling when I write messages
† Something else? (Please list them below.)

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
Appendix B
Coding textisms

Based on: Thurlow, C., & Brown, A. (2003). Generation txt? The sociolinguis-
tics of young people’s text-messaging. Discourse Analysis Online, 1(1). Retrieved
30 November 2010 from: [Link]
[Link].

Textism type Description Examples

Shortenings Word ending is dropped Mon for Monday

bro for brother
Contractions Letters are omitted from middle txt bk for text book
of word pls for please
g-clippings Final g is omitted from -ing ending talkin for talking
shoppin for shopping
Other clippings Other final letters are omitted wil for wll
stres for stress
Symbols Symbols on their own or made @ for at
into emoticons & for and
x for a kiss
emoticons such as ;-), ^_^
Initialisms First letter of each word used in a Lol for laugh out loud
compound word or phrase btw for by the way
Acronyms ‘Official’ initialisms BBC for British Broadcasting
Corporation
Letter/number Letter’s/number’s sound replaces c u for see you
homophones same sound in word l8 for late
Misspellings/typosa Common misspellings there for their
alot for a lot
Non-conventional Incorrect spellings that use nite for night
spellings accurate spelling-to-sound rules bloo for blue
for English
Accent stylisation Words from casual/accented wiv for with
speech style spelled as they sound dunno for (I) don’t know
innit for isn’t it

(Continued)

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
 (Continued)

Textism type Description Examples

Missing apostrophes Contractive or possessive cant for can’t

apostrophe missing Toms for Tom’s
Missing other Missing full stops, question marks, Hi how are you for Hi, how
punctuation commas that would be used in are you?
normal writing
Missing capitals Capital letter missing from start hi there for Hi there
of sentence or proper nouns richmond for Richmond

Note
a When we have calculated textism density in our own research, we have excluded this category as
these are interpreted as unintended errors rather than deliberate and playful alternative spellings.

One convention Thurlow followed was to categorise by the first change made to
a word, for example:

x cnt for can’t should be categorised as a contraction, even if it is also missing

an apostrophe.
x 2moro is a homophone, even if it is also a contraction and a clipping, or 2moz
also an accent stylisation.

Other research might count the number of changes made rather than the number of
words changed. Plester, B., Lerkkanen, M.-K., Linjama, L. J., Rasku-Puttonen, H.,
& Littleton, K. (2011: Finnish and UK English pre-teen children’s text message
language and its relationship with their literacy skills. Journal of Computer Assisted
Learning, 27, 37–48) found multiple changes per word were very frequent in
Finnish texts, so all changes were counted and categorised, even if they did not
agree with the number of words changed.

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
Appendix C
The coding of grammatical errors

As part of our Nuffield Foundation funded work, we (Kemp, Wood, Waldron &
Hart, submitted) developed a system for coding various types of grammatical vio-
lation and variation observed in children’s and young people’s text messages. This
system was developed until it was capable of capturing and representing all the
various types of grammatical variation observed in the samples of text messages
that we were given by the participants in the study. For ease of reference in Wood
et al. (submitted), we then organised these codes into three broad categories:
unconventional orthographic forms, incorrect punctuation and capitalisation, and
word errors. There are, of course, other ways in which the individual codes may be
usually grouped into different categories of violation and variation, but here we
have retained these three as descriptive headings for the tables.

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
 Unconventional orthographic forms

These types of grammatical variation were coded as unconventional orthographic

forms because, although they do not represent standard orthographic conven-
tions, they are not ‘errors’ as such. Rather, they represent a playful creativeness,
adopting new systems for punctuating sentences.

Violation type Examples

Ellipsis …
Start of sentence emoticon :D Hi there!
Start of sentence kiss X love you
End of sentence emoticon - (instead of punctuation)
End of sentence kiss X (instead of punctuation)
End of sentence initialism LOL, LMAO (instead of punctuation)
More than one question mark Are you coming out later???
More than one exclamation mark It was so awesome!!!
More than one emoticon - :D :x (instead of punctuation)
More than one kiss XXX (instead of punctuation)

Incorrect punctuation and capitalisation

This category represents true errors or violations, which often impede under-
standing. The standard rules about when to use capital letters and punctuation
are broken in these examples.

Violation type Example

Mid sentence missing full stop/comma It was ace are you coming out later?
End of sentence missing full stop I am going out later
Missing question mark Are you out later.
i for I i will be out later.
Missing proper noun capitals I am going to see tom tonight.
Missing start of sentence capitals it will be a great night.
Missing contraction apostrophe Im not coming out.
Missing possession apostrophe Robs books are blue.
Unnecessary apostrophe These shoe’s are comfy.

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
Word errors

As with the previous category, word errors represent actual mistakes in the com-
position of the sentence if we read it as a piece of conventional writing (rather
than the representation of a specific style of speech or regional dialect).

Violation type Examples

Missing pronoun/subject Am going out later.

Missing verb I going out later.
Missing function words You want to come with me?
(e.g. do, with)
Missing word endings I am go to school.
(e.g. -ed, -ing)
Missing other campin later (could be:
are you camping later?
I am camping later
I will see you at camping later)
Grammatical homonyms Their going to town too buy sweets.
(e.g. they’re/their, you’re/
your)
Ungrammatical word forms Does you want to go out later?
Word/verb reduction hafta, wanna, gonna

© 2014, Text Messaging and Literacy – The Evidence, Clare Wood, Nenagh Kemp and Beverly
Plester, Routledge.
References

Adams, M. (2009). Slang: The people’s poetry. Oxford: Oxford University Press.
Alonso, E., & Perea, M. (2008). SMS: Social and cognitive impact. Escritos de
Psicologia, 2, 24–31.
Anis, J. (2007). Neography: Unconventional spelling in French SMS text messages.
In B. Danet & S. C. Herring (Eds.), The multilingual Internet: Language, cul-
ture, and communication online (pp. 87–115). New York, NY: Oxford University
Press, Inc.
Associated Press (2007). Irish Government Blames Text Messaging for Teen Illiteracy.
Fox [Link]. Retrieved 20 December 2011 from: [Link]
story/0,2933,268733,[Link].
Baron, D. E. (2011). OED hearts OMG. The Web of Language. Retrieved 8 April
2011 from: [Link]
Baron, N. S. (2008). Always on. Oxford: Oxford University Press.
Baron, N. S. (2010). Attitudes toward mobile phones: A cross-cultural comparison.
In H. Greif, L. Hjorth, A. Lasen & C. Lobet, (Eds.). Cultures of participation (pp.
77–94). Frankfurt: Peter Lang.
Baron, N. S., & Campbell, E. (2010). Talking Takes Too Long: Gender and cultural
patterns in mobile telephony. Retrieved 19 December 2011 from: [Link]
[Link]/cas/lfs/faculty-docs/upload/[Link].
Baron, N. S., & Campbell, E. M. (2012). Gender and mobile phones in cross-national
context. Language Sciences, 34, 13–27.
Baron, N. S., & Hård af Segerstad, Y. (2010). Cultural patterns in mobile phone
use: The case of Sweden, the US, and Japan. New Media & Society, 12(1),
13–34.
BBC News (2012a). App helps blind to send text messages. [Link]
[Link]/news/technology-17105225. Posted 20 February 2012, accessed 29
January 2013.
BBC News (2012b). Smartphone to help blind people goes on sale. [Link]
[Link]/news/technology-18818279. Posted 12 July 2012, accessed 20 January
2013.
Berger, N. I., & Coch, D. (2010). Do u txt? Event-related potentials to semantic
anomalies in standard and texted English. Brain and Language, 113, 135–148.
Bieswanger, M. (2007). 2 abbrevi8 or not 2 abbrevi8: A contrastive analysis of differ-
ent space- and time-saving strategies in English and German text messages. Texas
 References 113

Linguistics Forum, 50. Retrieved 11 May 2012 from: [Link]

edu/salsa/proceedings/2006/[Link].
Binning, D. (2012). Textese aint killin our memory 4 newly learned words. Unpublished
Honours thesis, University of Tasmania.
Blair, B. L., & Fletcher, A. C. (2011). Meanings of cell phones in early adolescents’
everyday lives. Journal of Adolescent Research, 26, 155–177.
Bodomo, A. B. (2010). Computer-mediated communication for linguistics and
literacy: Technology and natural language education. Hershey, NY: Information
Science Reference. Retrieved 20 November 2012 from: [Link]/
sideeg/documents/[Link].
Bradley, J. M., & King, P. V. (1992). Effects of proofreading on spelling: How read-
ing misspelled and correctly spelled words affects spelling accuracy. Journal of
Reading Behavior, XXIV, 413–432.
Brown, A. S. (1988). Encountering misspellings and spelling performance: Why
wrong isn’t right. Journal of Educational Psychology, 80, 488–494.
Bryant, J. A., Sanders-Jackson, A., & Smallwood, A. M. K. (2006). IMing, text
messaging, and adolescent social networks. Journal of Computer-Mediated
Communication, 11, 577–592.
Bryant, P., Nunes, T., & Bindman, M. (2000). The relations between children’s lin-
guistic awareness and spelling: The case of the apostrophe. Reading and Writing:
An Interdisciplinary Journal, 12, 253–276. doi:10.1023/A:1008152501105
Bushnell, C., Kemp, N., & Martin, F. H. (2011). Text-messaging practices and links
to general spelling skill: A study of Australian children. Australian Journal of
Educational and Developmental Psychology, 11, 27–38.
Carter, R., & McCarthy, M. (2006). Cambridge grammar of English: A comprehensive
guide. Cambridge: Cambridge University Press.
Chapman, L. J., & Chapman, J. P. (1969). Illusory correlation as an obstacle to the
use of valid diagnostic signs. Journal of Abnormal Psychology, 74, 271–280.
Cingel, D. P., & Sundar, S. S. (2012). Texting, techspeak, and tweens: The relation-
ship between text messaging and English grammar skills. New Media and Society,
14, 1304–1320.
Cipielewski, J., & Stanovich, K. E. (1992). Predicting growth in reading ability
from children’s exposure to print. Journal of Experimental Child Psychology,
54, 74–89.
Clark, C. (2011). Children’s and young people’s reading today: Findings from
the 2011 National Literacy Trust’s annual survey. London: National Literacy
Trust.
Clayton, J. (2012). Students’ use of and opinions on textisms in text-messages, Facebook,
and emails. Unpublished Honours thesis, University of Tasmania.
Coe, J. E. L., & Oakhill, J. V. (2011). ‘txtN is ez f u no h2 rd’: The relation between
reading ability and text-messaging behaviour. Journal of Computer Assisted
Learning, 27, 4–17.
Crystal, D. (1998). Language play. London: Penguin.
Crystal, D. (2006a). Language and the Internet (2nd ed.). Cambridge: Cambridge
University Press.
Crystal, D. (2006b). The fight for English: How language pundits ate, shot and left.
Oxford: Oxford University Press.
Crystal, D. (2008). Txting: The gr8 db8. Oxford: Oxford University Press.
114 References

Cunningham, A. E., & Stanovich, K. E. (1990). Assessing print exposure and ortho-
graphic processing skill in children: A quick measure of reading experience. Journal
of Educational Psychology, 82, 733–740.
Danet, B., & Herring, S. C. (2007). The multilingual Internet: Language, culture and
communication online. Oxford: Oxford University Press.
Davies, N. (2011). How many wordz duz u no an wot iz sum of dem? In Rol& BRtZ
d PlsUR ov d Txt: A Cultural Experiment into Sociolinguistic Values (translated by
N. DAvEz). Vernacular Publications.
De Jonge, S., & Kemp, N. (2012). Text-message abbreviations and language skills in
high school and university students. Journal of Research in Reading, 35, 49–68.
doi: 10.1111/j.1467-9817.2010.01466.x
Dixon, M., & Kaminska, Z. (1997). The spell of misspelled words: Susceptibility to
orthographic priming as a function of spelling proficiency. Reading and writing:
An Interdisciplinary Journal, 9, 483–498.
Dixon, M., & Kaminska, Z. (2007). Does exposure to orthography affect children’s
spelling accuracy? Journal of Research in Reading, 30, 184–197.
Drouin, M. A. (2011). College students’ text messaging, use of textese and literacy
skills. Journal of Computer Assisted Learning, 27, 67–75.
Drouin, M., & Davis, C. (2009). R u txting? Is the use of textspeak hurting your
literacy? Journal of Literacy Research, 41, 46–67.
Drouin, M. & Driver, B. (2012). Texting, textese and literacy abilities: A
naturalistic study. Journal of Research in Reading. Available online. doi:
10.1111/j.1467-9817.2012.01532.x
Durkin, K., & Conti-Ramsden, G. (2007). Language, social behavior, and the quality
of friendships in adolescents with and without a history of specific language impair-
ment. Child Development, 78, 1441–1457.
Durkin, K., Conti-Ramsden, G., & Walker, A. J. (2011). Txt lang: Texting, textism
use and literacy abilities in adolescents with and without specific language impair-
ment. Journal of Computer Assisted Learning, 27, 49–57.
Ehri, L. C., Gibbs, A. L., & Underwood, T. L. (1988). Influence of errors on
learning the spellings of English words. Contemporary Educational Psychology,
13, 236–253.
Ellison, N. B., Steinfeld, C., & Lampe, C. (2007). Benefits of Facebook ‘friends’:
Social capital and college students’ use of online social network sites. Journal of
Computer-Mediated Communication, 12, 1143–1168.
eschool news (2010). Professors Not ROTFL at Students’ Text Language.
Retrieved 1 July 2011 at: [Link]
professors-not-rotfl-at-students-text-language/.
Frederickson, N., Frith, U., & Reason, R. (1997). Phonological Assessment Battery.
London: NFER Nelson.
Frehner, C. (2008). Email – SMS – MMS: The linguistic creativity of asynchronous
discourse in the new media age. Linguistic Insights. Studies in Language and
Communication. New York: Peter Lang Publishing.
Ganushchak, L. Y., Krott, A., & Meyer, A. S. (2010). Electroencephalographic
responses to SMS shortcuts. Brain Research, 1348, 120–127.
Ganushchak, L. Y., Krott, A., & Meyer, A. S. (2012). From gr8 to great: Lexical
access to SMS shortcuts. Frontiers in Psychology, 3. Available online. doi: 10.3389/
fpsyg.2012.00150
 References 115

Gergen, K. J. (2002). The challenge of absent presence. In J. E. Katz and M. Aakhus

(Eds.), Perpetual contact. Cambridge: Cambridge University Press.
Gertner, J. (2012). The idea factory: Bell labs and the great age of American innova-
tion. New York: Penguin.
Grace, A., Kemp, N., Martin, F. H., & Parrila, R. (2012). Undergraduates’ use of text
messaging language: Effects of country and collection method. Writing Systems
Research, 4, 167–184.
Grace, A., Kemp, N., Martin, F. H., & Parrila, R. (submitted, a). Undergraduates’
attitudes to text messaging language use and intrusions of textisms into formal
writing.
Grace, A., Kemp, N., Martin, F. H., & Parrila, R. (submitted, b). Undergraduates’
text messaging language and literacy skills. To Reading and Writing.
Grinter, R. E., & Eldridge, M. (2001). y do tngrs luv 2 txt msg?. In W. Prinz,
M. Jarke, Y. Rogers, K. Schmidt & V. Wulf (Eds.), Proceedings of the Seventh
European Conference on Computer-Supported Cooperative Work ECSCW ’01,
Bonn, Germany (pp. 219–238). Dordrecht, Netherlands: Kluwer Academic
Publishers.
Grinter, R. E., & Eldridge, M. (2003). Want2tlk? Everyday text messaging. Presented
at CHI 2003, Fort Lauderdale, Florida, USA.
Guron, L. M. (1999). Wordchains. London: NFER-Nelson.
Henry, J. (2002). Delete text message style, say examiners. Times Educational
Supplement, 16 August.
Herring, S. C., & Zelenkauskaite, A. (2008). Gendered typography: Abbreviation and
insertion in Italian iTV SMS. In J. F. Siegel, T. C. Nagel, A. Laurente-Lapole &
J. Auger (Eds.), IUWPL7: Gender in language: Classic questions, new contexts
(pp. 73–92). Bloomington, IN: IULC Publications. Retrieved from: [Link]
[Link]/ herring/[Link].
Herring, S. C., & Zelenkauskaite, A. (2009). Symbolic capital in a virtual heterosexual
market: Abbreviation and insertion in Italian iTV SMS. Written Communication,
26, 5–31.
Hokanson, L., & Kemp, N. (2012). Adults’ spelling and understanding of possession
and plurality: An intervention study. Reading and Writing. Available online. doi:
10.1007/s11145-012-9366-7
Holtgreaves, T. (2011). Text messaging, personality, and the social context. Journal
of Research in Personality, 45, 92–99.
Hough, A. (2010). Rehab clinic for children internet and technology addicts founded.
The Daily Telegraph, 18 March. Retrieved 29 January 2013 from: [Link]
[Link]/health/children_shealth/7467200/Rehab-clinic-for-children-
[Link].
Hsu, J.-L. (2013). Exploring the relationships between text message language and
the literacy skills of dyslexic and normal students. Research in Developmental
Disabilities, 34, 423–430.
Humphrys, J. (2004). Lost for words: The mangling and manipulating of the English
language. London: Hodder and Stoughton.
Humphrys, J. (2007). I h8 txt msgs: How texting is wrecking our language. Daily
Mail, 24 September 2007. Retrieved 1 July 2011 from: [Link]
[Link]/news/article-483511/I-h8-txt-msgs-How-texting-wrecking-language.
html.
116 References

Igarashi, T., Takai, J., & Youhida, T. (2005). Gender differences in social network
development via mobile phone text messages: A longitudinal study. Journal of
Social and Personal Relationships, 22, 691–713.
Jacoby, L. L., & Hollingshead, A. (1990). Reading student essays may be hazard-
ous to your spelling: Effects of reading incorrectly and correctly spelled words.
Canadian Journal of Psychology, 44, 345–358.
Johnson, S., & Ensslin, A. (2007) (Eds.). Language in the media. London:
Continuum.
Jones, G. M., & Schieffelin, B. B. (2009). Talking text and talking back: ‘My BFF Jill’
from Boob Tube to YouTube. Journal of Computer-Mediated Communication, 14,
1050–1079.
Junco, R., & Cotten, S. (2012). No A 4 U: The relationship between multitasking
and academic performance. Computers and Education, 59, 505–514.
Kapidzic, S. (2010). Non-standard features of English used in teen chatrooms. Master’s
thesis, English Department, University of Sarajevo.
Kasesniemi, E.-L., & Rautiainen, P. (2002). Mobile culture of children and teenagers
in Finland. In J. E. Katz and M. Aakhus (Eds.), Perpetual contact: Mobile commu-
nication, private talk, public performance (pp. 170–192). Cambridge: Cambridge
University Press.
Katz, J. E., & Aakhus, M. (2002) (Eds.). Perpetual contact. Cambridge: Cambridge
University Press.
Katz, L., & Frost, S. J. (2001). Phonology constrains the internal orthographic rep-
resentation. Reading and Writing: An Interdisciplinary Journal, 14, 297–332.
Kelland, K. (2008). Textonyms Give Texters a New Language. The Telegraph.
Retrieved 6 February 2008 from: [Link]
oukin-uk-texting-idUKL0537558320080206.
Kemp, N. (2009). The acquisition of spelling patterns: Early, late, or never? In
C. Wood & V. Connolly (Eds.), Contemporary perspectives on reading and spelling
(pp. 76–91). Oxford: Routledge.
Kemp, N. (2010). Texting vs. txting: Reading and writing text messages, and links
with other linguistic skills. Writing Systems Research, 2, 53–71. doi: 10.1093/wsr/
wsq002
Kemp, N. M., & Bryant, P. (2003). Do beez buzz? Rule-based and frequency-based
knowledge in learning to spell plural –s. Child Development, 74(1), 63–74.
Kemp, N., & Bushnell, C. (2011). Children’s text messaging: Abbreviations, input
methods and links with literacy. Journal of Computer Assisted Learning, 27,
18–27.
Kemp, N., Wood, C., Waldron, s., & Hart, L. (submitted). Producing and correct-
ing unconventional grammar: Children’s and adults’ understanding of grammar
in text messages and conventional written language tasks. To British Journal of
Educational Psychology.
Kids Speak Out (2007). [Link]
Posted 30 April, accessed 31 May 2007.
Lee, D., & Gavine, G. (2003). Goal-setting and self-assessment in Year 7 students.
Educational Research, 45, 49–59.
Lenhart, A., Arafeh, S., Smith, A., & Macgill, A. R. (2008). Writing, technology,
and teens. Washington, DC: Pew Charitable Trusts. [Link]/
PPF/r/247/report_display.asp. Posted April, retrieved 3 July 2010.
 References 117

Lenhart, A., Ling, R., Campbell, S., & Purcell, K. (2010). Pew Internet and American
life project: Teens and mobile phones. Retrieved 8 July 2013 from: [Link]
[Link]/Reports/2010/[Link].
Leong, C. K. (2009). The role of inflectional morphology in Canadian chil-
dren’s word reading and spelling. The Elementary School Journal, 4, 343–358.
doi:10.1086/593937
Leppänen, S. & Piirainen-Marsh, A. (2009). Language policy in the making: An anal-
ysis of bilingual gaming activities. Language Policy, 8, 261–284.
Lewis, B. (2012). Are your kids addicted to technology? [Link]
dpp/news/children-addicted-to-technology. Posted 7 November, accessed 29
January, 2013.
Lewandowski, G., & Harrington, S. (2006). The influence of phonetic abbre-
viations on evaluation of student performance. Current Research in Social
Psychology, 11, 215–226. Retrieved from: [Link]
crisp/crisp11_15.pdf.
Lewis, C., & Fabos, B. (2005). Instant messaging, literacies, and social identities.
Reading Research Quarterly, 40, 470–501.
Ling, R. (2007). The length of text messages and the use of predictive texting: Who uses
it and how much do they have to say? American University TESOL Working Papers,
Number 4. Retrieved 1 June 2007 from: [Link]
[Link].
Ling, R. (2010). Texting as a life phase medium. Journal of Computer-Mediated
Communication, 15, 277–292.
Ling, R., & Baron, N. S. (2007). Text messaging and IM: Linguistic comparison of
American college data. Journal of Language and Social Psychology, 26, 291–298.
Ling, R., Bertel, T. F., & Sundsøy, P. R. (2012). The socio-demographics of texting:
An analysis of traffic data. New Media and Society, 14, 281–298.
Los Angeles Times (2009). Why text messages are limited to 160 characters. Retrieved
1 July 2011 from: [Link]
[Link].
Marsh, J. (2004). The techno-literacy practices of young children. Journal of Early
Childhood Research, 2(1), 51–66.
Marsh, J. (2005). Digikids: Young children, popular culture and media. In N. Yelland
(Ed.), Critical issues in early childhood education (pp. 181–196). Maidenhead, UK:
Open University Press.
Massengill Shaw, D., Carlson, C., & Waxman, M. (2007). An exploratory investi-
gation into the relationship between text messaging and spelling. New England
Reading Association Journal, 43, 57–62.
Mindlin, A. (2008). Letting our fingers do the talking. New York Times. Retrieved
29 September 2008 from: [Link]
technology/[Link]?_r=1&th&emc=th&oref=slogin.
Morgan, J. (2011). Why did LOL infiltrate the language? BBC News. Retrieved 8
April 2011 from: [Link]
National Council of Teachers of English (2003). Rest assured – students, teachers,
and language are ‘alive and kickn’. Council Chronicle, May. Retrieved 5 April 2010
from: [Link]
National Literacy Trust (2008). Case study – Rhyme time series. Retrieved 11 July 2011 from:
[Link]
118 References

National Literacy Trust (2011). Practical ideas – Using technology to promote read-
ing: Mobile phones. Retrieved 11 July 2011 from: [Link]
uk/practical_ideas/3179_practical_ideas-using_technology_to_promote_reading_
mobile_phones.
Neville, L. (2003). Cn U rEd dis? The causes and consequences of a ‘text message
language’ in young teenagers. Unpublished undergraduate dissertation, University
of Oxford, Oxford, UK.
New York Times (2011). OMG!!! OED!!! LOL!!! New York Times. Retrieved 5 April
2011 from: [Link]
Nunes, T., Bryant, P., & Bindman, M. (1997). Morphological spelling strate-
gies: Developmental stages and processes. Developmental Psychology, 33(4),
637–649.
NZPA & Smith, M. (2006). Principals Oppose Text Language in Exams. New
Zealand Herald. Retrieved 1 July 2011 from: [Link]
news/[Link]?c_id=1&objectid=10409902.
Ofcom (2010). Children’s Media Literacy Audit. Retrieved 19 December 2011
from: [Link]
medlitpub/medlitpubrss/ukchildrensml/.
Paton, G. (2007). England slides down world literacy league. The Daily Telegraph,
29 November. Retrieved from: [Link]
uknews/1570895/[Link].
Peeters, V. E. (1983). The persistence of stereotypic beliefs: A cognitive view. In R.
P. Bagozzi & A. M. Tybout (Eds.), Advances in Consumer Research (Vol. 10, pp.
454–458). Ann Arbor, MI: Association for Consumer Research.
Perea, M., Acha, J., & Carreiras, M. (2009). Eye movements when reading text messag-
ing (txt msgng). The Quarterly Journal of Experimental Psychology, 62, 1560–1567.
Pew Internet and American Life Project (2008). Retrieved 19 December 2011 from:
[Link]
Plester, B., Lerkkanen, M.-K., Linjama, L. J., Rasku-Puttonen, H., & Littleton, K.
(2011). Finnish and UK English pre-teen children’s text message language and its
relationship with their literacy skills. Journal of Computer Assisted Learning, 27,
37–48.
Plester, B. & Wood, C. (2009). Exploring relationships between traditional and new
media literacies: British preteen texters at school. Journal of Computer-Mediated
Communication, 14, 1108-1129.
Plester, B., Wood, C., & Bell, V. (2008). Txt msg n school literacy: Does texting
and knowledge of text abbreviations adversely affect children’s literacy attainment?
Literacy, 42, 137–144.
Plester, B., Wood, C., & Joshi, P. (2009). Exploring the relationship between chil-
dren’s knowledge of text message abbreviations and school literacy outcomes.
British Journal of Developmental Psychology, 27, 145–161.
Powell, D., & Dixon, M. (2011). Does SMS text messaging help or harm adults’
knowledge of standard spelling? Journal of Computer Assisted Learning, 27, 58–66.
Provine, R. R., Spencer, R. J., & Mandell, D. L. (2007). Emotional expression online:
Emoticons punctuate website text messages. Journal of Language and Social
Psychology, 26, 299–307.
Ramus, F. (2001). Outstanding questions about phonological processing in dyslexia.
Dyslexia, 7, 197–216.
 References 119

Ramus, F. (2003). Developmental dyslexia: Specific phonological deficit or general

sensorimotor dysfunction? Current Opinion in Neurobiology, 13, 212–218.
Ramus, F., Rosen, S., Dakin, S. C., Day, B. L., Castellote, J. M., White, S., & Frith, U.
(2003). Theories of developmental dyslexia: Insights from a multiple case study of
dyslexic adults. Brain, 126, 841–865.
Reid, D. J., & Reid, F. J. M. (2007) Text or talk? Social anxiety, loneliness, and
divergent preferences for cell phone use. CyberPsychology & Behavior, 10(3),
424–435.
Revelle, G., Reardon, E., Green, M. M., Betancourt, J., & Kotler, J. (2007). The
use of mobile phones to support children’s literacy learning. Persuasive Technology
Lecture Notes in Computer Science, Vol. 4744/2007, 253–258.
Roschke, K. (2008). The text generation: Is English the next dead language? Available
online. Retrieved 8 May 2013 from: [Link]
pdf Accessed most recently.
Rosen, L. D. (2007). Me, MySpace, and I: Parenting the net generation. New York:
Palgrave Macmillan.
Rosen, L. D., Chang, J., Erwin, L., Carrier, L. M., & Cheever, N. A. (2010). The
relationship between ‘textisms’ and formal and informal writing among young
adults. Communication Research, 37, 420–440.
Rosenberg, T. (2011). Everyone speaks text message. New York Times. Retrieved
11 December 2011 from: [Link]
[Link]?_r=2&pagewanted=all.
Sacre, L., & Masterson, J. (2000). Single word spelling test. London: GL
Assessment.
Safire, W. (2008). The seamy side of semiotics. New York Times. Retrieved 25 May
2008 from: [Link]
html?scp=1&sq=Safire+The+seamy+side+of+semiotics&st=nyt.
Safire, W. (2009). Abbreve that template. New York Times. Retrieved 24 May 2009
from: [Link]
scp=1&sq=Safire+Abbreve+that+template+&st=nyt.
Shafie, L. A., Azida, N., & Osman, N. (2010). SMS language and college writing: The
languages of the college texters. International Journal of Emerging Technologies in
Learning, 5, 26–31.
Shaw, P. (2008). Spelling, accent and identity in computer-mediated communication.
English Today, 24, 42–49.
Shea, A. (2010). The keypad solution. New York Times. Retrieved 24 January 2010
from: [Link]
html?th&emc=th.
Shortis, T. (2001). The language of ICT: Information and communication technology.
London: Routledge.
Silverstone, R., & Haddon, L. (1996). Design and domestication of informa-
tion and communication technologies: Technical change and everyday life. In
R. Silverstone and R. Mansell (Eds.), Communication by design: The politics
of information and communication technologies (pp. 44–74). Oxford: Oxford
University Press.
Stanovich, K. E., & Cunningham, A. E. (1992). Studying the consequences of lit-
eracy within a literate society: the cognitive correlates of print exposure. Memory
and Cognition, 20, 51–68.
120 References

Stanovich, K. E., & West, R. F. (1989). Exposure to print and orthographic process-
ing. Reading Research Quarterly, 24, 402–433.
Steinhauer, J., & Holson, L. M. (2008). As text messages fly, danger lurks. New York
Times. Retrieved 1 July 2011 from: [Link]
us/[Link]?pagewanted=1.
Stuart, M., Dixon, M., & Masterson, J. (2004). Use of apostrophes by six to nine year
old children. Educational Psychology, 24, 251–261.
Sutherland, J. (2002). Cn U Txt? The Guardian. Retrieved 31 July 2007 from:
[Link]
Tagg, C. (2009). A corpus linguistics study of SMS text messaging. Unpublished PhD
dissertation, University of Birmingham, UK.
Thurlow, C. (2006). From statistical panic to moral panic: The metadiscursive con-
struction and popular exaggeration of new media language in the print media.
Journal of Computer-Mediated Communication, 11, 667–701.
Thurlow, C., & Bell, K. (2009). Against technologization: Young people’s new
media discourse as creative cultural practice. Journal of Computer-Mediated
Communication, 14, 1038–1049.
Thurlow, C., & Brown, A. (2003). Generation txt? The sociolinguistics of young
people’s text-messaging. Discourse Analysis Online, 1(1). Retrieved 30 November
2010 from: [Link]
[Link].
Thurlow, C., Lengel, L., & Tomic, A. (2004). Computer mediated communication:
Social interaction and the internet. London: Sage.
Thurlow, C., & Poff, M. (2009). The language of text-messaging. In S. C. Herring,
D. Stein, & T. Virtanen (Eds.), Handbook of the pragmatics of CMC. Berlin and
New York, NY: Mouton de Gruyter.
Tong, S. T., Van Der Heide, B., Langwell, L., & Walther, J. B. (2008). Too much
of a good thing? The relationship between number of friends and interpersonal
impressions on Facebook. Journal of Computer-Mediated Communication, 13,
531–549.
Underwood, M. K., Rosen, L. H., More, D., Ehrenreich, S. E., & Gentsch, J. K.
(2012). The BlackBerry project: Capturing the content of adolescents’
text messaging. Developmental Psychology, 48(2), 295–302. doi: 10.1037/
a0025914
Varnhagen, C. K., McFall, G. P., Pugh, N., Routledge, L., Sumida-MacDonald, H.,
& Kwong, T. E. (2009). lol: New language and spelling in instant messaging.
Reading and Writing, 23, 719–733.
Veater, H. M., Plester, B., & Wood, C. (2010). Use of text message abbreviations and
literacy skills in children with dyslexia. Dyslexia, 17, 65–71.
Wagner, R. K., Puranik, C. S., Foorman, B., Foster, E., Wilson, L. G., Tschinkel, E.,
& Kantor, P. T. (2011). Modeling the development of written language. Reading
and Writing, 24, 203–220.
Wechsler, D. (1999). Wechsler Abbreviated Scale of Intelligence (WASI). San Antonio:
Harcourt Assessment Inc.
Wilde, S. (1988). Learning to spell and punctuate: A study of eight- and nine-year-old
children. Language and Education, 2, 35–59.
Wolf, M., & Bowers, P. (1999). The double deficit hypothesis for the developmental
dyslexias. Journal of Educational Psychology, 91, 415–438.
 References 121

Wood, C., Jackson, E., Hart, L., Plester, B., & Wilde, L. (2011). The effect of text
messaging on 9- and 10-year-old children’s reading, spelling and phonological
processing skills. Journal of Computer Assisted Learning, 27, 28–36.
Wood, C., Kemp, N., & Waldron, S. (submitted). Grammatical understanding,
literacy and text messaging in children and adults: A concurrent analysis.
Wood, C., Meachem, S., Bowyer, S., Jackson, E., Tarczynski-Bowles, M. L., &
Plester, B. (2011). A longitudinal study of children’s text messaging and literacy
development. British Journal of Psychology, 102, 431–442.
Wood, C., Wade-Woolley, L. & Holliman, A. (2009). Phonological awareness:
Beyond phonemes. In C. Wood & V. Connelly (Eds.), Contemporary perspectives
on reading and spelling (pp. 7–23). Oxford: Routledge.
Woronoff, P. (2007). Cell phone texting can endanger spelling. Retrieved 10
January 2013 from: [Link]
[Link].
Yule, V. (2007). Can literacy be made easier? The Psychologist, 20(4), 212–214.
Index

abbreviations 88, 89; categories 18–19; Davies, Nick 100

as time-saving 9–10 deictic expressions 8
‘absent presence’ 6 dictionaries: on phones 57
academic achievement: texting and 26–9 distance: communication at 1–2
accent stylisations 19–21, 47–9, 107 dyslexia 55; texting by children with 7,
acronyms 18–19, 107 30–2, 95
‘addiction’ to mobile phones 57–60, 65
adolescents: textisms use 41–52 Edison, Thomas 1
agreeableness 97 education: texting as tool in 98–9
apostrophe: use 69, 70, 108 emoticons 13, 71–2
articulatory suppression 55 emotional nuance: in texting 9
AT&T: adverts 15 enjoyment: of mobile phones 60–1,
64–5, 66
Baron, Dennis 15 errors in grammar: coding 109–11; and
Barthes, Roland 100 ignorance 74–7
Bell, Alexander Graham 1 ethics committees 83
Bell Telephone Laboratories 3 event-related potentials (ERPs) 91
Braille: texting in 96 examinations 15–16
extraversion 97
capitalisation 72–3, 108, 110
car phone 3 Finnish: textisms in 21, 44
cell phones (cellular phones) 3; see also focus groups 79
mobile phones French: textisms in 43
cells 3 future gazing 99–101
children: links between spelling, reading
and texting 23–33; see also mobile g-clippings 107
phone behaviours study games: playing on mobile phones 56,
Chinese 21–2, 31 62, 65
clippings 88, 107 gaming 21
CMD 12–13 German: textisms in 43–4
code switching 21–2 Global System for Mobile
Cognitive Abilities Test (CAT) 27 Communications (GSM) 3, 5
Computer Mediated Discourse (CMD) glossaries 18, 50
12–13 grammar 68–78; capitalisation 72–3,
contractions 17, 19, 69, 107 108, 110; coding of errors 109–11;
cordless phones 3 conclusions 77–8; errors and
Critical Discourse Analysis 12 ignorance 74–7; omission of
Crystal, David 7, 17–18 words 73, 111; punctuation 68–72,
 Index 123

108, 110; ungrammatical word message production 82–3; message

forms 74 translation 81–2, 86–7, 90;
Gray, Elisha 1 naturalistic messages 84–6; points
Groupe Spéciale Mobile 3 for attention 96–7; self-report of
GSM 3, 5 behaviours 79–80, 85; self-report
of number of messages 80–1; sex
hearing-impaired people 95 differences 89–90
heart symbol 14–15 misspellings 20, 23–6, 32, 63, 92, 107
Henry, Julie 15 mobile phone behaviours study 53–67;
Herriman, Michelle 16 conclusions 65–7; enjoyment and
heteroglossia 21–2 motivation 60–1, 64–5, 66; levels
hieroglyphs 13, 17 of ‘addiction’ 57–60, 65; patterns
Hillebrand, Friedhelm 5 of texting 55–6; predictive text
homonyms: grammatical 74 use 57; primary vs. secondary
homophones 19–21, 43–4, 49, 88, 107 school comparison 61–2; research
Humphrys, John 14 participants 53–4; text exposure via
phones 56; text messaging behaviour
illusory correlation 39–40 and literacy outcomes 62–5; types of
initialisms 18, 107 technology 56–7
Instant Messaging (IM): punctuation mobile phones: age of acquisition
in 69–70 4, 54, 63–4, 66–7; analysis of use
internet: communication through 6–7 97; arrival 3–4; children’s access
interviews: individual 79 to 39; domestication 4; keypad
Italian: punctuation in text messages in vs. keyboard use 42–3, 49, 57;
70, 71; textisms in 43 questionnaire on use 103–6; social
standards for use 6; as technological
Kelland, Kate 22 intervention 36–8; use by children
35; see also cell phones
language compression 17–18 Moral Panic 13
language play 17, 21–2 Morgan, James 15
languages: preservation of 17 morphological awareness 31, 46, 49,
lessons learned 92–3 75, 95
Likert scales 81 multi-press entry 30, 46, 49; print
literacy: misspellings exposure and exposure via 56
23–6; predictive text use and
29–30; text messaging behaviour National Literacy Trust (NLT) 16–17
and outcomes 62–5 neologisms 3, 12
literacy development: mobile phone use networks of friends: size 94
and 34–9 neuroticism 97
literacy skill: textism use and 44–52 N’Ko 17
longitudinal work: need for 76–7, Nokia 5
87–8, 93–4
Obama, Barack 14
Marks of Distinction 12 orthographic awareness 46
media: text language and 12–18 orthographic exposure effect 24–5
metalinguistic awareness 94 orthographic forms: unconventional 110
methodology 80–91; comparison orthographic processing 55, 62–4, 66
across methods 85–8; Oxford English Dictionary (OED)
comprehension of textisms 90–1; 14–15
counting and categorising textisms
88–9; message collection 83–4; ‘perpetual contact’ 5–7; setting limits
message elicitation 82, 85–7; to 10
124 Index

personalities: of texters 97 smart phones 56, 95–6

phonics 28, 55, 98 social networking 6, 56, 61
Phonological Assessment Battery Spanish: textisms in 43
(PhAB) 54, 55 Specific Language Impairment (SLI)
phonological awareness 27–8, 31, 32, 95
54–5, 64; longitudinal analysis speech: writing versus 7–8
34–6; spoonerisms task measure speech rhythm 95
54, 63 spelling ability: textism use and
phonological fluency 38 27, 63
phonological processing: longitudinal spellings: change in 17; creation
analysis 34–6 of alternative 25–6; misspellings
poetry: in textspeak 100–1 20, 23–6, 32, 63, 92, 107; non-
possession 69 conventional 19, 107
predictive text entry 22; and spoonerisms 54, 63
apostrophes 70; and capitalisation surveys 79
72; literacy and use of 29–30; and symbols: in text messages 71–2, 107
message length 9
prosody awareness 95 telephone: domestication 2–3; rise of
punctuation 68–72, 108, 110 1–3
text language: and media 12–18; and
questionnaire: on mobile phone use texters 18–19
103–6 text messages: arrival 4–5; collection
83–4; elicitation 82, 85–7; length 9;
rapid automatised naming (RAN) naturalistic 46–8, 84–6; production
55, 62 82–3; sending rates 80; translation
reading ability: textism use and 81–2, 86–7, 90
29, 31 textese 13
reading difficulties: textism use and texting: and academic achievement
30–2 26–9; as educational tool 98–9; and
recipients of messages: sensitivity to grammar see grammar; as life phase
96–7 phenomenon 42–3; social effects 8;
research: conclusions 92–3; issues talking versus 7–11
needing investigation 99; textisms 10–11, 12–13; adolescents’
methodological points for attention and young adults’ use 41–52;
96–7; outstanding questions 93–6; appropriateness 50–1; coding 89,
texting as educational tool 98–9 107–8; comprehension 90–1;
ROTFLMAO 13 counting and categorising 88–9;
false 91; frequencies by type 19–21;
Saffo, Paul 14 information collection methods 81;
Safire, W. 17–18 intervention study of use 36–8, 84;
Standard Attainment Test (SAT) 27 in languages other than English
Statistical Panic 13 43–4; literacy skill and use of 44–52;
school: mobile phones in 59 longitudinal analysis of use 34–6,
self-report: of behaviours 79–80, 85; 87–8; as misspelling 24; naturalistic
number of messages and textism use studies 46–8; prosodic 95; reading
80–1 ability and use of 29, 31; reading
sex differences: in texting behaviour difficulties and use of 30–2; social
89–90 nature 51–2; spelling ability and use
shortenings 88, 107 of 27, 63; in student work 16; types
sign language 95 27–8, 107–8
Silver, Norman 100–1 textspeak 12–13; poetry in 100–1; in
slang 10 writing 16
 Index 125

toy mobile phones 4 Intelligence 54

truncation 88 ‘whateverism’ 7
word decision tasks 90
ungrammatical word forms 74 word errors 111
word omission: in text messaging
visually-impaired phone users 73, 111
95–6 word play 10–11
voice feedback 96 writing: speech versus 7–8

Wechsler Abbreviated Scales of young adults: textisms use 41–52

This page intentionally left bank