ArticlePDF Available

Abstract and Figures

This paper examines material from two companies, HSBC and Orange, to reveal how far the 'Tone of Voice' of these brands expresses their different brand positions. Four linguistic techniques are applied: (i) analysis of chains of reference; (ii) analysis of transitivity and participant roles within clauses; (iii) analysis of presupposition and assumption; and (iv) the analysis of features of tenor such as 'conversationalisation' and informal grammar. It is suggested that these approaches are useful in describing the links between language and brand position. Four research hypotheses are proposed: (i) brand strength, i.e. the strength of the reader's impression of the brand varies depending on the number and nature of referring expressions that relate to the brand; (ii) participant roles, i.e. the reader is influenced by the types of processes ascribed to the brand and its related concepts; (iii) assumption, i.e. presuppositions about reader lifestyles, beliefs and values affects engagement with a brand; and (iv) conversationalisation, i.e. greater frequencies of 'conversational' linguistic features influence readers' affinity with the brand and a willingness to associate the brand with 'human' qualities such as 'approachable'. The paper investigates linguistic markers that relate to these hypotheses in detail, but also proposes two further, as yet untested, hypotheses: (v) explicit evaluation , i.e. the impression of a brand is influenced by the linguistic markers of 'point of view' such as those identified within Appraisal Theory, and (vi) unusualness, i.e. there may be a correlation between low-frequency vocabulary items and collocations and readers' tendencies to associate with the brand qualities such as 'quirky' or 'fun'. The paper also presents a discussion of the use of linguistic techniques for commercial aims, which is a departure from the critical tradition in applied linguistics.
Content may be subject to copyright.
 ()  –
 ()  –
L O N D O N
Journal of
Applied
Linguistics
doi : 10.1558/japl.2005.2.1.1
Article
Brand Tone of Voice:
a linguistic analysis of brand positions
1
Judy Delin
Abstract
is paper examines material from two companies, HSBC and Orange, to reveal how
far the Tone of Voice of these brands expresses their dierent brand positions. Four
linguistic techniques are applied: (i) analysis of chains of reference; (ii) analysis of
transitivity and participant roles within clauses; (iii) analysis of presupposition and
assumption; and (iv) the analysis of features of tenor such as conversationalisation
and informal grammar. It is suggested that these approaches are useful in describing the
links between language and brand position. Four research hypotheses are proposed: (i)
brand strength, i.e. the strength of the readers impression of the brand varies depending
on the number and nature of referring expressions that relate to the brand; (ii) partici-
pant roles, i.e. the reader is inuenced by the types of processes ascribed to the brand
and its related concepts; (iii) assumption, i.e. presuppositions about reader lifestyles,
beliefs and values aects engagement with a brand; and (iv) conversationalisation,
i.e. greater frequencies of conversational linguistic features inuence readers anity
with the brand and a willingness to associate the brand with human qualities such as
approachable. e paper investigates linguistic markers that relate to these hypotheses
in detail, but also proposes two further, as yet untested, hypotheses: (v) explicit evalu-
ation, i.e. the impression of a brand is inuenced by the linguistic markers of point
of view such as those identied within Appraisal eory, and (vi) unusualness, i.e.
there may be a correlation between low-frequency vocabulary items and collocations
and readers tendencies to associate with the brand qualities such as quirky or fun.
e paper also presents a discussion of the use of linguistic techniques for commercial
aims, which is a departure from the critical tradition in applied linguistics.
K: ,  , , ,  ,
 ,  
Aliation
Enterprise IG, London & University of Leeds
Correspondence: Judy Delin, Centre for Translation Studies, School of Modern Languages and Cultures,
University of Leeds, LS2 9JT, UK.
email: j.delin@leeds.ac.uk
  . : –
©,  
2    
1 Introduction
is paper is about brands and the way they are expressed linguistically to
potential consumers the language that is used in everything from advertising
to letters and forms. In the developing eld of corporate Tone of Voice, in
which specic language styles are created in an attempt to convey to a range of
audiences the brands ‘personality’, techniques from linguistics can be applied
rst to describe, and then to prescribe, the language of brands. e paper
focuses on one channel for brand expression that service brands such as tele-
communications companies and banks are using with the aim of attracting and
informing customers: the leaets that are distributed in branches and shops that
describe new and existing services. ese are but one channel that attempts to
communicate a distinctive ‘voicefor brands, but the research method described
here is applicable to a large range of written genres.
e overall aim of this research has been to put in place the foundations
for examining the relationship between the linguistic features of text and the
characteristics of ‘brand essence’ that are in common use in the commercial
world. For example, one company may wish to project its brand as reliable,
professional, and dynamic, while another may wish to be fun and easy. It is the
ultimate aim of the research programme of which this report forms a part to
see which linguistic phenomena can reliably be connected to such qualities,
and whether, once a company feels that these qualities have been ‘written in
to their materials, they are consistently appreciated to be there by consumers.
In this rst exploratory study, rather than seek to work with specic brand
qualities, I have used a more general idea of brand positionthat locates
brands within a framework, the Enterprise IG Brand Analytics continuum,
which categorises brands as ranging from ‘Tangibleto ‘Intangible. e aim,
then, has been to see whether two very dierent brands – which Enterprise
IG understand
2
to be located at either end of the continuum do indeed show
dierences in their language. In order to detect whether such dierences are
present, I have used some ideas from mainstream linguistics and applied
linguistics to analyse a small sample of language and rm up four hypotheses
for further research.
In what follows, I rst describe the data for the study, and describe the
relationship between the language of these texts and other available resources
for meaning-making, such as typography and graphics. I then go on to
look more closely at the notion of brands and branding, and the relation-
ship between brand and language in the commercial world. In Section 3, I
present four linguistic approaches to the selected texts: chains of reference,
participant roles, presupposition and assumption, and the features of tenor
that are used to construct social relationships with readers. Finally, in Section
.  3
4, I discuss conclusions, and propose some further linguistic techniques that
would also be of use for examining the relationships between brand position
and linguistic expression.
1.1 Customer leaets for service brands
is paper concentrates on the brochures and leaets made available to existing
and potential customers by service industries such as mobile phone companies
and banks. ey encompass descriptions of accounts, facilities, and services,
and are available to pick up in shops and branches. As such, they are used
both by potential customers choosing an account or provider, and by existing
customers looking to extend their use of existing facilities or take out a new
service, such as a dierent account, insurance, on-line banking, or loans.
e two companies I have selected to examine are the telecommunications
provider Orange and the bank HSBC. ese two brands were chosen because
of their dierent ‘brand positions’: Orange positioning itself as an accessible,
friendly, ‘peoplebrand, and HSBC as a solid, reliable but more impersonal one.
While the texts examined are derived from two dierent sectors, the genre of
their introductory materials is similar.
e texts that were analysed were an extensive 20-page brochure from Orange
describing its services, What do you want to do? and three shorter leaets avail-
able from branches of HSBC: Bank account; Accessible banking, whatever your
needs (about TV and PC banking); and Starting your own business. e sample
of language is therefore a small one, but sucient to generate hypotheses for
further research on the relationship between language and brand positioning.
My own work as a professional consultant in commercial Tone of Voicehas
involved developing not just the copy for such documents, but the underlying
strategy about how a company’s language should in general reect their chosen
brand position. I have not, however, been involved in the development of either
copy or strategy for any of the brands referred to in this paper.
I am not making any claims that these brochures and leaets are privileged
in terms of their power to establish brands in the minds of potential custom-
ers they are just one of many types of document that companies use in
an attempt to do so. It might be argued that leaets and brochures are, in
practice, ignored a lot more frequently than they are read or used. Apart from
specic studies undertaken by individual companies, and as such subject to
commercial condentiality restrictions, there is no research available on the
consumption of such information or on its impact. However, the increasing
number of companies that devote time and attention to extending, refreshing,
and rewriting such material indicates that they are felt to be commercially
important, and that they are a valuable site for expressing brand positions. To
4    
take two recent examples, in January 2006 the Chelsea Building Society, one of
the top six in the UK, appointed a branding agency to undertake a review of its
brand position and to give a more consistent look across the brand’s print and
digital collateral’, aiming ‘to give more cohesiveness to the brand, developing
a literature system and reviewing material’
3
. e following week, Design Week
reported that Argos, a catalogue retailer, had commissioned a consultancy
to ‘look at ways of interpreting the Argos brand across its entire portfolio of
printed collateral, including point-of-sale items, in-store elements, catalogues
and promotional literature … to develop more of a personality for the brand
and create a cohesive tone of voice.
4
e genre of leaets and brochures is a shiing one, and crosses the bound-
ary between marketing and information. In fact, as Varey (2002: 4) suggests,
there is no such clear boundary: One part of marketing communication is
concerned with eectively and eciently providing information about the
business and the products to chosen customer groups. Indeed, some deni-
tions of advertising appear very similar. Vestergaard and Schroder (1985:
2) quote Harris and Seldons (1962: 40) denition of advertising as a public
notice designed to spread information with a view to promoting the sales of
marketable goods and services.is view would suggest similarities between
advertising and marketing, to the extent that they would appear to share a
commonality of purpose sucient to classify them within the denition of
the same genre, according to Swales’ (1990: 58) denition of genre as ‘a class
of communicative events which share some set of communicative purposes.
However, as Cook (2001: 7) points out, genres oen merge into each other
and defy exact denition.
ere is, therefore, signicant overlap between the documents described
in this research and the genre of advertising, but this potentially adds to the
power of the approach: if a branded language analysis and synthesis of its
principles is sucient to capture language choices in marketing materials,
it may also be a fruitful approach to apply to advertising materials relating
to the same brand. Many of the observations made about the language of
advertising texts can provide pointers for the analysis of branded language
(see for example, Cook, 2001; Myers, 1994; Vestergaard & Schroder, 1985;
Goddard, 1998; Delin, 2000), since both have in common the desire to express
a particular feeling’ or ‘personality’ behind brand or product, to persuade
potential customers to position themselves and the product in certain aspira-
tional relationships (cf. Fairclough, 1989: 202), and to call people to action.
In what follows, I will be looking in some detail at word, text and clause level
phenomena that reveal both commonalities and dierences between advertis-
ing and branch-level marketing materials. is, however, is not the primary
aim of the research: what I am seeking to do is interrogate how specic brand
.  5
positions do, or could, inuence language choices made by producers of
such materials. e research reported here is intended as a starting-point for
generating hypotheses about the links between brand position and language
style and form, and as such has a wide application beyond the current data
to all the written materials generated by service industries.
1.2 Language and multimodality
As van Leeuwen and Kress (1995: 25) have pointed out, All texts are multimo-
dal.While this paper focuses specically on language, it is important to note
that other modes of communication do play a role in creating the meaning of
the kinds of texts that we will be looking at: in particular, images, typography,
and layout. As Cook (2001: 4) points out in his discussion of the context of
advertising, other modes such as pictures and paralanguage also play a role, as
do the kinds of texts that readers perceive to be part of the same co-text (for
example, the relationship between a leaet, a TV ad, and a set of posters in a
branch all relating to the same bank).
e leaets and brochures analysed for this study contain images, dia-
grams, and tables. ey are printed in colour and use a range of typefaces and
typographical styles, including bold, italics, caps, and underlining. ey are
structured by headings and punctuated by quotations. All of these are not only
important in creating the meaning, purpose and eect of the document, but
are the most prominent means of communicating the identity of a brand (see
for example, Floch, 2000, on the creation of brand visual identity, and Allen
& Simmons, 2003, on the relationship of visual and verbal identity, among
many others).
ere is as yet no theory extant that is able to articulate the connection
between a particular brand position and the multiplicity of visual and graphical
resources that brand designers choose to express it. is is a fascinating area
that is likely to benet a great deal from the combined perspectives of research-
ers and practitioners working on advertising (cf. for example Cook, 2001, who
makes extensive comment on both visual and verbal elements), on typography
(e.g. McLean, 1980; Walker, 2001), on information design (e.g. Mijksenaar,
1997; Joshi, 2003), on the relationship between documents, information, layout,
and typography (e.g. Waller, 1987; Delin et al., 2002) and, of course, on language
and semiotics in multimodal documents and environments (e.g. Kress & van
Leeuwen, 1996). In addition, some useful recent work has focused on the visual
images used in corporate contexts (see Koller, 2004).
While not debating the importance of the multimodal communicative
resources employed in documents I have studied, the purpose of the current
research has been to generate hypotheses that attempt to relate language in
6    
particular to specic brand positions. In some cases, I make reference to the use
of colour and typography, but it is beyond the scope and intent of the current
work to go further into the operation and signicance of modes of meaning-
making beyond the linguistic. It is the aim of the study presented here to address
in particular how brand values are communicated linguistically, with a view to
situating such research in the wider semiotic context.
2 Brands and brand Tone of Voice
2.1 What is a brand?
According to Varey (2002: 152), a brand ‘is dierentiated from a commodity by
having associated with it, as an integral part of the total oering, values that are
signicant to the consumer and buyer.ese values oer a larger perception of
worth, and therefore a higher prot margin, than the unbranded commodity
would have done. De Chernatony (1993) suggests some of the elements that
appear to consumers to add this value, including trust that the product will
do what it says, assistance in recalling and evaluating the commodity easily
by using the brand as a shorthand, and the symbolic role that the brand plays
in building the identity of the buyer and appearing to help them achieve a
desirable self-image.
Many writers on brand agree, however, that creating these perceptions of
brand value in the minds of consumers is not at all straightforward. As Varey
(2002: 154) suggests, meanings and uses of a brand cannot be xed by the
provider. Jeremy Bullmore of WPP Group
5
describes the process of mental
brand-building as somewhat disorganised: ‘People build brands as birds build
nests: from the scraps and straws they chance upon(Bullmore, 2001: 5). On
this view, brands are personal: a network of concepts and associations derived
from all kinds of thoughts and encounters, from things that the brand owner
seeks to convey (such as through advertising and packaging) to what is read
about brands in the media or heard from friends. e network of concepts that
constitute one individual’s version of a brand may be anywhere from entirely
absent if they have never heard of the brand through semi-developed on the
basis of a few encounters, to highly intricate and developed, as is the case, for
many, with ‘superbrands’ such as McDonalds and Coca-Cola, or old faithfuls
such as Marmite and Fairy Liquid (a dish-washing liquid) in the UK.
e importance of the scraps and strawsview of brands is that, as Bullmore
has said elsewhere, every brand encounter counts(Bullmore, 2000). is
means that both good and bad experiences contribute to an individual notion
of a brand. While brand owners may strive to inuence these encounters,
.  7
they cannot keep control of them all. is is why, on Bullmores view, brands
are out of control: they are what people make of them. For this reason, he
stresses the importance of recognizing that every corporate action, every
corporate decision, every corporate communication will be seen as a clue as
one of those all important scraps and straws from which people build brands
(Bullmore, 2001: 5).
We can derive from this discussion a perception of brands not as static objects,
but as sets of qualities that a provider seeks to project and which are interpreted
by potential consumers both dynamically and partially. e appreciation is
dynamic because it changes constantly and varies a great deal between consum-
ers, and partial because the match between the intended values and the actual
values perceived is not complete. Brand communications – visual, verbal, and
physical are avenues through which providers can attempt to express intended
qualities, but consumersencounters with the dierent channels through which
brands are communicated, the attention they pay them, and the meanings they
attribute to them are not under the control of the provider.
2.2 Product and service brands
Much of the attention that has been paid to brands in the branding literature
involves product brands such as Nike, Adidas, Coca-Cola, IBM, and Apple (see
for example, the collection of essays by Floch, 2000, on ‘industrial semiotics’).
is paper, by contrast, focuses on service brands such as banking and insur-
ance, mobile (cell) phone and other telecommunications brands, and domestic
energy (gas, electricity) brands. As branding guru Wally Olins (2003: 75) points
out: ‘Product brands are about products. Service brands are about people.’ e
reason that service brands are potentially of more interest to linguists is that
the experience of the brand is almost completely constituted in acts of com-
munication. As Rob Waller and I have argued elsewhere (Waller & Delin, 2003),
while the experience of a product is constituted substantially in other activities,
such as the drinking of the drink or the wearing of the shoes, the experience
of a service lies in the kinds of interaction that the consumer has with its
human representatives – in using the bank, in making the insurance claim, in
ordering the delivery. is relationship is not only constituted in face-to-face,
telephone and internet encounters, but in documents such as the bills, letters,
statements, brochures, and forms that remotely enact the relationship between
the consumer and the brand. is means that the communications surrounding
service brands are an important site for their expression.
8    
2.3 Brand values and brand position
All kinds of brands, whether product or service, have in common the fact
that they are developed to achieve a particular brand position: to appeal to a
particular audience for a particular purpose. To this end, most brands at least
to their owners are motivated by some explicit statement of their position:
a set of characteristics, traits, or behaviours that expresses what is unique or
distinctive about the brand. Table 1 shows the brand values for TXU Energi
6
,
an electricity company trading in the UK until 2003, when it was merged
with Powergen. Table 2 gives the brand values of Orange, the global mobile
telecommunications company.
TXU Energi values Elaboration
Straight Talking They know what they are talking about, speak my language, get to
the point and make things happen.
Energy Experts Plus
Whatever they do you can trust TXU to get it right and they make
it all so easy.
Switched On They’re a bright group of people who use their expertise for my
benet.
Customer Champion TXU goes the extra mile for me. They know what I need and deliver
it in the way that I want it.
Table 1: TXU Energi brand values
Orange values Elaboration
Honest We are always open and honest. We say what we do and we do
what we say. We have nothing to hide and we behave responsibly.
Straightforward For us, clarity comes through simplicity. We recognise that we are
people communicating with other people. We are always direct and
easy to understand.
Friendly We enjoy working and succeeding together by building close
relationships. While we have a sense of purpose, we also have a
sense of humour. We consider the needs both of our customers and
of each other.
Refreshing We constantly look to do things dierently and in a better way. We
give colour to all that we do. We are ready to push the boundaries
and take risks.
Dynamic We want to make a dierence to people’s lives. Our optimism is
contagious. We are passionate about what we do and we have
condence in ourselves.
Table 2: Orange brand values
.  9
Other brand values that occur frequently in service brands are human, approach-
able, warm, clear, friendly, professional, expert, and committed. Less frequently
present are fun, courageous, vital, agile, and passionate. Not all brand values are
adjectives (see TXU’s, above) and not all are single words. Some are mixtures of
dierent grammatical categories: for example, First Direct (a UK online bank)
has respect, right rst time, responsive, contribution, openness, and kaizen (this
last being a Japanese word meaning continuous improvement’). e idea of
values such as all these is that they govern everything the brand does – not just
in communications, but in the way it operates all of its business processes.
As an example of the way branding agencies seek to describe, analyse and
reposition the brands they work with, Figure 1 shows a diagram of Enterprise
IGs Brand Analyticsframework, which is used by the agency to locate brands
along a continuum stretching from tangibleon the le to intangible’ on the
right (for an overview of this and other descriptive tools in current use by
agencies, see Michel and Ettenson, 2005).
Figure 1: Enterprise IG’s ‘Brand Analytics continuum
Tangible brands are those whose oer to customers is focused on their own
infrastructure – for a phone company, this could be the coverage of its mobile
network; for a bank, it could be the number of branches and ATMs they have.
It is clear that this is a very dierent oer from one at the ‘intangibleend of the
scale, in which the brands position is based on a ‘mission. For example, a phone
company could attempt to communicate that they are passionate about keeping
people in touch with each other’; or a bank could try to express a desire to help
customers ‘achieve their dreams. In terms of brand values, brands towards the
right of the scale may have more abstract values such as ‘loveand ‘passion,
while brands on the le of the scale may have more concrete values such as
reliability, accessibility, and ease of use. In fact, in their analysis of the brand
Tangible Intangible
What you
have
What you do How you do it
Who you are Why you do it
infrastructure products and
services
approach and
process
people and
values
purpose,
mission, cause
10    
language of Nordea Bank and Bang and Olufsen, Henriksen et al. (2004) have
replaced ‘Tangible’ with ‘Concrete’ and ‘Intangible’ with ‘Abstract’.
Brand positioning and brand values are typically contained within a deni-
tive statement of the brand, a ‘brand handbook. is is likely to have been
produced either by the brand owning company itself, or by a brand consultancy
(well-known examples in the UK are Interbrand, Enterprise IG, Futurebrand,
and Wol Olins). e brand handbook is the denitive resource for making
sure the brand ‘stays put’ in terms of how its employees and agencies develop
all branded communications, including packaging and advertising. As well as
describing the main traits of the brand, the handbook usually denes the visual
elements of the brand in detail: permissible typefaces and when to use dierent
families, sizes, weights, and styles of type, how the logo should look, what
colours should be used and in which combinations, how images (photography,
line drawings) should be selected and used, and design guidelines for how to
produce the kinds of documents that the brand uses most oen, such as posters,
brochures, and so on.
2.4 Brand Tone of Voice
e idea of brand Tone of Voice has surfaced relatively recently as a means of
communicating the distinctive nature of a brand.
7
e aim of a brand Tone
of Voice is to make sure that the values, personality, or essence of the brand
is uppermost in every situation in which people come into contact with the
brand’s language. e term Tone of Voicehas come to be used commercially to
refer to the language styles or registers that a company uses to express a distinc-
tive personality or set of values that will dierentiate its brands from those of
competitors. Somewhat confusingly, Tone of Voiceis becoming established
as the term of choice to refer mainly to written language, although it may also
be applied to speech. Other terms, such as verbal identity, brand language and
language identity are also regularly used.
Most service companies perceive the most important set of contacts or
touchpoints’ to be those that customers see and hear: phone calls, face-to-face
encounters, letters, bills, statements, brochures, marketing materials such as
leaets, web sites and other interfaces such as phone menus, forms, packaging,
instructions, and all forms of advertising. However, many are also interested in
internal communications – those that take place between sta, or between sta
and others who are not end-consumers, such as retailers, organisations to whom
work is outsourced, and other company stakeholders. e idea here is that what
is done internally will aect external perceptions, and sta who are fully engaged
with the brand can be expected to adopt ‘brand behaviours’ as routine.
It is worth noting that the aim of a Tone of Voice is not just to sell things.
is is obviously so if it is used across a wide range of non-selling communi
-
.  11
cations. In fact, the adoption of Tone of Voice is being investigated not just
by commercial companies, but by charities and public sector bodies such as
government departments. e aim of the Tone of Voice is to engage people
with messages perhaps about products and services they may wish to buy,
but also about benets and services that they are invited to take part in or
claim, such as pensions, tax credits, or advice, health services and screenings,
safety information, and more. Tone of Voice dierences can inuence the most
mundane of communications such as forms, as the following examples from
the Tone of Voice guidelines for a government department show:
Don’t use Do use
apportioned accordingly worked out accordingly
expected connement the date your baby is due
bona de genuine
undertake to agree to
at such time as when
commence start
Extensive jargon busterssuch as these are frequently found in guidelines on
corporate Tone of Voice, particularly for applications such as forms, letters,
and information leaets. It is important to note, though, that avoiding jargon
or over-formal language is not all there is to a Tone of Voice. For this particular
government department, consumer research that I undertook as part of a com-
mercial contract
8
showed that many people were perfectly happy with some of
the more technical language that was normally used in relation to the subject
matter which meant that some plainor clear’ English terms were seen as
over-simple and possibly patronising. People preferred complete to ll in, receive
to get, consecutive to in a row, inform and state to tell, and deduct to take o or
take away. What was apparent was that they did not wish to see obvious ‘work-
arounds’ in over-simple language to replace what was already understandable,
and attitude research to the language revealed that they thought the simplest
terms were just too friendly for a government agency and therefore came
across as false. For the same reason, consumers did not want to see contractions
such as youre and well, and preferred please call us to give us a ring. ey liked
you are likely to be entitled to a further amount, but found you can get more money
actually oensive, because they felt that they were being addressed as if they were
not only childish, but greedy. Given the nature of the ‘brand personality’ in this
case, people preferred language that was ‘direct, but not too chatty’, ‘direct and
accessible, but still business-like, and ‘not too pally’.
It is important to note, therefore, that the development of a Tone of Voice
is not the same enterprise as writing in Plain English, the logical conclusion
of which is that all brands sound the same. Because Tone of Voice is aimed at
12    
communicating a personality that is accessible and liked by consumers, the
clarity of the language chosen is not the only concern.
3 Linguistic analysis of brand positions
In this section, I apply some linguistic frameworks with a view to elucidating
how far the language of the two brands is dierent, and whether this com-
municates each position as more ‘tangible’ or more ‘intangible.
ere are many possible analytical frameworks to apply, but I have selected
four which have already borne fruit as a means of analysis of other genres and
which appeared intuitively likely to reveal something of interest in the docu-
ments of study. ese are:
1. Chains of reference, examining how references to the brand and closely
associated elements are made throughout the text (cf. Halliday & Hasan,
1976; Prince, 1981; Cook, 2001);
2. Participant roles, examining how the brand, its services, providers and
audiences are ‘positioned’ in relation to one another through the gram-
mar of the clause (cf. Halliday, 1994; ompson, 1996; Fairclough, 1989;
Clark, 1992; Delin, 2000);
3. Presupposition and assumption, looking further at how audiences are
positioned in relationship to the brand through the kinds of implicit
assumptions that are made about participants (cf. Vestergaard &
Schroder, 1985; Fairclough, 1989; Delin, 2000; Cook, 2001);
4. Tenor, looking at how social relationships are assumed or constructed
through the texts by means of markers of social distance such as formal-
ity or informality (cf. Halliday & Hasan, 1989; Myers, 1994; Goodman &
Graddol, 1996; Fairclough, 1992);
In what follows, we will look at these approaches in turn, showing what each
reveals about the way language is being used by Orange and HSBC and its
possible relationship with the position of each brand.
3.1 Chains of reference
e rst linguistic tool examines chains of reference in the text to see how the
brand and concepts related to it are evoked and kept activated in the mind
of the hearer/reader. e examination of chains of reference has long been a
foundational means of description of text cohesion (see in particular Halliday
& Hasan, 1976; Halliday, 1994: 310) and has also been discussed by Cook (2001:
151) in relation to advertising texts.
.  13
What is important for the current purposes is not so much that texts are
cohesive, but that the appreciation of this cohesion depends on the ability of the
reader to retrieve the relationship between semantically-underspecied referring
expressions such as pronouns, and denite referring expressions which assume
shared knowledge, with concepts that are currently activated in short-term
memory. is analysis looks at how brand concepts are referred to, activated, or
allowed to decay, by the choice of dierent types of referring expressions.
e key elements in this process are the noun phrases in the text, whose con-
tent and form can serve either to strongly evoke a brand (which is the concept
we are interested in), reinforce it, or not evoke it at all. I developed the categories
shown in Table 3 for application to branding from seminal frameworks in the
linguistics and psychology literature (e.g. Halliday & Hasan, 1976; Prince, 1981;
Gundel, Hedberg & Zacharski, 1993). e table shows seven dierent relations
that can hold between a noun phrase and a brand concept, and predicts how
strongly the brand will be evoked as a result. I have also included some features
of graphical identity, including fonts, colours, and logos, where it appears that
they are being used signicantly to evoke the brand.
Link Denition Example Evokes
brand?
Repetition Repeating the full reference to
the brand
Orange… Orange
HSBC… HSBC
Strongly
evokes
Partial
repetition
A phrase contains a reference
to the brand, but refers to
something other than the brand
concept
Orange… the Orange
Service Promise
HSBC… HSBC branches
Strongly
evokes
Co-reference Where a concept is reinforced
by referring to it again, but not
using a full descriptive NP
Orange… we
HSBC… with us
Strongly
evokes
9
Brand elements Where a link is created by
language or visuals recognisably
special to the brand
Special terms such as
Orange TalkShare;
Distinctive font
Brand colour
Logo
Strongly
evokes
Possessive
inferrables
Where a link is created by
referring to something that the
brand has, does, or has given to
the customer, using a possessive
NP
Orange… our network
Your phone
HSBC… our branches
Your account
Evokes
General
inferrables
Where a link is created by
referring to something that
things of this type (e.g. banks,
telecomms companies in
general) have or do
Orange… the network
HSBC… branches
Does not
evoke
14    
Link Denition Example Evokes
brand?
Other NPs Referring to other things relevant
to the topic, but not related to
the brand
Your current address
Your laptop
Does not
evoke
Table 3: Forms of referring expression and brand evocation
Figures 2 and 3 give an illustration of the way in which Orange and HSBC use
referring expressions to reinforce the brand a technique Fairclough (1989:
28) refers to in the context of advertising as ‘building the image’ of the product
(see also Delin, 2000: 130 on the use of language in advertising to reinforce
product image).
Orange has branded all its services by explicitly repeating its name, so these
strongly evoke the brand through partial repetition (Orange Internet, Orange
Answer Fax). It also repeats the brand name in the context of things belonging
to the user, so that instead of possessive inferrables (your phone) it upgrades
them to co-references (
your Orange phone).
Figure 2: Evoking Orange brand
e Orange texts use brand keywords to claimordinary territory and brand it:
for example through the creation of proper nouns with capitalisation:
Orange (from N, the fruit, or Adj, the colour)
Orange International Text Messaging
Orange Multi Media services
we/us
Orange phones
Orange TalkShare
Orange Internet
Orange Answer Fax
our network
our cigar-lighter charger
orange colour
Orange
your laptop
your friends
your oce computer
system
typography
your Orange phone
your Orange account
your Orange Answer Phone
strongly evokes
evokes
doesn’t evoke
.  15
Brand elements are therefore created out of phrases that could otherwise
just be possessive inferrables (our text messaging service) or general infer-
rables (text messaging). Orange also brands strongly through colour and
typography.
Figure 2 shows that the range of techniques of brand reinforcement can
include signicant ‘upgrading’ of what would otherwise be ordinary ‘life-
world’ language (cf. Habermas, 1984) to co-referential or brand element
phrases. e specialnessof these services when related to Orange is therefore
constantly reinforced through a range of means of evoking the brand name
and its related concepts.
By comparison, HSBC is a more reticent brand. In one 22-page brochure,
HSBC is mentioned only ten times, and none of these references are in
the main body text: as part of the logo on the front of the leaet, twice in
images, once in its address on the cover of the leaet, four times in regulatory
statements, and twice in URLs
10
. e identity is not strongly reinforced by
typography, with the choice of the neutral Times and Univers as fonts. ere
is some reinforcement of the brand colour (the use of red for titles and bullet
points).
Where Orange services are branded by mentioning the brand name,
the HSBC brand does not appear in the names of its own special services.
Spacing and typography have been used to make service names dierent
from ordinary language, but these do not strongly evoke HSBC. Typography
and spacing are not used suciently consistently, either, to make special
services’ seem like any kind of set.
ServicePlus no space, both elements capitalised, italics on second
element
Talk Direct space, both elements capitalised, no italics on either
element
Typetalk no space, rst element capitalised, no italics on either
element
Concepts and elements that are ascribed to the user are not branded, so
possessive inferrables (your Basic Bank Account) are used, for example, rather
than co-referential phrases (your HSBC account). In addition to these few
possessive inferrables that do weakly evoke brand, however, there is a much
larger group of general NPs, oen also possessives, that refer to things that
could be had from any bank (your account, your branch). I have grouped these
(perhaps controversially) as non-evoking general NPs, because the concepts
they refer to could relate to any company providing similar services. e
result is a set of much weaker referential evocations of the brand. ese are
illustrated in Figure 3.
16    
Figure 3: Evoking HSBC brand
Interestingly, the language used by Orange serves to transform its brand identity
through use of chains of reference. rough co-reference by pronouns, the
brand becomes identied with the company as a group of people:
Orange covers… And you dont have to take our word for it.
e Orange Network Performance Promise means that if we ever lose your
call…
Orange literature gives the impression that there is no machinery involved in
managing telephone calls, as the network is also run by ‘we’ or the brand:
if we ever lose your call in mid conversation, we will…
we keep you connected
is means that chains of reference can alternate between the brand, the people,
and the telephone system and its handsets and treat them all as co-referring:
Orange… we… our network… the Orange Network Performance
Promise… if we lose your call…
Once identity is achieved between the brand, network, handset, and people,
connecting customers with this is straightforward, through possessives:
call from your Orange phone
your call
HSBC
we/us
our branches
our self-service machines
our/your Basic Bank Account
our buer zone
our Customer Service Centres
typography
HSBC
red colour
Talk Direct
ServicePlus
Typetalk
your local branch
your bank account
your rst personal loan
your statement
.  17
or through explicit assertion of the relationship:
Orange Global Travel Services is your one-stop travel service
e only other actor present in the text apart from Orange and the customer is
Wildre, a voice recognition ‘virtual personal assistant. is service is explicitly
personied with ‘she’: distinct from Orange, but working for them:
She can help you make calls, take and sort messages and store
contact details…
HSBC, by contrast, uses simple chains of reference that do not repeat the banks
name. e bank establishes itself at the outset as a kind of global topic by
minimal use of its name. To make sense of the text, the reader has to assume the
global referent of HSBC when encountering expressions such as the following
pronouns (underlined):
we
our special services
all our customers
Let us help you
our branches
Lack of repetition of the name assumes that the brand concept is at a very high
level of activation in the reader’s consciousness. is could be quite a power-
ful position: it forces the reader to adopt the attitude that no other possible
referent (Barclays bank, for example) can be intended. Less charitably, it is
an assumption that may not be warranted – and is not well salvaged by visual
evidence such as colour, layout, and typography. It is possible, therefore, that
the reader could forget which bank they were reading about halfway through
the document, or at least be le with a very weak impression of the brand.
is analysis showed that the dierences in referring expression usages
between the companies are marked. e highly noticeable presence of Orange
as a brand, and the highly developed way in which it constantly re-evokes the
brand in its language, is in keeping with a ‘Who we are’ position on the brand
continuum. HSBC’s lack of presence indicates that it does not currently occupy
a ‘Who we areposition, but this is negative evidence, rather than positive
evidence for any other positioning.
If HSBC’s position is deliberately sought, the background nature of the brand
(a ‘supportive presence’ rather than the ‘you’re on our team’ position adopted
by Orange) could account for its reticence in using its name. While there is a
general assumption that all brand texts would represent attempts to ‘push’ the
brand name, it is a known strategy in branding that this would not suit every
18    
brand position. Brands seeking a more subliminal presence might be better
reinforced by typography, layout, image, and colour.
is preliminary survey of referring expressions suggests that further system
-
atic research would be useful in revealing the relationship between patterns of
lexical cohesion and brand position. We can suggest, then, our rst hypothesis,
testable with consumers:
Hypothesis 1: Brand strength. e strength of the reader’s impression of
the brand will vary depending on the number and strength of referring
expressions that activate the brand or brand-related concepts.
Testing this hypothesis will, of course, require a clear notion of what ‘strength
of impression’ might be in cognitive terms, and how to measure it.
3.2 Participant roles
e second technique for correlating language use with brand position is
the analysis of participant roles and the processes they are associated with in
clauses. What kinds of processes is the brand associated with, particularly as
agent or doer? What kinds of roles are le to or ascribed to readers? What other
participants are there in the processes reported in the text?
is kind of study is related to what Fairclough (1989: 202) has called ‘build-
ing relationsbetween the producer/advertiser and the consumer, ‘building
imagesof products (in our case, brands), and ‘building consumersthemselves
by associating them with ideal ‘subject positionsin which they might be likely
to accept or reject the product and their relationship with it. We can capture
its manifestation in language through applying the framework for looking at
clause structure oered by Systemic Functional Grammar (Halliday, 1994:
101). is framework categorises verbs (known in the framework as processes)
not through their form, but through their meaning, so that we can look at the
eect certain kinds of verbs have on the meaning of the text.
e technique of looking at process types and their doers and done-tos has
been used successfully to analyse bias in newspaper reporting (Fowler et al.,
1979), in the apportioning of blame in the reporting of crime of domestic
violence (Clark, 1992), and in an analysis of presumption of innocence or guilt
in the Tracey Andrews ‘road rage’ murder (Delin, 2000: 28). Vestergaard and
Schroder (1985: 27) have a useful discussion of the application of a similar
approach to an advertising text. A clear introduction to the categories of analysis
(a useful supplement to Halliday) is given by ompson (1996: 78).
We can start our analysis with mental processes those processes that are
concerned with mental states. Because mental processes are only available to
.  19
animate objects, we would expect these to be associated closely with ‘personi-
ed’ brands – i.e. those that are aimed at the rightmost, least tangible, position
on the Brand Analytics continuum. However, it does not seem from this that
Orange as a brand occupies this position: in the document examined, Orange
actually gives itself only two mental processes:
Orange believes…
We know
Apart from these, Orange does not then have any other thoughts or feelings, at
least as a brand. ere is no we think, we feel and so on. ese might be expected
from a brand on a mission. Interestingly, though, while the brand itself is not
projected as having a rich mental life, its handsets and services are:
your Orange phonebook remembers your numbers for you
Wildre… learns about how you want her to help
When it comes to doing, most Orange sentences in my sample have the cus-
tomer as doer. Orange is not a doer in many clauses at all: in fact, its services
and handsets are doers more than the brand itself is. e following represents
a brief survey of the clauses in Oranges customer guide ‘What do you want to
do?’. Table 4 below summarises the kinds of processes that Orange customers
are involved in.
Process type Material:
action verbs
Mental:
thinking verbs
Verbal:
saying
verbs
Relational:
being and
connecting
verbs
Orange customers get through to
people
call back
nd another
provider
call
enter (names)
save (to memory)
get the message
listen to messages
return (call)
listen
record (a greeting)
change (greeting)
select
take someone’s
word for
something*
need to*
need to track
down
worry*
forget
respond have (a
phone)
rely on*
be connected
* asterisked processes = negatives
Table 4: Processes ascribed to Orange customers
20    
Customers are given practical action verbs (material processes), which appear
to convey that they can get on with what they have a phone for. e following
material processes (underlined) all have customer as doer:
you can access the information you want via text messaging..
if youre planning a last minute weekend away
you can keep up with whats going on…
you can access Orange WAP services…
Customers are not apparently expected to interact much with Orange (there
is only one verbal process respond which links the customer to Orange).
Customers’ mental processes are mostly couched in terms of things Orange is
trying to prevent: many of the processes are negatives, such as ‘you won’t need
to. All the negatives in Table 4 (e.g. You wont have to worry about… represent
Orange, or one of its services, preventing things that the customer would not
want, such as With the Orange Answerphone service, you wont have to).
Negative processes are marked in Table 4 with asterisks.
e processes that the Orange texts ascribe to the brand itself are summarised
in Table 5. As I noted above, the Orange brand does not appear as doer of many
material processes: it is doer in only four of the clauses looked at. Neither is
the brand a big thinker: only two examples of mental processes were found in
a long brochure: We believe that no other UK digital mobile network can match
us on service… and Orange knows the major cities of the world inside out. e
Orange brand is mainly associated with processes that state its relationship to
its customers, and imply for it a helpful disposition. is means that, gram-
matically, Orange is a subject (doer) in a few clauses, but most oen appears
in a subordinate clause or phrase and with the customer as the subject of the
main clause. Orange is therefore backgrounded as a doer, while the customer
is the active party. e company, instead, is providing services that are at the
customers elbow for them to use to do what they want to, and while appearing
to have an overseeing, caring role.
Process
type
Material:
action verbs
Mental:
thinking
verbs
Verbal:
saying
verbs
Relational: being
and connecting
verbs
Behavioural:
having some
disposition
Orange lose (your call)*
credit
give
text
believe
know
oer keep (you
connected)
cover
be here
take care of
keep our
promises
waiting to help
Table 5: Processes ascribed to the Orange brand
.  21
Oranges subordinate role comes out as a range of facilitating relationships
arranged around the main clause. ese relationships are usefully described
rhetorically. Using the terms for the moment of Mann and ompsons (1987)
Rhetorical Structure eory (RST), we nd Orange appearing in clauses or
phrases expressing the relation of circumstance:
When you register for Orange Internet…youll nd
as means:
With Orange Mail you can access your personal email account
check out the comedy, theatre, and gig listings with SceneOne.
as destination:
popular tunes that you can download to your Orange phone
and as enablement:
Orange Multi Media Services can also help you to nd your way
While the customer has agency of the main’ action – even if it is not in every
case syntactically the main clause these examples show the Orange brand
placed in a range of facilitative relationships with the things that customers
might want to do.
While Orange as a brand is not always directly foregrounded, the services
Orange oers are an important protagonist in clause structure. Table 6 shows
the kinds of processes ascribed to brand-evoking services such as
the Orange
Network Performance Promise, our Customer Services Team, your Orange phone-
book, Orange Directory Enquiries, Caller id, Line 2, our xed rates, Orange
Answer Phone, Orange Multi Media Services, and Orange International Text
Messaging.
Process
type
Material:
action verbs
Mental:
thinking
verbs
Verbal:
saying
verbs
Relational: being
and connecting
verbs
Behavioural:
having some
disposition
Orange
service
store
show
give
remember keep in
touch
mean
take care of
allow
enable
help
have
be waiting
Table 6: Processes ascribed to Orange services
22    
While most of the processes that Orange services take part in are relational
rather than material, there are many more clauses that have the Orange service
as a protagonist doing and relating than there are clauses that have Orange
itself in this position. Doing and relating is therefore heavily loaded onto the
services, rather than the brand directly. We have seen the strong evoking link
between the services and the brand already, so evocation of the brand is not
weakened by placing the services rather than the brand itself in this position.
However, Orange services are positioned as key delegates enacting relationships
with customers. ey therefore act as the bridge that builds the relationship,
linguistically and conceptually, between the Orange brand and its customers.
Turning now to the processes ascribed to HSBC customers (Table 7), there
is no great dierence in material processes: as with Orange, customers are
expected to do things associated with their account. Across the whole spread
of other processes, however, the onus is on the customer to be agentive in
the operation of their accounts. While Orange was reassuring that customers
would not have to take someones word for something, would not need to, would
not have to worry, would not have to rely on, and could respond, be connected,
have, and forget, HSBC customers need to see and call in, complete, pay, apply,
request, arrange, agree, and set up. ere is a sense that vigilance is required on
the customers part (control (money), review, check).
e processes associated with HSBC as doer (Table 8) appear to be mainly
regulatory in nature. ere are no relational processes in the sample looked
at, where Orange used these processes to set up a caring relationship with the
customer. Likewise, HSBCs only disposition is that it needs to see (as in check)
an alternative form of identication when customers are opening an account,
while Orange is keeping promises and waiting to help.
Process
type
Material:
action
verbs
Mental:
thinking
verbs
Verbal:
saying
verbs
Relational: being
and connecting
verbs
Behavioural:
having some
disposition
HSBC
Customers
see
call in
complete
pay
get
apply
withdraw
print
transfer
use
think
review
check
tell
call
phone
request
arrange
order
speak
control take advantage
nd it hard to
beat
agree
need to set up
Table 7: Processes ascribed to HSBC customers
.  23
Process
type
Material:
action
verbs
Mental:
thinking
verbs
Verbal:
saying
verbs
Relational: being
and connecting
verbs
Behavioural:
having some
disposition
HSBC pay
resolve
use
will not
charge
verify
monitor
check
promise
oer
reserve
the right
contact
advise
need to see
Table 8: Processes ascribed to the HSBC brand
e services HSBC oers are predominantly described in terms of what the
customer can do. However, in the context of banking communication, the
verbs through which this is communicated (may, let, can) are notoriously
ambiguous between expressing possibility and expressing permission. While
it may be the intention to express a realm of possibility, the permission reading
is also available, and is perhaps more prominent when the reader knows that
the ‘voice’ is that of a bank:
11
Our Instant Access Savings Account lets you pay in money
With 24-hour access via our self-service machines you can…
Telephone banking…you can control
Looking at the analysis of the HSBC brochures, it may appear that there is in
fact a shortage of doers – the bank is not oering to do much, but neither are
its services. So how is the bulk of the language made up? e answer to this
riddle lies in the grammar of the sentences HSBC is using: many of the clauses
are imperatives addressed to the customer:
Phone us on 0800 130 130
Complete and return your application form to us
Take advantage of our special fee-free credit card and introductory interest
rate oer
Ask any branch for a leaet
ere are also many clauses that are passive in nature, if not in grammar, of the
form (some service) is available. ere are also many real passives:
Details will be shown on the cash machine screen
Cash can be withdrawn
24    
ese do not show who is doing the showing or withdrawing. ere is also a
great deal of listing and use of non-clause grammar, so that nobody is doer, or
other participants are le vague:
Cash withdrawals with no penalty.
Quarterly account statements.
Supported by our hassle-free account opening guarantee.
e balance between dierent clause types comes out heavily in favour of
directives with the customer as doer. e text, then, is predominantly telling
the customer what to do.
In contrast, Orange uses two kinds of sentence prominently: structures that
list virtues of the product, and structures that go into more detail to tell the
customer how to do things. e basic informative sentence ‘frame’ is given in
Table 9.
Circumstance Enablement Clause kernel Purpose Means
When you register
for x service
we’ll even help
you
with an x from
Orange
when you’re
doing x
you can (even) do x so that
(advantage)
with Orange’s x
you’ll be able to with your Orange
x
Table 9: Orange informative sentence frame
Everything in this structure except the clause kernel is grammatically optional.
However, in the actual text, there is always an element included that shows the
benet to the customer, either a purpose or enablement element. is means
that the resulting statement is a declarative sentence (Well even help you to
set up an extra line, so that you can take social and business calls on separate
numbers) rather than an imperative (Set up an extra line).
When the customer has to be asked to do an action, the general action
sentence structure is that given in Table 10. e request to do something is
nearly always minimised by some statement about how simple it is, and there
is frequently some statement of the benet to the customer.
.  25
Minimiser Action Result
simply do x (and (then) do y) your result will happen.
just Orange x service makes this
easy.
all you have to do is then/so you can do x.
Table 10: Orange directive sentence frame
Orange language, therefore, serves to create a link between customer and
desired services, by positioning those services as enabling desired action.
When customers are asked to do something, it is described as simple and
advantageous. HSBC language, on the other hand, places the customer in
the position of responsible doer, and the bank as allower of those deeds.
e brand is not constructed as apparently oering to reciprocate, and the
benets oered to readers are oen those of withholding penalties (we will not
charge) rather than positive advantages. It is a Tone of Voice that constructs
a paternalistic, legislative and withholding/allowing position for the brand,
which this analysis suggests is likely to create social distance between the brand
and its customers.
is short analysis also gives rise to a testable hypothesis for further research
which should reveal more about the way brands are constructed in relation to
services, customers, and other participants in the world of the text. If we assume
that this kind of textual construction does indeed have an eect on readers, we
can suggest that the eect might inuence both perceptions of the brand, and
perceptions of the self in relation to the brand. is proposal, then, is captured
in Hypothesis 2:
Hypothesis 2: Participant roles. e reader’s impression of the brand will
be inuenced by the type of processes ascribed to the brand and its related
concepts. Further, the reader’s construction of their own self-image will be
inuenced by the types of processes ascribed to them in the text. Finally,
the reader’s understanding of their own relationship with the brand and
its associated concepts will be inuenced by the way all these participants
are positioned by association with processes within particular kinds of
sentence frame.
e diculty in testing this hypothesis lies primarily in what we mean by
readers being inuenced by’ the construction of them in the text. While this
remains vague, this approach does t within a tradition of research that assumes
that linguistic means of construction of readers and hearers do inuence the
26    
construction of mental worlds the ‘relational values(cf. Fairclough, 1989:
179) that are created through lexical and grammatical choices.
3.3 Presupposition and assumption
As Vestergaard and Schroder (1985: 24) note, presupposition is ‘an extremely
frequent featureof advertising texts. Also in relation to advertising, Cook
(2001: 178) suggests that any text must make assumptions about the knowl-
edge of its readers, and it is sometimes very informative to spell out exactly
what this may be.While this process can lead to the innite expansion of a
text, both Cook (2001) and Fairclough (1989: 78) concur that it is ‘what is
omitted in discourse, the gaps within it, which constitute the shared ideology
of the participants’ (Cook, 2001: 179). As with all texts, the use of presupposi-
tion and assumption can be a convenient means of recruiting genuine shared
knowledge as an aid to economy of explanation. However, texts can also use
these resources to attribute to readers apparent shared knowledgethat is in
fact newly constructed in the process of interpretation.
To see how this ‘shared knowledge’ is constructed, we can look rst at what
is explicitly asserted in the texts. e HSBC texts make claims for the banks
coverage and size, which would suggest its placement in the ‘what we havearea
of the Brand Analytics continuum:
…one of the worlds largest banking and nancial services organisations
with more than 6,500 oces in 78 countries and territories
we oer 1,700 branches in the UK, and in-store branches at some
Morrisons supermarkets
However, the Orange literature also makes similar claims for coverage of the
brand:
Orange now covers over 99% of the UK population
more transmitters than any other UK network
Earlier, we looked in at the kinds of ‘brand elementsthat were associated with
Orange: these were mainly characterised in terms of Orange services with
explicit branding (the Orange Answer Phone, for example). We saw then that the
HSBC name was not used in this way, with possessive inferrables (our service,
your self-service machines) appearing instead. is means that the services
are not securely tied to the brand, as this was only weakly established at the
outset, while the Orange texts positioned the Orange brand, its network, and
its services as interchangeable.
.  27
e crucial point here is the means of introduction of the services: HSBC
texts refer to them using possessives without introducing them earlier (not
we have a service… our service but simply our service). What the and our have
in common is that they are both presuppositional: they both linguistically
presuppose, require, or force the reader to infer the existence of their referents
without these things having been introduced beforehand. In the case of Orange,
however, these things are presupposed as belonging to the Orange brand, which
has been positioned as belonging equally to the customer and the company. As
a result, there is a sense of shared ownership of these referents. HSBC language,
on the other hand, introduces many services for the rst time using possessives
that describe the services as belonging to the brand alone. is, together with
the position of authority it has constructed, may encourage customers to feel
that they can take advantage of these services only if HSBC lets them.
Table 11 summarises HSBCs presupposed services, etc. in terms of the brand
continuum. ere is a strong concentration of what is presupposed about
the brand in the central area of the continuum: products and services, and
approaches and processes.
Infrastructure Products and services Approaches and
processes
Intangibles
our self-service
machines
our service(s)
our Instant Access
Savings account
our Individual Service
Review
our Bank Account
our Cardholder Payment
Protection Insurance
our 24-hour automated
telephone banking
service
our Basic Bank Account
our Telephone, TV or
Internet banking
our special fee-free
credit-card interest rate
oer
the services of the HSBC
group
our hassle-free account
opening guarantee
our buer zone
our unauthorised
overdraft fee
our personal terms and
conditions
our paid-up capital and
reserves
our security
our discretion
Table 11: Presuppositions about what HSBC has
It is also interesting to look at the assumptions that texts from both compa-
nies make about their customers. Both, understandably, make assumptions
that their customers have things that have to do with banking and phoning
28    
(your cards, repayments, account balance, and your calls, phone, user guide, for
example). Beyond this, HSBC texts make assumptions that may be linked to
customerslifestyles: your needs, future plans, spare cash, friends and family, for
example. Orange texts presuppose that its customers have a home, an oce,
they are on the move, they have groups of friends, they forget to take their phone
with them, they get some exciting news, they want to have some fun, they need
to call from abroad, they want to know the state of the football scores, and many
more. Arguably, Orange lifestyle presuppositions construct a subject position
that is more young, fun, and exciting than that constructed for consumers in
the HSBC texts.
Going beyond presupposition, Orange uses implicature, a weaker form of
assumption that nevertheless still presents facts as uncontroversial. In the
examples below, again and any more give rise to implicatures that the situation
has happened before:
You need never forget a birthday or miss an important email message ever
again.
You wont need to rely on your old paper address book – or your memory
– any more.
e Orange texts also assert beliefs on the part of the customer: in the brochure,
the following are given as the customer’s ‘voice’:
I want to stay in touch
I want to get through to people
I want technology that helps me
I want to be able to call abroad
ey suggest scenarios to appeal to the customer as familiar:
Youve just nished a tough day at work and fancy checking whats on at
the local cinema.
Your child is feeling ill and you want to know where to nd the nearest
late-night chemist.
ese are lifestyle suppositions which suggest a ‘knowerwith detailed under-
standing of the personal circumstances of the customers. ere is no such
knower implied in the HSBC texts, as their suppositions are of the most
neutral kind.
e kinds of presuppositions that appear in the HSBC texts suggest that
HSBC occupies a central position on the brand continuum: while there are
infrastructure claims, there is an emphasis on products and services. e
introduction of services as through possessive pronouns (our) rather than
to the brand (HSBCs) excludes customers from ownership of them. While
.  29
many of the products and services mentioned in the HSBC texts are positive,
the approaches and processes it mentions have a negative tone (security, fees,
terms and conditions) that refer to the bank protecting itself rather than oering
advantages to the consumer. While the Orange texts assume attractive lifestyles
for its customers (busy, with friends, popular), HSBC texts assume little more
than that their readers need to go to a bank. is may be in keeping with a
brand position that is not closely associated with the ‘who we arelocation on
the continuum – the implication being that if HSBC is not presenting itself as
a group of people, then they cannot know customers personally.
is survey of an initial sample suggests a third hypothesis for further
research:
Hypothesis 3: Assumption. e closeness of readersengagement with a
brand relies on the predominance of textual assumptions and presuppositions
that are either (a) accurate about those readerslifestyles and beliefs or (b)
represent attractive extensions of those lifestyles and beliefs.
In fact, the decision about what kinds of lifestyles and beliefs to assume, and
whether to stretchthose to aspirations about desired lifestyles, may itself be
dependent on the brand position a company is seeking to adopt. A factual,
tangibleposition may rely more on accurate understandings of customers
current needs and lifestyles, while an ‘intangibleposition may seek more
frequently to project apparent beliefs that are not accurate, but which readers
might like to aspire to.
3.4 Constructing social relations: tenor
e fourth and nal analytical approach applied in this research was to look
at the way in which social relationships in general were constructed between
brands and potential customers through the medium of the texts. is involves
the use of constructions of authority positions, which have been described
by Fairclough (1989: 182), in his discussion of political speeches, as existing
in tension with claims for solidarity with an audience. In addition, there has
been much comment on the mixing of ‘everyday languagewith more public,
institutional genres of discourse in a range of contexts: Fairclough (1992: 217)
refers to the appropriation of conversation by institutions, while Goodman
(1996: 141), following Fairclough (1994), describes extensively the way in
which ‘informalizationand conversationalizationof language is used to create
an impression of social closeness, even in institutional, public texts. Conversely,
as Fowler (1991: 128) suggests, the use of formal, impersonal language forms
serves to construct and reinforce hierarchical relationships of power, in which
30    
greater social distance is created and maintained between the ‘voice’ of the text
and its readers.
Analysing Orange and HSBC texts exposes dierences in the way the lan-
guage voiceallows each to construct social relationships with readers. We have
seen something of this in the kinds of position that each company constructs
for its customers, and in how much they seem to know about customer life-
styles. ere are other determinants of social relations, some implicit and some
explicit.
HSBC texts, as we have seen, construct a position of authority for the brand
in that they do not hesitate to mention statutory and legal obligations and
expectations, and the banks right to refuse services to the customer. In fact,
the rst body text of one new accounts brochure mentions just such a right in
the rst few lines:
£50 Buer Zone
Although our buer zone is not a right to go overdrawn, should your
account dip into the red by £50 at our discretion, we will not charge you
our unauthorised borrowing fee.
In fact, terms and conditions, regulatory statements, and other boilerplate
appear throughout the HSBC documents, reinforcing the position that the bank
gives (or withholds) permission for the customer to use its services. Some of the
banks actions are given as concessions: we will not charge you our unauthorised
borrowing fee, for example. Likewise, attempts to state what the customer is
entitled to might reect negatively on the bank: on opening an account, the
customer is reassured that HSBC will pay you £10 every time we get it wrong.
e nal phrase presupposes that there will be such cases, and more than one
for each customer.
is material solidly reinforces HSBC as relatively powerful and the customer
as relatively powerless. I would suggest that the inclusion of legal elements
such as this then constrains the coherent positions that the language voicecan
occupy: it is dicult to then be informal and friendly. However, HSBC does
attempt some conversationalizationthrough informal language use, such as
by using clause fragments instead of full clauses:
Easy to open. Simple to understand. Easy to manage.
To help you keep track of your money.
In the context of regulatory statements, however, any conversational, informal
position can come across as inauthentic and false.
e Orange texts use informal grammar quite consistently. Sentences can
begin with conjunctions:
.  31
And you dont have to take our word for it.
Informal, conversational clause structure, contractions, and informal punctua-
tion such as dashes are common in the Orange texts:
It happens. Phones get lost…
dont, wont, youre, cant
If you need to contact someone abroad, a friend on holiday or a relative
living outside the UK – it makes perfect sense to call from your Orange
phone.
To further reinforce the eect of conversation, the text mimics question-answer
structures across two speakers:
I want to get through to people (with photo of person)
With Orange you will.
e texts use rhetorical questioning to simulate understanding of a problem
that a customer might have:
Unexpected meeting? Orange can reserve an oce suite for you…
Home from home? Youll want our international apartment reservation
service.
A few days to spare? We can even charter a yacht for you.
Orange language uses well-known phrases and sayings that might occur in
everyday speech:
whether its time for morning coee or aernoon tea
write home about it
keep your nger on the pulse
seven days a week
when youre a stranger in a strange town
peace of mind
home from home
ere is also language play, using gures from poetic and oratorical language,
such as alliteration:
So if you need directions in Dublin, a room in Rotterdam…or haute cuisine
in Harare
be it Birmingham or Bali
Another rhetorical device (this time in the classical, rather than the RST sense)
is the use of the ‘rule of three’ when creating lists, such as lists of benets:
32    
for emergencies, the oce or when youre travelling.
within each zone, day and night, seven days a week.
oering city guides, executive services and international emergency
assistance across the globe
we can direct you to a local pharmacy, medical centre or hospital
Orange sentences vastly favour conjunction (and, or, and much more rarely
but) over subordination. is, coupled with the fact that the preferred sentence
length rarely goes beyond two clauses, means that sentences oen begin with
connectives:
en you can schedule text message reminders…
And because Answer Fax can receive multiple faxes…
So, if you are on a Talk Plan from Talk 150 upwards…
is adds to the impression of simplicity and informality, and more closely
mimics the style of conversation. HSBC grammar ranges from the list-like:
Alternatively, you can:
phone us on 0800…
email us…
to longer sentences with conditionals and circumstances:
If you choose not to hold your account at a particular branch, you can…
Before we can complete your application we will need to see…
When you open your account by phone or by sending in your application
form, you can…
ese sentences tend to be structurally more complex than Orange sentences
and less apparently like speech. Finally, about a third of HSBC’s sentences in
this sample consist of recognisably ‘legaltext, much more formal in nature than
conversational English. Legal text tends not to use pronouns, for example, and
tends not to break sentences where speakers might:
Payments under the Scheme are limited to 90% of the depositors total
deposits subject to a maximum payment to any one depositor of
£18,000…
Most deposits denominated in Sterling and other European Economic
Area currencies and euro made with oces of us within the European
Economic Area are covered.
By contrast, the Orange texts use conversational devices, word play, and infor-
mal grammar and vocabulary to create the sense of a close relationship with
readers. e presence of conversational features in the Orange texts, and their
.  33
relative absence in the HSBC texts, do seem to suggest that Orange is seeking
a more personal, ‘who we are’ position for the Orange brand than HSBC.
is brief overview allows us to suggest a fourth hypothesis for further
research, which allows us to bring in some specic brand qualities that may be
associated with close social relationships.
Hypothesis 4: Conversationalization. e greater the frequency of conver-
sationallinguistic features and their written correlates in a brand text, the
more socially close the reader will feel to the brand, and the more likely they
will be to associate the brand with ‘humanbrand qualities such as ‘warm
and ‘approachable.
Testing such a hypothesis would require careful control of the content of texts and
perhaps presenting consumers with dierent versions to rate for ‘humanness.
3.5 Orange and HSBC brand positions
e analysis reported in this paper does seem to support the view that Orange
and HSBC language establishes them in dierent positions on the Brand
Analytics continuum. Orange language positions its brand as oering services,
help and support, and being in close connection with its customers’ lifestyles.
e texts strongly associate the Orange name with its services, and those
services as belonging to the consumer. is creates a chain that encourages
the consumer to have ownership of what Orange oers, rather than Orange
owning the services itself. While the Orange brand aspires to occupying the
rightmost intangible position on the continuum, however, its language does not
completely support this. e texts do not explicitly foreground its values and
people, do not make claims for its sta, and do not repeat any set of beliefs about
what it stands for. e language simply positions the brand as a supportive pres-
ence that does not dierentiate between people, phones, and services – all are
interchangeable. Inanimate and ethereal services are personied, sharing the
human qualities of the company and the brand. For Orange, it is as if phones are
people. is places the customer, who has his or her hand on her phone most of
the time, in physical proximity with the brand identity and values. Orange does
not therefore need to talk explicitly about its employees, having successfully
turned services into people. e brand speaks informally, even conversationally,
perhaps much as customers might do to their friends on the phone.
e HSBC texts position the company as a provider of services that enjoys a
large infrastructure. e infrastructure and services clearly belong to HSBC, but
it allows its customers to use them subject to certain conditions being met. e
language does not suppose any particular qualities about consumer lifestyles
beyond that the customer wants the services to be accessible. It claims posses-
34    
sion of the services it oers without branding them explicitly, and it is not clear
from the presentation how extensive or limited the list of services is. It is up to
HSBC to oer or withdraw services from its customers at any time. It does not
expect a close relationship with customers as it has a regulatory obligation over
them, but at times can talk to them quite casually. Although positioning on the
Brand Analytics continuum cannot be done precisely, HSBC’s language seems
to suggest a products and services emphasis rather than an infrastructure
claim, which would place it in the lemost position but its positioning of
services in relation to consumers is completely dierent from Oranges in that
consumers are not at liberty to call these services their own. is is perhaps
a dierence between the two brands that the Brand Analytics continuum has
not the delicacy to reect.
We have also seen that language analysis can detect whether the positions
that are being constructed for brand consumers are attractive and appropriate.
Orange heavily constructs its consumers as young, busy, and popular, and
itself as helping them do what such people need to do. e assumptions HSBC
has about customers are either neutral or negative such as the assumption
that they may want to go into the red without authorisation, or the assertion
that the bank cannot take only one form of identication (presumably in case
customers are lying about who they are) to open an account. Its own position
is one of checking up on them before allowing them to take advantage of its
services. It is not in dispute that banks need to operate such procedures, but
the careful management of Tone of Voice is about how foregrounded and how
well counterbalanced these messages are.
4 Discussion and conclusion
ere are some important points to note about the methodology of this small
study, and of future, larger ones. Companies may not be using language that
suits their brand position on the continuum, and so it cannot be assumed
that any existing text is a perfect example – or a perfect test of the theory that
language can predict and reect brand position. However, linguistic analysis
can expose shortcomings, and can act as useful means of triangulation, testing
other means of auditing brand position.
We should also bear in mind that any sample of language might be untypical.
Companies produce large amounts of language, written and spoken. We know
that language varies between genres and registers and so it would be odd to
assume that a company’s whole language output, given that the document types
and purposes are dierent, will oer a consistent result. A company’s brand
positioning may also be deliberately inconsistent throughout its output. It may
.  35
be part of brand positioning to reect genre and register in dierent ways.
Dierent brands output dierent genres of texts, and the pattern of genres that
are chosen are part of the brands communication identity. is means that we
may not always be able to nd texts that are comparable between brands.
However, I hope it is clear, even given these provisos, that the application
of some well known language frameworks chains of reference, transitivity
analysis, presupposition, features of tenor being but four can add to our
insights about how brands communicate and position themselves, deliberately
or otherwise. What I have not done so far is to attempt to link brands to distinct
brand values, such as warm, expert, and friendly. Neither have I shown that
particular language features appearing in the brand Tone of Voice will have a
predictable eect on readers in general (or readers of certain kinds). at is, it is
not possible to say as yet whether deliberate use of, say, certain participant roles
or features of tenor will prompt readers to feel that the brand is communicating
the values that the brand owners hope it is. It is crucial that consumer research
on the hypotheses outlined above actually tests whether dierent language
choices are having the eects that we intuitively think they might.
e four linguistic approaches outlined above represent only a small sample
of the possible ways in which linguistic theory can be used to interrogate the
linguistic construction of brand position. While there is not the room in this
paper to go into the interesting area of Appraisal eory (see for example,
Iedema et al., 1994; Christie & Martin, 1997; White, 1998; Macken-Horarik
& Martin, 2003, inter alia), the kinds of features identied within the eory
(which might be more appropriately termed a model, or an extension of the
various networks within a systemic grammar) are suggested to be those that
show the writer’s or speaker’s point of view or evaluation of the content of
what they are saying. is is likely to be a feature of branded texts, and if so, it
will have a role to play in developing a notion of authorial presence (or brand
presence?) to readers and hearers. at is, if the brand (or author, or narrator) of
the text is constructed as having a point of view, it is likely that they may come
across as being more (or less) certain, human, authoritative, and so on – some
of which features may relate closely to the brand values identied earlier on in
this paper. is view can be stated as Hypothesis 5:
Hypothesis 5: Explicit evaluation. e reader’s/hearers impression of the
brand will be inuenced by linguistic markers of the brands point of view’
such as those identied within Appraisal eory.
e weasel word here is ‘impression– we would of course need to nd a way
of identifying what an ‘impressionis, and how it might vary, in order to test
with groups of consumers.
36    
Another possible way of identifying textual characteristics with brand values
is through the lexicon, and through collocation studies. We know from literary
stylistics (cf. Carter, 1982; Carter & Nash, 1990; Carter & Simpson, 1989, for
example) that poetic eects are oen created through the use of unusual lexis
and unusual collocation. While the nature of the data examined for the study
reported above did not really reveal much in the way of unusual lexis, there are
many branded texts (such as the advertising and marketing copy referred to in
Section 2, for example) that do oer more lexical potential. In relation to such
texts, then, Hypothesis 6 seems to be a reasonable assumption:
Hypothesis 6: Unusualness. e higher the number of low-frequency
vocabulary items and collocations that occur in the brand text, the more
likely the reader/hearer will be to associate the brand with ‘unusualbrand
qualities such as ‘quirky’ or ‘fun.
Testing Hypothesis 6 will clearly involve using a large-scale corpus, either of the
language in general or, ideally, of the language of brands from a similar sector, to
benchmark the frequency and collocational patterns of the terms of interest.
ese six hypotheses do not exhaust the language features potentially appli-
cable to describing how brand positions are constructed through the medium
of language. ere are several other elds of study which may well yield ways
of tracing the links between language form, brand values, and the perception
of brand personalityby readers. For example, we know from research in
psychology
12
that personality traits are known to aect, and be realised by,
linguistic behaviour (e.g. Dewaele, 2002; Pennebaker & King, 1999; Gill, 2003),
although the kinds of personality trait that psychologists are interested in such
as extroversion and aggression are not always those that brand owners would
require a text to communicate. Linguistic patterns have been found to relate to
personal qualities such as sociability, anxiety and other emotions, the intention
to deceive, and even the state of the writers/speaker’s physical or psychological
health (e.g. Gill & Oberlander, 2003; Newman et al., 2003; Hancock et al.,
2003; Graybeal et al., 2002; Hemenover, 2003; Oxman et al., 1988; Campbell &
Pennebaker, 2003). It will also be relevant to try to relate the kinds of qualities
that appear in branding to the current theories of personality, such as the ‘Big
Fiveeory (cf. Goldberg, 1993; John, 1990) and the Five Factor Model (cf.
McRae & Costa, 1997).
4.1 Commercial uses of linguistic techniques
e application of linguistic frameworks to commercial texts has not been
without controversy, and the use of this research to work with commercial
.  37
clients to further their aims in business goes against the traditional application
of these methods.
e development of linguistic analysis of advertising and media such as
newspaper reporting has been to some extent and certainly was in its incep-
tion intended to expose the practices of organisations that in some ways
seek to manipulate, or even deceive, the general public. For example, the
outer cover of Faircloughs (1989) Language and Power describes the book as
about how language functions in maintaining and changing power relations
in modern society, about ways of analysing language which can reveal these
processes, and about how people can become more conscious of them and
more able to resist and change them.’ Cook (2001: 208) presents a review of
academic attitudes to advertising and the ways in which these attitudes have
shaped research. Concentrating on the work of Geis (1982) and Williamson
(1978) as examples, Cook (2001: 210) suggests that the ideological position
represented in such work is that the senders of ads are villains, intent on
deception; the recipients are uneducated, vulnerable, and easily deceived;
the observers are superior to both, and unaected by the text, because they
can decode it.’ However, Cook also points out that, as ads have become more
mainstream, and there is something more to them than mere commercial
activity’ (ibid: 214), there is less problematising of them in the more recent
academic literature and more apparent acceptance that they are but one
genre worthy of scholarly attention. While the assumption that linguistic
analysis should necessarily be a means of ghting against the ubiquity, morals,
economics or politics underlying the production of a genre of text has appar-
ently become less strong, there is little so far that takes the opposite view: that
linguistics could actually be applied to help advertisers or the brand owners
behind them to create texts that do a particular job, and to understand
better how texts might communicate in certain ways. e perspective taken
in this paper represents such a view – that the insights available from within
linguistics can, and should, be made available to a commercial audience,
who seek to improve their understanding of the language upon which their
businesses depend.
Attention to the language that companies used in areas outside creative copy-
writing tasks such as advertising really began with the widespread adoption of
‘Plain Englishor Clear Englishinitiatives in the 1980s, which were mainly
intended to help customers understand important communications such as
forms, letters, and the terms and conditions of credit cards and mortgages. At
the same time, as Cameron (2000) reports, the language used in call centres was
also receiving prescriptive attention, with mixed results. Neither of these initia-
tives had behind them any linguistic underpinning that we might recognise
38    
as coming from academic research in the eld. e development of ‘branded
languageor Tone of Voiceis a much more recent phenomenon, developing
only perhaps since 2000, and only two agencies in the UK, Enterprise IG
and Linguistic Landscapes, have applied to it any techniques, approaches, or
knowledge from linguistics.
Fairclough (1989: 211–23 et seq) has characterised the use of explicit lan-
guage knowledge to produce specic eects as a technologization of discourse,
in which everyday’ orders of discourse become increasingly regulated by
researchers, designers, and trainers, and that the resulting styles of discourse
simulateelements of private discourse in the public sphere (Fairclough, 1992:
215). It is important to note that few of the genres through which brands
are communicated are easily describable as everyday’ in the sense that they
are the site of personal identity work for the individuals and groups that
produce them, with the exception of spontaneous speech (e.g. face-to-face
and telephone interactions). While the notion of discourse technologist’
can imply a negative view of the use of language knowledge in commercial
contexts and to serve commercial aims, there are positive benets to the pres-
ence of knowledgeable practitioners who use linguistic insights to improve
the coherence, consistency, brevity, clarity, appropriateness and interest and
therefore the eectiveness of the communications that companies use.
Research informed by knowledge about language can, for example, nd out
the specic request forms that customers prefer when they are being asked
to do something, or nd out whether the ‘Plain Language’ that is being used
is in fact over-simple and condescending for some customer groups. It seems
clear that knowledgeable practitioners working in commercial contexts can be
inuential in providing linguistically-enlightened advice such as counselling
against the use of call centre scripting when it is otherwise doubtful that
companies would ever come into contact with, let alone be inuenced by,
academic views on the subject.
is is not to say that the application of linguistic theory in commercial
contexts is without its practical problems. Applying linguistics to commercial
communications still needs to be further developed if companies and organi-
sations are to feel fully comfortable with it. Such development will need to
overcome at least the following challenges:
1. Using linguistics potentially adds a costly and time-consuming step
to the process of understanding a brands communication needs and
current patterns, which is something that could conceivably be done
‘by eye, using a linguist, copywriter, or other language-aware profes
-
sional who would simply read or listen to communications to gain a
.  39
general impression. e linguistic approach will undoubtedly produce
an analysis at a level of detail that the client company will rarely want
to see.
2. Even when a satisfactory linguistic description of a desired Tone of Voice
has been completed, it is dicult to express it in terms that the people
who need to use it will understand. is is because few of the adult work-
ing population in the UK have any background in grammar and related
concepts. While my own experience has shown it to be possible to
discuss passives and reexive verbs with the branding team of a Spanish
bank, I have not encountered a British client with such a well-developed
understanding of grammar. Interestingly, design specialists have their
own technical vocabulary, and are happy to talk about reverse blocks,
half-tones, picas, and call-outs, but communications specialists in the
industry appear content to manage without a metalanguage at any simi-
lar level of specicity.
However, if these challenges can be overcome, there are also advantages that
can be gained from the application of linguistic frameworks in the planning and
production of commercial texts. Firstly, systematic linguistic analysis exposes
patterns that are not casually visible ‘by eye, and can yield genuinely new
insights into the ways commercial texts work. Secondly, because branding is
based on the idea that brands need to be distinctive from one another, it is surely
important to brand owners to have some reliable way of describing what is
distinctive about a brands language. Linguistic analysis provides for an explicit
description of language patterns, and hence for such an explicit statement, in
place of the rather vaguer statements currently extant. irdly, an explicit,
linguistically-sound understanding of a language style is both potentially com-
municable for training and information within the company, and usable for
detecting (and planning) shis in the brands language across time. Fourthly,
experience suggests that companies are oen grateful for evidence, rather than
just opinion, about the best language to use to communicate their brands – to
settle internal arguments about what is, and what is not, ‘on brand, if nothing
else. Finally, it is important for any organisation with global aspirations to
think about how its communications might work in other languages. Explicit
statements about Tone of Voice in the source language can help translators to
see what kind of ‘feel’ they might need to create in communications in target
languages. Ideally, a Tone of Voice would be worked out independently for each
language of application, and specied at a level of linguistic detail that enabled
translators to use it for guidance on their written style.
40    
Notes
1 I am grateful to several people, linguists and others, who have helped rene
the framework reported here, particularly those who have helped apply it to
commercial branded texts. anks in particular to John Bateman in Bremen,
Susana Murcia-Bielsa in Madrid, and David Lewis of Enterprise IG Informa-
tion Design. I would also like to thank four anonymous reviewers for their
detailed and most helpful comments on a rst dra of the paper.
2 Personal communication, Enterprise IG internal presentation, 2001.
3 Design Week, 26 January 2006, page 6: ‘Building Society strategy by Lloyd
Northover’.
4 Design Week, 2 February 2006, page 4: ‘e Core tries to instil personality
into Argos brand’.
5 WPP is a group of more than 125 communications services companies
(including, but not limited to, marketing, advertising, branding, and media)
in 106 countries, employing over 91,000 people worldwide. Key oce details
and contacts can be found from the WPP Group company directory at www.
wpp.com/directory.
6 TXU Energis brand values are no longer available on-line, due to its merger
with Powergen (now E.ON) in the UK. Orange brand values are available on
the Orange web site. http://www.orange.com/English/aboutorange9/philoso-
phyandvalue2.asp?bhcp=1.
7 See for example http://www.brandchannel.com/features_eect.asp?pf_
id=281 for a description of brand Tone of Voice from the branding agency
Interbrand.
8 Research undertaken by Enterprise IG Information Design in 2003 with 120
users of a government service, involving detailed language questionnaires
and interviewing of both department sta and the public. e research is,
however, subject to condentiality.
9 Although the brand is not explicitly named in this case, this kind of referen-
tial link only works when the concept that is being re-specied is active in
short-term memory. is accounts for its categorisation of ‘strong evocation
– it can only be successfully used when the concept is strongly evoked.
10 Interestingly, BT Cellnet has three explicit mentions in running text. ey
oer a free dedicated phone to HSBC customers. BT Cellnet is therefore the
only brand name mentioned in body text.
11 Stubbs (1996) discusses a judges use of these forms in a courtroom and a
jury’s likely interpretations of them: you may think that… for example, is
interpreted as I am allowing you to think that… rather than you might think
that…because of the jury’s assumptions about the authority of the judge and
their uncertainty about their own role and rights.
12 I am grateful to Alastair Gill for introducing me to this literature.
.  41
References
Allen, T. and Simmons, J. (2003) Visual and verbal identity. In R. Clion and J. Simmons
(eds) Brands and Branding 113–26. London: Prole Books/e Economist.
Bullmore, J. (2000) Why Every Brand Encounter Counts: seductive, anarchic, or cata-
strophic. London and New York: WPP Group PLC.
Bullmore, J. (2001) Posh Spice and Persil. e Brands Lecture, British Brands Group. www.
britishbrandsgroup.com
Cameron, D. (2000) Good to Talk? London: Sage.
Campbell, R. and Pennebaker, J. (2003) e secret life of pronouns: exibility in writing
style and physical health. Psychological Science 14: 60–65.
Carter, R. (ed.) (1982) Language & Literature: an introductory reader in stylistics. London:
Longman.
Carter, R. and Nash, W. (1990) Seeing through Language. Oxford: Blackwell.
Carter, R. and Simpson, P. (eds) (1989) Language, Discourse and Literature: an introduc-
tion to discourse stylistics. London: Unwin & Hyman.
de Chernatony, L. (1993) e seven building blocks of brands. Management Today March:
66–8.
Christie, F. and Martin, J. R. (eds) (1997) Genres and Institutions: social processes in the
workplace and school. Sydney: Cassell.
Clark, K. (1992) e linguistics of blame: representations of women in e Suns reporting
of crimes of sexual violence. In M. Toolan (ed.) Language, Text and Context: essays in
stylistics 208–26. London: Routledge.
Cook, G. (2001) e Discourse of Advertising. (Second edition) London: Routledge.
Delin, J. (2000) e Language of Everyday Life. London: Sage.
Delin, J., Bateman, J. and Allen, P. (2002/3) A model of genre in document layout.
Information Design Journal 11(1): 54–66.
Dewaele, J.-M. (2002) Psychological and sociodemographic correlates of communication
anxiety in L2 and L3 production. International Journal of Bilingualism 6: 23–8.
Fairclough, N. (1989) Language and Power. London: Longman.
Fairclough, N. (1992) Discourse and Social Change. London: Polity.
Fairclough, N. (1994) Conversationalization of public discourse and the authority of
the consumer. In R. Keat, N. Whitely and N. Abercrombie (eds) e Authority of the
Consumer. London: Routledge.
Floch, J.-M. (2000)
Visual Identities. London: Continuum.
Fowler, R. (1991) Language in the News. London: Routledge.
42    
Fowler, R., Hodge, R., Kress, G. and Trew, A. (1979) Language and Control. London:
Routledge & Kegan Paul.
Geis, M. (1982) e Language of Television Advertising. New York: Academic Press.
Gill, A. J. (2003) Personality and language: the projection and perception of personality in
computer-mediated communication. PhD thesis, University of Edinburgh.
Gill, A. and Oberlander, J. (2003) Perception of e-mail personality at zero-acquaintance.
Extraversion takes care of itself; neuroticism is a worry. Proceedings, 25th Anniversary
Conference Cognitive Science Society. Boston, MA.
Goddard, A. (1998) e Language of Advertising. London: Routledge.
Goldberg, L. R. (1993) e structure of phenotypic personality traits. American
Psychologist 48: 26–34.
Goodman, S. (1996) Market forces speak English. In S. Goodman and D. Graddol (eds)
Redesigning English: new texts, new identities. Milton Keynes and London: Open
University/Routledge.
Goodman, S. and Graddol, D. (eds) (1996) Redesigning English: new texts, new identities.
Milton Keynes and London: Open University/Routledge.
Graybeal, A., Sexton, J. and Pennebaker, J. (2002) e role of story-making in disclosure
writing: the psychometrics of narrative. Psychology and Health 17: 571–81.
Gundel, J., Hedberg, N. and Zacharski, R. (1993) Cognitive status and the form of refer-
ring expressions in English. Language 69: 274–307.
Habermas, J. (1984) eory of Communicative Action. Volume 1. (Translated by T.
McCarthy) London: Heinemann.
Halliday, M. A. K. (1994) An Introduction to Functional Grammar. (Second edition)
London: Edward Arnold.
Halliday, M. A. K. and Hasan, R. (1976) Cohesion in English. London: Longman.
Halliday, M. A. K. and Hasan, R. (1989) Language, Context, and Text: aspects of language
in a social-semiotic perspective. Oxford: Oxford University Press.
Hancock, J. T., Boucher, E. and Dunham, P. J. (2003) Impact of conversational perspective
and interpersonal relationship on the perception of irony. Poster presented at 44th
annual meeting of the Psychonomic Society, Vancouver, BC (abstract p. 44).
Harris, R. and Seldon, A. (1962) Advertising and the Public. London: e Institute of
Economic Aairs.
Hemenover, S. H. (2003) Individual dierences in rate of aect change: studies in aective
chronometry. Journal of Personality and Social Psychology 85: 121–31.
Henriksen, L., Jongejan, B. and Maegaard, B. (2004) Tone of Voice. VID report No. 4,
Center for Sprogteknologi, University of Copenhagen.
Iedema, R., Feez, S. and White, P. R. R. (1994) Media Literacy. Sydney: Disadvantaged
Schools Programme, NSW Dept of School Education.
.  43
John, O. P. (1990) e ‘Big Five’ factor taxonomy: dimensions of personality in the natural
language and in questionnaires. In L. A. Pervin (ed.) Handbook of Personality: theory
and research. New York: Guilford.
Joshi, Y. (2003) Communicating in Style. Delhi: e Energy and Resources Institute.
Koller, V. (2004) Metaphor and Gender in Business Media Discourse: a critical cognitive
study. London: Palgrave Macmillan.
Kress, G. and van Leeuwen, T. (1996) Reading Images: the grammar of visual design.
London: Routledge.
van Leeuwen, T. and Kress, G. (1995) Critical layout analysis. Internationale
Schulbuchsforschung 17 25–43. Frankfurt: Diesterweg.
Macken-Horarik, M. and Martin., J. R.
(eds) (2003) Negotiating heteroglossia: social
perspectives on evaluation. Text, 23, Berlin and New York: Mouton de Gruyter.
Mann, W. C. and ompson, S. A. (1987) Rhetorical Structure eory: a theory of text
organization Technical Report Number RS-87-190 203–11. Marina del Rey, CA: USC/
Information Sciences Institute.
McCrae, R. and Costa, P. (1997) Personality trait structure as a human universal.
American Psychologist 52: 509–16.
McLean, R. (1980) e ames and Hudson Manual of Typography. London: ames and
Hudson.
Michel, S. and Ettenson, R. (2005) e value and validity to organizations of measuring
brand strength: an independent collaborative investigation with a top 5 global brand
consultancy. Paper presented at the 10th International Conference on Corporate
Marketing Communications (CMC), Intercollege, School of Business Administration,
Nicosia, Cyprus, April.
Mijksenaar, P. (1997) Visual Function: an introduction to information design. Del: Del
University.
Myers, G. (1994) Words in Ads. London: Edward Arnold.
Newman, M., Pennebaker, J., Berry, D. and Richards, J. (2003) Lying words: predicting
deception from linguistic styles. Personality and Social Psychology Bulletin 29: 665–75.
Olins, W. (2003) On Brand. London: ames and Hudson.
Oxman, T., Rosenberg, S., Schnurr, P. and Tucker, G. (1988) Diagnostic classication
through content analysis of patients’ speech. American Journal of Psychiatry 145:
464–68.
Pennebaker, J. W. and King, L. (1999) Linguistic styles: language use as an individual dif-
ference. Journal of Personality and Social Psychology 77: 1296–312.
Prince, E. (1981) Toward a taxonomy of Given-New information. In P. Cole (ed.)
Radical
Pragmatics. New York: Academic Press.
Simmons, J. (2000) We, Me, em & It. London: Texere.
Simmons, J. (2003) e Invisible Grail. London: Texere.
44    
Simmons, J. (2004) Dark Angels: how writing releases creativity at work. London: Cyan.
Stubbs, M. (1996) Text and Corpus Analysis. Oxford: Blackwell.
Swales, J. (1990) Genre Analysis. Cambridge: Cambridge University Press.
Varey, R. (2002) Marketing Communication: principles and practice. London: Routledge.
Vestergaard, T. and Schroder, K. (1985) e Language of Advertising. Oxford: Blackwell.
Walker, S. (2001) Typography and Language in Everyday Life. London: Longman.
Waller, R. (1987) e Typographical Contribution to Genre. Unpublished PhD dissertation,
Department of Typography and Graphic Communication, University of Reading.
Waller, R. and Delin, J. (2003) Co-operative brands: the importance of customer informa
-
tion for service brands. Design Management Journal 14(4): 63–9.
White, P. R. R. (1998) Telling Media Tales: the news story as rhetoric. PhD dissertation,
University of Sydney.
Williamson, J. (1978) Decoding Advertisements. London and Boston: Miriam Boyars.
... To this end, the paper exploits several classification algorithms to check the brand persona of a content. Delin [6] describes how brand personality and brand value are built based on brand content. This research on content articles published by brands follows four linguistic frameworks: chains of reference, participant roles, presupposition and assumption, and tenor. ...
... We extract the following set of linguistic features from each article. We use the concepts proposed by Delin [6] to formulate a number of linguistic features, which aim to capture the trait of the underlying article in a more compact form. ...
... The key elements in this process are the noun phrases in the text, whose content and form can serve either to strongly evoke a brand, reinforce it, or not evoke it at all. Different kinds of relations that can hold between noun phrases and brand concepts as categorized by [6] have been summarized in Table 4. For our model, we use repetition, partial repetition, co-reference, and possessive inferrable as four different features. ...
Conference Paper
Brands produce content to engage with the audience continually and tend to maintain a set of human characteristics in their marketing campaigns. In this era of digital marketing, they need to create a lot of content to keep up the engagement with their audiences. However, such kind of content authoring at scale introduces challenges in maintaining consistency in a brand's messaging tone, which is very important from a brand's perspective to ensure a persistent impression for its customers and audiences. In this work, we quantify brand personality and formulate its linguistic features. We score text articles extracted from brand communications on five personality dimensions: sincerity, excitement, competence, ruggedness and sophistication, and show that a linear SVM model achieves a decent F1 score of $0.822$. The linear SVM allows us to annotate a large set of data points free of any annotation error. We utilize this huge annotated dataset to characterize the notion of brand consistency, which is maintaining a company's targeted brand personality across time and over different content categories; we make certain interesting observations. As per our knowledge, this is the first study which investigates brand personality from the company's official websites, and that formulates and analyzes the notion of brand consistency on such a large scale.
... will show, brand consulting, document design and corporate communications are established industries that increasingly rely on linguistic expertise. Yet hardly any of this work has been published (but see Delin, 2005;Ereaut & Segnit, 2006), partly for reasons of confidentiality and partly because few professionals straddle academia and the private, public or third sectors. ...
... My answer would be in the affirmative: while intervening in a company's discourse does not change the broad economic structures in which it operates, any increase in respect towards, and agency for, stakeholders is to be welcomed. To take but one example, Delin (2005) has shown how 'deeply branded' utility bills not only help with corporate impression management but are also consumer-friendly, helping customers to make informed decisions about budgeting and price plans. Consulting work can also involve, for instance, reducing gender stereotypes when marketing financial products to affluent women, and while such work does not address the economic, political and social structures facilitating gender pay gaps and female poverty, it does reduce the reproduction of sexist concepts in one particular domain. ...
Article
This paper argues that linguistic skills and awareness are essential requirements for professionals whose work centres on language as a product. Brought about by the commodification of language in developed economies, language work such as brand consulting, text design or online marketing requires linguistic knowledge and resources that many current teaching materials do not provide. Extracts from interviews with a diverse group of language workers allow for first insights into their kind and level of language awareness, but also show that they are more concerned about what they perceive as a lack of language awareness in their clients. This finding suggests a non-linear model of teaching and learning relations between academic linguists, language workers, clients and students. The paper further discusses the options that applied linguists in academia have if they want to work with/as language workers and argues that engaging with language work(ers) can be an opportunity to bring critical language awareness and discourse analytical skills to bear on professional practice and training.
Article
A consumer-dependent (business-to-consumer) organization tends to present itself as possessing a set of human qualities, which is termed the brand personality of the company. The perception is impressed upon the consumer through the content, be it in the form of advertisement, blogs, or magazines, produced by the organization. A consistent brand will generate trust and retain customers over time as they develop an affinity toward regularity and common patterns. However, maintaining a consistent messaging tone for a brand has become more challenging with the virtual explosion in the amount of content that needs to be authored and pushed to the Internet to maintain an edge in the era of digital marketing. To understand the depth of the problem, we collect around 300K web page content from around 650 companies. We develop trait-specific classification models by considering the linguistic features of the content. The classifier automatically identifies the web articles that are not consistent with the mission and vision of a company and further helps us to discover the conditions under which the consistency cannot be maintained. To address the brand inconsistency issue, we then develop a sentence ranking system that outputs the top three sentences that need to be changed for making a web article more consistent with the company’s brand personality.
Chapter
Full-text available
This primacy of the semiotic clearly suggests the usefulness of discourse-based approaches for branding scholarship (Turlow & Aiello, 2007:310), even if branding —like other aspects of corporate communication (Koller, 2008b:155)— has received relatively little attention from critical discourse scholars, and no comprehensive theorisation of “the connection between a particular brand position and the multiplicity of visual and graphical resources that brand designers choose to express it” (Delin, 2007:5) is available. Mumby (2014) argues that only a focus on the discursive processes through which brands are constructed, disseminated and negotiated can bridge the gap between organisational studies oriented to the strategies organisations employ to achieve their desired market position and cultural studies that explore the meanings that commodities acquire in the lived experience of consumers. In the following sections, I first seek to conceptualise branding as a semiotic activity; I then provide an overview of critical-discursive approaches to the branding of products, places and politics; and, finally, illustrate these approaches through an analysis of the practices employed by the John Lewis retail chain to manage internal audiences' perception of the brand.
Technical Report
Full-text available
abstraction:instance whole:part process:step object:attribute 8It is also anomalous in another way: the widely used pattern of presenting a problem and its solution does not occur in this text The Conditional Schema --- 6; T