Peddling a semiotics of fear: a critical examination of scare tactics and commercial strategies in public health promotion

In this study, we critically examine the ways in which a nationwide health promotion campaign – the 2013 Diabetes UK/Tesco diabetes campaign, the largest of its kind in the UK – seeks to raise the general public's awareness of Type 2 diabetes. We subject a series of six campaign images (including their layout and accompanying text) to a multimodal discourse analysis, identifying the presence of a range of fear-inducing, stigmatising and commercial strategies, through which the campaign emphasises the dangers of diabetes and advocates personal responsibility for assessing both individual and others' risk of the disease. Specifically, we describe, in multi-semiotic detail, three discursive techniques deployed in the campaign to achieve these ends: (1) the depiction of grief and amplification of diabetes-related danger, (2) the promotion of diabetes risk and localisation of individuals' responsibility for their health and (3) the commercial branding and framing of the Diabetes UK/Tesco partnership – including the promotion of goods and services – as a means of diabetes prevention and management. Our findings raise concerns about the moral legitimacy of using fear-inducing and commercial strategies in order to (effectively) raise public awareness of and responses to Type 2 diabetes, strategies which do little to address the environmental factors which are associated with increasing rates of the disease.


Introduction
Diabetes mellitus (henceforth diabetes) is a condition which affects over 371 million people worldwide (International Diabetes Federation [IDF 2012]). In the UK alone, over 3 million people are estimated to have the disease (Public Health England 2013). The prevalence of diabetes is growing and is estimated to rise to 552 million worldwide by the year 2030 (IDF 2012). Although the disease can have significant biological, personal and social consequences, people with diabetes who manage the condition effectively are able to live fairly normal lives, and achieve normal life expectancy (Lutgers et al. 2009).
A number of risk factors are associated with developing diabetes. 1 Epidemiologists have related the increasing rates of the disease to increasing rates of obesity (Hill et al. 2003), attributing, in turn, the prevalence of these two health problems to environmental factors, to wit, recent societal and lifestyle changes (Butland et al. 2007). For example, changes over the last five decades in patterns of work, transport, and food production and sales have contributed to an "obesogenic environment" -the influences that individuals' surroundings and life opportunities have on promoting obesity (Butland et al. 2007). In

Health promotion, risk and the rhetoric of fear
Health promotion is a pervasive, clamouring form of contemporary public discourse, realised in and through a range of semiotic means. Nowadays, one encounters mass media health promotion messages in virtually every kind of public space (e.g. shopping centres, libraries, restaurants, supermarkets, pharmacies, car parks, bus stops, billboards and other poster sites) and in a variety of material forms and genres (radio, television and cinema broadcasts; newspaper and magazine articles and advertisements; leaflets and posters; and health-related and charity websites). Indeed, such is the relentless ubiquity of health promotion texts in everyday life that individuals are regularly and unavoidably exposed to a plethora of health education messages, ranging from exhortations to monitor and enhance their well-being, through to messages that warn the public about specific risks to health. Although present in many guises and performing a number of functions, health promotion discourse can broadly be defined as a form of communication which seeks "to inform and persuade intended audiences to change habits or adopt new routines" (Finan 2002, 16). Accordingly, health promotion campaigns are pedagogical in function, with campaigners assuming a position of knowledge and authority over others, targeting populations who are perceived to be in need of instruction and information (Lupton 1995).
In recent times, at least from the turn of the twenty-first century, health promotion has changed its emphasis from curing and containing disease to inciting people to take personal responsibility for maintaining their health (Breslow 1999;Lupton 2003;Nettleton 2006). This shift can be situated within the wider movement towards a neoliberal model of public health in Western societies. The logic behind such an approach is premised on the idea that rates of illness will be reduced if individuals can only be persuaded to practise self-care, that is, modifying their lifestyles in accordance with healthy living advice and thereby exerting control over their bodies (Lupton 2003, 25). Such an individualist stance arguably reflects the neoliberal approach to public health whereby the onus for well-being is placed firmly on the shoulders of the self-determining citizen, a corollary of which is to absolve the government of responsibility towards the health of its citizens (Kemshall 2002, 42;Inthorn and Boyce 2010, 84). The maintenance of health thus becomes the concern of the individual, rather than the state. With its laissez-faire, individualistic values, neoliberalism's emphasis on assuming personal responsibility for health downplays the significance of environmental factors in the onset of disease and health problems, and also ignores the complicated genetic aetiologies of conditions like diabetes.
The reconfiguration of attitudes towards this neoliberal ideal is accomplished, in no small way, through public health education campaigns, such as the one we analyse in this study. First, such campaigns extend individuals' capabilities, rights and responsibilities in relation to their wellbeing by providing them with the knowledge and pedagogical tools (e.g. do-it-yourself risk assessments) to "stay healthy" and prevent ill-health from occurring. Yet at the same time such campaigns are also very restricting, by stressing the link between disease and individual lifestyle behaviours, and bombarding the public with constant reminders of what they should not do: condemning, for example, the consumption of high-fat and low-nutrition food, unprotected and promiscuous sexual activity, and excessive alcohol consumption. This so-called "healthist" philosophy compels people to constantly strive for "perfect" health, and hence to conduct all aspects of their lives in ways which, first and foremost, improve their health (Crawford 1980; see also Burchell [1996, 29] on the notion of "responsibilisation"). These principles have become the major thrust of almost all contemporary public health campaigns, and further strengthen the influence of neoliberal ideals on the way we think about health and disease (Brown and Baker 2012, 17-18).
Yet health-related action takes place, of course, in a specific social, political and economic context, a cultural context which inevitably influences individuals' health behaviours. For example, conditions such as obesity and Type 2 diabetes, which are potentially preventable through lifestyle choices (Zimmet, Shaw, and Alberti 2003), are not solely the problem of the individual, but are also directly connected to broader commercial and environmental factors, such as the rising sugar content in prepared-food and the increasing availability, and extensive variety of, cheap high-fat, high-calorie and hyper-palatable foods, 4 along with the relentless marketing and advertising of such lowgrade produce (Blythman 2012) -factors that have led to increased sales and consumption of energy-dense foodstuffs, in turn contributing to an obesogenic environment (Inthorn and Boyce 2010, 90). Rather than responding to these cultural and economic realities (by, say, introducing regulatory measures on sugar and salt content in food or stricter rules on the advertising and promotion of junk foodresponses, of course, liable to shrink profit margins for the food industry and retailers alike), many health promotion policies continue to emphasise personal responsibility for health, even though the shortcomings of the self-deterministic approach are cast into sharp relief when environmental factors such as the aforementioned are taken into consideration (Minkler 1999, 128).
The issue of encouraging self-determining health action raises the question of which health promotion strategy best serves this neoliberal ideal. Successful communication in health promotion comprises the elements of gaining and retaining the audience's attention, and ensuring their correct interpretation of the intended message (Finan 2002, 16). However, it is a point of contention which approach is precisely the best way of realising this notoriously difficult-to-achieve goal. Health promoters have drawn upon a repertoire of persuasive strategies to encourage the public to adopt certain behaviours and to raise their awareness of risk, the most common being fear appeals and the "unvarnished" presentation of facts (Monahan 1995, 81). Despite the documented benefits of using more positive appeals to promote health, employing, for instance, supportive messages that underscore hope, reward and positive consequences (Hastings, Stead, and Webb 2004;Rothman et al. 2006), arguably the most commonly (and certainly the most controversially) used strategy in recent times is that of scare tacticsdiscourses designed to elicit fear and anxiety in people (Hill, Chapman, and Donovan 1998;Hastings, Stead, and Webb 2004). The appeal, as it were, of the fear appeal is that it is deemed to be an effective means of securing attention and provoking attitude and behaviour change (Sutton 1982;Borland and Balmford 2003). A premise at work in many mass media campaigns that draw on shock tactics can, if crudely, be formulated as follows: the greater the amount of fear that can be aroused in the audience, the greater the audience's intention to execute the recommended course of action. For example, antismoking advertisements that feature graphic images in lurid anatomical detail (pictures of expansive malignant tumours, dark tar-coated lungs, sticky arterial linings and so forth) charge facts with emotion and potentially render them much more impactive, hence motivating change in health behaviour (Borland and Balmford 2003, 45).
However, the use of shock tactics is delicately double-edged. Both the ethics and effectiveness of fear-arousing communications in health promotion are disputed. For example, there might be unintended adverse effects in using emotionally charged, fearinducing campaigns which, in failing to be wholly accurate and sincere, potentially exaggerate risks to the extent of scaring and misleading people (Guttman and Salmon 2004). Further, a number of researchers in public health argue that fear-arousing health messages are more likely to be effective only when the message recipients have a high degree of self-efficacy. In other words, individuals who are better placed, socially and psychologically, to respond to persuasive messages are better able to act on and profit from health promotional discourses (Hastings, Stead, and Webb 2004, 975). For those individuals who are less socially and psychologically able to do so, fear-inducing messages might well make them feel worse, engendering, for example, feelings of anger and defensiveness (2004,975).
Although not dismissing fear in health promotion per se, we take a critical view of the particular use of fear in the Diabetes UK/Tesco campaign, viewing its use of scare tactics as a glib, quick-fix solution to the problem of effectively raising public awareness of diabetes. Moreover, the use of fear, we will argue, is liable to be counter-productive, tending to alienate, stigmatise and guilt-load certain groups of people who are believed to be more susceptible to the disease. And when fear is harnessed not solely to promote awareness but is also used for apparent commercial reasons, the moral legitimacy of such a promotional strategy is further (and more urgently) called into question.
A multimodal approach to contemporary health promotion discourse As with other traditionally non-commercial aspects of society, health and health promotion have become increasingly colonised and influenced by commercial discourses and interests (Chouliaraki and Fairclough 1999). Ritchie, Swami, and Weinberg (1999) note how, originally, health charities and other so-called not-for-profit organisations were put-off commercial discourse, due to its association with corporate capitalism and profitmaking. However, forced to compete for audiences' attention with countless other commercial and non-commercial organisations (what Schroeder terms the "battle of the brands" [2008,278]), these organisations eventually began to harness the linguistic, semiotic tools of commercial advertising. Indeed, as Lupton (2003) observes, such persuasive commercial advertising techniques have become a central strategy of much contemporary health promotion and education. In particular, the use of arresting, visceral visual imagery is increasingly common, since such semiotic elements have been shown to influence the public's uptake of a particular promotional message, and help to send people along a more emotive pathway than might be accomplished by health promotion texts which are strictly verbal in communication (Joffe 2008, 84).
Despite their common use in health promotion discourse, however, the pictures used in health communication texts have often been left, from a visual semiotic perspective at least, to "speak for themselves" (Singh 2007, 134), while textual elements have principally been examined in terms of their content, rather than the subtle ways in which they convey meaning. One important reason for this semiotic neglect is because of the deeply entrenched tradition of treating pictures (at least in risk and health communication) as little more than illustrations of the surrounding, anchoring text, a belief which fails to recognise and appreciate the encompassing and autonomous meaning potential of other semiotic modes besides language (Ferreira, Boholm, and Lofstedt 2001, 283). The model of critical multimodal discourse analysis we employ in this study combines both visual and textual analysis, recognising the interplay between text and image as well as the propensity for images (photographs in particular) to consistently induce interpretations of their own. Our analytical approach is underpinned by the principles of critical discourse analysis (CDA) and a social semiotic theory of multimodal communication. We view language as just one semiotic mode of communication, alongside images, layouts and sounds, etc., that are available to text creators. Meaning is the product of the interplay between these various semiotic modes, in and across texts, which each have their own cultural and social histories (Machin and van Leeuwen 2007, 158-159;Bezemer and Kress 2008). The meanings of modes are not fixed, but instead carry what Barthes terms "floating chain[s]" (1977, 39) of meaning potentials, which are open to both the text creator and audience. We must acknowledge, therefore, that while we offer our own interpretation of these texts, this may not be shared by other audiences, or indeed the texts' creators. Meaning is a necessarily contestable concept which exists only in the ways in which it is "made" by individuals, a process that is subject to social and cultural influences (Chandler 2007, 195;Martinec and van Leeuwen 2009;Bezemer and Jewitt 2010).
The analytical approach we take can be divided into two complementary aspects: the analysis of visual choices and the analysis of lexical choices. When considering visual choices, we pay close attention to the photographic and layout choices made in the design of these texts. We draw on van Leeuwen's (1996, 2006) "visual grammar", an approach which views a text's visual components, or modes, as working collaboratively, in a rule-based system, or grammar, to convey messages. Following others, such as Machin and Mayr (2012), we combine the aforementioned analysis of visual choices with linguistic analysis. Specifically, we investigate the dominance of particular types of words and broader lexical fields in these campaign images. Just like visual elements, we view lexical choices as made (1) from a number of other possible options and (2) in a way that benefits the text creators' own, or their institution's, social and ideological preoccupations (Blommaert and Bulcaen 2000).

General semiotic characteristics of the campaign images
Although the Diabetes UK/Tesco campaign makes use of a range of media, in this paper we focus on its most prominent and richly multi-semiotic mode of communication: that of the billboard poster and leaflet. Apart from their difference in size (the poster being a larger version of the leaflet), these two text types are near-identical respecting content (i.e. layout, image and linguistic text), and so we will use the general term "campaign images" when referring to them. A total of six photographs featured in these campaign images and these, ideally, need to be considered together in order to appreciate fully the communicative import of the campaign as a whole. Nevertheless, each image also functions as an autonomous text in itselfdesigned to persuasively communicate the dangers associated with Type 2 diabetes and to encourage people to check their risk of developing the disease. Accordingly, we first consider some of the general semiotic features common to all of these campaign images, before interrogating their more specific visual and linguistic aspects.
Arguably, the most noticeable similarity across these images is the use of studiostaged photographyand the particular configuration of social actors and scenes so photographed. Photography, as John Berger (2001, 217) reminds us, deals in "the language of events", with the camera bestowing "authenticity upon any set of appearances, however false" (1982,97). Although a significant amount of arrangement by the photographer will have taken place before the campaign photographs were taken (and of course editing afterwards), these images, by dint of their being sharp-focused, colour photographs, as opposed to drawings or illustrations, possess a high degree of visual modality, and are hence presented as a seemingly "naturalistic, unmediated, uncoded representation of reality" (Kress and van Leeuwen 2006, 158). The viewer, in other words, is being asked to see the participants, situations and events depicted in these photographs, albeit studio staged and constructed, as inherently plausible and truthful, reliably presenting some kind of recognisable and tangible reality (specifically the hard, harsh reality of living with and reacting to diabetes) in a naturalistic way.
The depiction of authenticity in these six imagesthe credible, real-life scenarios the photographs present to viewersis emphasised by the choice of settings in which the participants appear. Although two of the images (Figures 1 and 4) possess blurred, decontextualised backgrounds (from which, viewing them in isolation, it is difficult to identify with complete certainty a specific context: a hospital or domestic scene, say), the remaining pictures display identifiable backgrounds with distinguishing features: a lamp on a desk and electric socket in one image, an internal (living room?) door, a dining table in the others. These "objects" (Barthes 1977, 22-23) help to signify a domestic context, suggesting that the events being played out are taking place in an intimate, familiar everyday setting. There is no evidence of medical intervention in these images, no sense that we are witnessing a scene taken from an impersonal, institutional setting such as a doctor's surgery or hospital. The problem of diabetes, these images suggest, is one managed primarily by sufferers and their families alone, not through the state support of medical professionals, an ideological underpinning in keeping with the neoliberal ideal.
Each campaign image features a highly individualised, close-up shot of two represented participants, who, we believe, are supposed to depict close relatives (e.g. father and son in Figure 2, father and daughter in Figure 3). This design choice is evidently intended to individualise the problem of diabetes for viewers of these images, enforcing the neoliberal notion that this condition is one that is solely managed by the individual sufferer and their close familyand one that the individual is responsible for preventing in the first place. Moreover, the represented participants appear in an "offer" picture (Kress and van Leeuwen 2006), that is, a picture in which the participants' gaze is not directed at, and hence does not visually address, the viewer (see, for example, Figures 1 and 2). The participants' looking away off frame invites the viewer to imagine what these individuals are thinking and experiencing (Machin and Mayr 2012, 72), to ponder their predicaments and reconstruct their stories. A feature of most narratives of illness is that they commence at the point of diagnosis (Radley 1999, 781), and indeed it might well be the case that what the viewer sees in all of these images are people variously reacting to a diagnosis of diabetes, those anguished moments in which bad news is broken. There are other possibilities, though. We might, for example, be witnessing the consequences of diabetes further down the line, or else at some other significant kernel in these participants' personal illness trajectories.
The images are thus compellingly ambiguous and open to interpretation, detaining the viewer and compelling him or her to actively search for and recover some sort of meaning. Yet arriving at a final, clear-cut interpretation is not straightforward. Each photograph appears to have been, in Berger and Mohr's (1982) words, "taken out of a possible continuity, or set of life stories, and is hence depleted of history" (91). There are no words that supply a particular, authoritative interpretation or produce the effect of certainty; no captions to anchor, to precisely illustrate the scenes depicted. Whatever the case, the search for meaning in these pictures illustrates Radley's (2002) point that the visual portrayal of illness suffering "does not merely find its meaning but retains it through its reconstruction of the sensuous experiences of those who contemplate the image" (19). In other words, viewers of these images produce their own particular context to make sense of them, and in the process play out, as it were, the suffering for themselves. Indeed, it is this very process of the viewers' vicariously playing out and experiencing the fallout of diabetes upon which the campaign so heavily depends, both as a means of inducing fear and prompting actiona theme which we develop in greater detail in the following sections.

The danger of diabetes and depictions of grief
Through their depictions of shock, anguish and grief, each of the six images confronts the viewer with the possibility of loss and death. Looking at these intensely, if artificially, emotionally laden pictures, deliberately instinct with overtones of bereavement and mourning, one would be forgiven for thinking that to be diagnosed with diabetes is to be subject to nothing less than some kind of incontrovertible death sentence. The pathos of these constructed-as-harrowing scenes is expressed in the faces, gestures and postures of the actors. In Figure 3, for example, the participants pose as the (presumably) mutually devastated father and daughter.
Here, the actors' heads are angled inwards towards one another, converging, touching, as if in reciprocal support. A tear from the daughter's left eye, clearly visible in the brightly illuminated setting, streaks down the length of her cheek to underneath her chin. The participants appear not to be speaking, but rather inhabit a stunned silence, numbed, as it were, by the awfulness of their situation, a predicament (whatever it precisely is), presumably, too dreadful for words. Similarly, in both Figures 4 and 5, two people (again, we assume, family members) are depicted in a close, consoling embrace, however in these photographs the embrace is much more physical, firmer, deepera hug rather than a touching of heads. Figures 4 and 5 clearly aim to portray the emotional distress of a family (father, mother, daughter) reeling from the terrible impact of diabetes. In both photographs, one participant's back is squarely positioned towards the camera: we see only the back of their head and upper back, their bodies pressed into the other participants' chests. Over the left shoulders of these away-facing participants appear the agonised visages of the mother and father: both rheumy-eyed, the mother's mouth distressfully agogtwo undeniably disconsolate looks which, since theirs are the only faces presented, the only countenances we can read and interpret, mournfully characterise the tableaux entirely. These scenes are made all the more tragic and anxiety-inducing by virtue of the fact that this is an entire familynot just an individualthat is being devastated by diabetes. This aspect of the image is emphasised further by the wedding ring, clearly visible on the mother's left hand, an object which symbolises family ties (and the possibility of their being undone), quietly stressing the tragedy and pathos of the scene.
Quite clearly, then, these visual tableaux draw on conventional pictorial formula for grief: the clear distortion of physiognomic forms, the tearful eyes and downward cast gazes (Barasch 1987). Yet despite this rather stereotypical form of grief representation, the photographs nonetheless remain powerfully connotative. To see people crying is, of course, to infer that they are in a certain mood, a visual spectacle at once arresting and disquieting. As Berger and Mohr (1976, 116)

observe:
A man or woman who is sobbing reminds one of a child, but in the most disturbing way. This is partly because of the particular social convention which discourages adults […] from breaking into tears but permits children to do so […], [t]here is a physical resemblance between a sobbing figure and a child. The 'bearing' of the adult falls away and his movements are limited to certain primitive ones.
It is, moreover, worth noting that the act of viewing the campaign photographs of distressed, weeping people involves (symbolically at least) the camera invading the private space of others, revealing what is typically unobservable, a potentially discomforting act of trespass on the part of the observer. The probing, all-seeing gaze of the camera turns photographic subjects into objects (Nunn 2004, 275), a gaze which makes public the misfortune of others, exposing a threatening yet hidden aspect of life (Hockey 1997, 107). Contemplating these people in these specific contexts and configurations arguably contributes to the pity we are supposed to feel for them; the intrusive camera catches them unawares in their private misfortune, offering them to us as items of information or objects of contemplation (Kress and van Leeuwen 2006). They do not return our gazethey are unaware of our presence, our visual violation, casting their gaze off frame, elsewhere.
Whether or not these images (artificially composed as they are) have the desired effect of moving and unsettling the viewing public, they are clearly designed to portray diabetes as a dangerous, indeed fatal, disease rather than as, say, an eminently manageable condition with which people are able to live long and full lives. But it is not only these images that depict diabetes as a deadly disease with devastating consequences; they work in conjunction with other semiotic means to steer viewers towards such an understanding. For example, in Figure 2, reader-viewers are verbally enjoined to evaluate their risk of diabetes to avoid its serious consequences (or to quote from the text directly: "BEFORE IT HITS YOU AND YOUR FAMILY"). Although not necessarily obvious on a quick initial reading, this material action process, specifically the use of the dynamic transitive verb "hits", is an instance of metaphoric conceptualisation; that is, the understanding of one, sometimes unfamiliar and usually abstract thing in terms of another, more familiar and usually tangible thing (Lakoff and Johnson 1980). In this case, diabetes is conceptualised as a violent killer that "hits" sufferers (see, for instance, Wallis and Nerlich 2005). This militaristic trope conceptualises diabetes as something that is capable of vigorous and aggressive movement, wilfully colliding with the viewer and his/her family to cause sudden and "devastating" impact (the phrase "devastating consequences" appears in the main body copy of the advert). Furthermore, the role of the viewer (the potential diabetes sufferer, in this construction) is omitted, rendering the person with diabetes as a "passive victim" of the conditiondefenceless, perhaps, against this forceful and harmful agent of disease. 5 Since metaphors have a cognitive reality, and thus constitute "vehicles for understanding" (Reisfeld and Wilson 2004, 4024), this particular figurative construction might well instil in readers a sense of fear and helplessnessdiabetes as an unremittingly potent disease against which they are effectively powerless.

Promoting diabetes risk and localising responsibility for health
Complementing the portrayal of diabetes as a devastating disease with dreadful consequences is the way in which the campaign images localise this sense of dread and danger to viewers. We, the audience, are invited to consider ourselves as sufferers of the conditionto experience the represented participants' fear and anguish for ourselves. In viewing and seeking to make sense of these pictures, we are very much pitched into the private worlds of the characters portrayed. As we described earlier, a number of the photographs depict domestic settings. These everyday scenes help to construct a familiar, quotidian contextnot, for example, the alien environment of the hospital, but a common, mundane setting which the viewer will more likely be able to identify with and (literally!) feel at home in (see, for instance, Figure 6 in which the participants are positioned by a dining table). This intimate distance between the viewer and the participants also results from the use of close-up camera shots and the positioning of the viewer, frontally (and at eye level), rather than at an oblique angle with the photographed scene (Jenkins 2007, 97). This visual-spatial effect aligns the observer with the represented, including them directly, head-on, in the domain of the suffering participants. In fact, what we have is a kind of blurring, merging of worlds (the viewers' and the participants'), a visual configuration which suggests that diabetes is your (the viewers') problem toothat the disease is liable to affect you and your family in the same way that it has affected, and continues to affect, these people. In other words, the images promote the idea that you, the audience, live "in the shadow of serious illness" (Radley 1999, 780) that there (for the moment at least) but for the grace of God go you.
Although, as viewers, we are aligned very closely with the represented participants in these campaign images, we are given very little information regarding who they actually are. They resemble "generic" sufferers of diabetes. These people are neither named nor otherwise specifically referred to in the text which is co-extensive with the images. Nor does the accompanying text offer details of their biographies, their storieshow they came to find themselves in the situations in which they are depicted. Viewers have to fill in and guess at these details themselves. If any of the participants have diabetesand presumably we are meant see some of them as having and suffering from the diseasethe images alone do not make this clear. In fact, looking at these images in isolation from the accompanying text, one would be hard-pressed to actually identify any of these depicted scenarios as representative of a distinct diabetes illness narrative. Moreover, the deliberate blurriness and lack of focus in which certain participants are depicted (see, for instance, the older participant in Figure 2) means that the diabetic sufferer is not defined. Instead, viewers are permitted a degree of interpretative licence, and might envisage themselves, or a family member, in the role of either diabetes sufferer or familial carer.
Despite this lack of visual definition of and information about the represented participants in these campaign images, what is clear is that the diabetes sufferers depicted therein are ethnically and generationally diverse and are presented in a range of apparent family relationships/configurations: for example, Asian mother and daughter, Asian father and daughter, black father and son, white father and daughter, and white elderly husband and wife. When the images are viewed in conjunction with the text on the reverse of the leaflet versions of these texts (a series of bullet point questions which highlight the susceptibility of particular demographics for developing diabetes -"Are you over 40 years old?", "Are you South Asian, Black African or African-Caribbean?"), then it can be reasonably assumed that the viewer is being asked to make the connection between ethnicity and age and increased diabetes risk, and accordingly attribute the disease to the corresponding individuals in the images. Ethnicity and age thus, in these photographs, come to symbolise diabetes, the images equating certain groups of individuals with the disease, without, at the same time, highlighting the specific lifestyle and environmental factors that contribute to the onset of the condition. Furthermore, although the participants depicted in these images are ethnically diverse, they nonetheless represent quite a narrow proportion of the UK population: principally nuclear families (parents and children, no extended relatives such as grandparents); no ethnic diversity within the families (the participants in each image are only ever from one ethnic background and are dressed in typically traditional Western attire). The strong focus on the "at risk" ethnic groups, at the cost of representing the truly diverse nature of the UK population today (ethnicity aside)might potentially contribute to stigmatising such groups, signalling them out and showing them as victims rather than as, say, empowered individuals successfully managing and living with diabetes, which in reality many people do.
Yet as well as alluding to individuals ethnically and generationally more likely to be predisposed to diabetes, the campaign images also vaguely broaden out the notion of susceptibility to the disease, conveying the sense that no one is safeeveryone is at risk. This effect is achieved by the use of textual "anchorage" (Barthes 1977) which accompanies and supports the photographs, in particular the recurring headlines which appear at the top of each campaign image. Taking the grammatical form of an unmitigated imperative, these headlines take on an instructive tone, assuming authority over the audience by ordering us to "CHECK YOUR RISK OF TYPE 2 DIABETES", with the use of capitalisation and striking white bold lettering lending the text not only visual salience but also adding extra emphasis and urgency to its (already) peremptory message. Directly beneath this imperative statement, in contrasting black type (but again emphatically capitalised), appears the second part of the headline, a proposition which develops the premise set out in the clause above and points up the reasons for and consequences of not assessing one's risk of diabetes: "IF NOT FOR YOU FOR YOUR FAMILY", "BEFORE IT HITS YOU AND YOUR FAMILY", "SO YOU AND YOUR FAMILY DON'T SUFFER", "IGNORING IT AFFECTS THOSE CLOSEST TO YOU", etc. The second-person pronoun forms "you" and "your", which are repeated throughout these examples, can be interpreted in both their singular sense, picking out individual readers, the "you" reading the text, as well as their plural sense, thus incorporating a broader frame of reference, to include people in general (Harvey 2013). At the bottom of each advert, a conventional red and white warning sign contains, in bold, black lettering, the message: "CHECK YOUR RISK AT PHARMACIES OR ONLINE TODAY", another bold second-person injunction which further communicates the sense of imminent danger while reinforcing the urgent necessity for readers to act immediately. 6 Each advert provides a different reason why "YOU" (whoever you are) should "CHECK YOUR RISK", but at the heart of these injunctions is the recurring neoliberal idea of assuming personal responsibility of taking care not only for oneself but also one's family. The potential effect of these synthetically personalised 7 messages, we suggest, is that each individual viewer of these campaign images might feel as though they are being addressed, personally and directly, in relation to their and their family's diabetes risk. Furthermore, the pronoun "your" in the imperative construction "VISIT YOUR GP" (in all figures) transforms the referent "GP" into each individual viewer's own particular General Practitioner, thereby instructing each viewer, personally, to visit their own reallife GP. This personalised formulation can be contrasted with the hypothetical alternative "visit a GP"where the indefinite article "a", as opposed to the possessive pronoun "your", means that the referent of the phrase "a GP" is generic and non-specific, relating to any GP. Of course, one might read this alternative sentence and still think of his/her own GP. However, they would not necessarily be directed to do so in the same specific and direct way as in the actual sentence containing the localising possessive pronoun.
In short, these campaign images communicate to us that we (and our families) are personally, directly at risk of diabetes, and that we should consult our own GP because of this very real and inevitable threat to health. Consequently, readers of the adverts are responsibilised (Brown and Baker 2012) in terms of their health, and are constructed as being accountable for adverse outcomes that affect both their personal health as well as the health of their immediate family. The texts therefore reinforce a moral commitment to others' needs, guilt-loading the readers addressed by the texts (Guttman and Salmon 2004, 545). Indeed for readers who actually have a relative with Type 2 diabetes, the injunction to evaluate risk of the disease might well take on an accusatory, reproachable tonethe shame-inducing sense that they have failed in their moral obligation to care for their families.
Creating a brand and promoting a service: commercialising health promotion Establishing a brand has long been viewed as a feature of health charity and health promotion campaign material (Hevey 1992;Waltz 2013), and the Diabetes UK/Tesco campaign is certainly no exception. First, on the textual level (Halliday 1978), the Diabetes UK and Tesco logos are aligned, both spatially and in terms of colour, for the purpose of establishing an allied, composite "brand". They are both rich and striking in colour, particularly when set against the white background of the textbox graphic, drawing the attention of the viewer. The small "i" in the Diabetes UK logo can be contrasted against the rest of the lettering in the logo; it is in lower case, coloured a lighter shade of blue, and is more rounded (Lim 2004;van Leeuwen 2005). We interpret this "i" as standing out as the first-person pronoun form, subtly connoting the importance of the individual and, by extension, the support that Diabetes UK claim to provide to individual sufferers. When combined with the triplet-structure slogan located below the Diabetes UK logo, which reads "Care. Connect. Campaign.", itself resembling a commercial slogan in its design (Russell and Lane 1990), these co-occurring components convey a "brand image" of Diabetes UK, and by extension the partnership with Tesco, as organisations that are wholly benevolent in their focus on the health of the individual and, perhaps more specifically, in helping individuals by providing them with the tools (e.g. the online risk assessment and branded diet plans) to manage their diabetes risk and fulfill their obligations as "good" neoliberal subjects.
While we have already discussed the photographs featured in this campaign in greater detail earlier, it is useful to consider, briefly, the way in which these contribute to the Diabetes UK/Tesco partnership visual brand. Imagery, and in particular photography, play a big part in creating a brand. Schroeder (2008, 277), for instance, refers to a "visual turn in marketing", in which brand designers employ a "compelling visual rhetoric" to create a striking and memorable visual representation of the company or organisation with which they are associated. Charities and other non-profit organisations are often keen to distance themselves from the conventions of traditional corporate branding, such as highly saturated, colourful, almost unrealistic imagery (Tapp 1996, 335). Therefore, such organisations often use, as Diabetes UK and Tesco do here, less saturated, more "realistic" imagery, such as photographs, in their campaign material. Rather ironically, this style of image has itself come to constitute a type of commercial brandinga high modality genre of charity commercialism. This use of realistic, documentary-style photographs is often accompanied by colourful graphical elements akin to more traditional commercial advertising, such as the textbox located at the bottom of the texts we analyse here. Through this type of low-saturated, naturalistic photography, the Diabetes UK/Tesco partnership situates itself in what Evans (1999, 279) calls "the tradition of social documentary", deliberately branding itself as helping "real" people. The apparent authenticity of these photographs performs an important role in visual branding. In continuation of the theme of localising the diabetes risk to each individual viewer, these photographs invoke the sense that diabetes not only affects ordinary and normal people but also that ordinary and normal people, like each audience member, make use of the goods and services offered by the Diabetes UK/Tesco alliance. There is no greater product endorser than the ordinary person and, as Schroeder (2008, 282) notes, this visual element is able to promote brands as "authentic, to invoke the 'average consumer' as a credible product endorser, and to demonstrate how the brand might fit in with the regular consumer's lifestyle". This is a key function of any branding strategy since it is through such authentic and realistic imagery that brands, and in particular health-related and charity brands, which rely on consumer trust, present themselves as un-staged, genuine and trustworthy (Holt 2002;Elliott and Davies 2006).
An integral part of commercial discourse is also the way in which brands are framed as the means to happiness and a better lifestyle or, in this case, the solution to social or health-related problems. As Hevey (1992) contends, the mere presence of the charity logo will, to some extent, achieve this effect. However, we believe there to be a more intricate layout design mechanism at work here, namely, the ideal/real construction. According to Kress and van Leeuwen (2006), in commercial advertising, images or textual elements aligned on a vertical axis can relate to one another, with the top element (the ideal) presenting a possible future scenario and the bottom element (the real) presenting the means or information audience members require to either realise (if it is desirable) or avoid (if it is undesirable) the scenario presented above it. As Kress and van Leeuwen (2006, 186) put it, "[t]he upper section tends to make some kind of emotive appeal and to show us 'what might be'; the lower section tends to be more informative and practical, showing us 'what is'". The top elements of these images are, accordingly, the photographs of crying and grieving; the message that diabetes is a dangerous, indeed fatal, condition that can devastate people's lives. In these health promotional texts, a diabetes diagnosis thus constitutes an undesirable potential future scenario which viewers will undoubtedly want to avoid. In the quest to avoid this frightening eventuality, they might, by dint of this visual grammar, turn to the main body copy information located at the bottom of the imagethe "real", as it were. Here, readers are greeted by each company's logo, a link to the Diabetes UK/Tesco shared website, and explicit instructions to readers to "CHECK YOUR RISK TODAY", and to "GO ONLINE, TO ANY TESCO PHARMACY, OTHER PHARMACY OR VISIT YOUR GP". 8 The internal logic of this ideal-real structure means that readers might assume that the best way to avoid the undesirable future scenarios presented in the photographs is to follow these explicit instructions, the order of which, it is worth noting, privileges those options associated with Tesco: first preference is to go online, second to visit a 'Tesco pharmacy' and finally, if all else fails, visit your GP. Not only does the order of preference of these options reinforce the neoliberal ideal of keeping people out of doctors' surgeries by providing them with the tools for self-maintenance and care (conveniently available online), but it also serves a commercial purpose by allowing Tesco to promote and sell its goods and services online.
Yet, the offer of the Diabetes UK/Tesco brand as the preventative measure against the frightening consequences of diabetes is not always made so subtly in these texts. The clearly visible presence of these logos, positioned at the base of each poster, casts the Diabetes UK/Tesco partnership as the wholly benevolent solution for those who are concerned about, or wish to check, their risk of developing diabetes (Hevey 1992, 34;Lupton 1995, 126-127). Visual branding, at this stage, does not need to do any more than make readers aware of the alliance; there is no need to make claims about what Diabetes UK and Tesco actually do, instead it is the mere presence of these logos, in opposition to the fear-inducing photographs above them, which constructs the Diabetes UK/Tesco partnership as a benevolent "help" to viewers.

Discussion
In this study, we have critically examined the multi-semiotic means through which the (as of 2014) on-going Diabetes UK/Tesco campaign discursively constructs diabetes and people with (or at risk of acquiring) diabetes. We have sought to show how, in drawing on fear-and anxiety-inducing imagery and text, this prominent campaign emphasises the dangerous and potentially fatal consequences of individuals not taking responsibility for personally assessing, and responding to, their (and their family's) risk of diabetes, whether they are actually at risk of contracting the disease or not. Consequently, the campaign presents, we argue, only a partial picture of both diabetes and the actual lived experiences of people who have the condition. This portrayal obscures the reality of diabetes as a disease which, although can indeed have very serious long-term biological consequences, people are able to manage successfully (Ellison and Rayman 1998), living full and long lives (Lutgers et al. 2009). Thus, the campaign mirrors the somewhat sensationalist way in which some quarters of the media report illnesseswith a strong focus on the debilitating and fatal aspects of disease at the expense of reporting how people actually live with and overcome the effects of illnesses (Entwistle and Hancock-Beaulieu 1992;Seale 2002, 45-47).
One objection to our critical multimodal analysis is that if the campaign raises people's awareness of diabetes and prompts those at very real risk of the disease to act accordingly, then surely it is justified in harnessing a rhetoric of fearthe end, in other words, justifies the means. This is a fair rebuttal. However, one can take issue with the way in which such fear-and anxiety-inducing strategies are calculatingly usedthe very deliberate semiotic means through which such scare tactics are realisedand question the extent to which they are genuinely, usefully awareness-raising. Our analysis of the campaign photographs, for instance, revealed how these emotion-laden images failed to make clear to viewers who the represented participants, as well as what the precise contexts of the events in which they are visually depicted, are. As viewers of these pictures we are told very little about these people. Even looking at the text and images together, seeking some clue in complementary meanings, we do not know who the represented participants are, why exactly they are in a state of emotional distress, and, perhaps most vitally, we do not know how they came to develop diabetes. In fact, in order to determine which participant has diabetes, we are, to some extent, forced to rely on our pre-existing knowledge of the condition, and hence risk drawing on inaccurate or, worse still, stigmatising, assumptions about diabetes and who "gets" itwhom diabetes, in the pithy wording of the campaign, "hits".
These deliberately vague and, to some degree ambiguous, photographic representations of diabetes-related suffering might thus actually serve to perpetuate the stigmatising discourses which surround diabetes while simultaneously failing to challenge these in a way that health-related awareness-raising should (Lupton 1995). Another potentially negative consequence of these images is that, in their attempt to convey the sense that diabetes is something that can affect anyone, they obscure the environmental factors which actually underlie the development of the condition. Having inspected the bulletpointed risk factors detailed on the reverse of the leaflet versions of these campaign texts, one is able, perhaps, to attribute certain characteristics to various represented participants (such as ethnicity, age and body weight). However this, once again, forces us to cast moral judgements and further stigmatise certain ethnic and age groups. Gagnon, Jacob, and Holmes (2010, 254) remind us that fear is regaining momentum within the domain of public health promotion, a trend "symptomatic of a broader political context where public health campaigns are inadequately funded and relegated to private enterprises that apply advertising techniques to non-commercial issues". As we have sought to show, the use of various commercial communicative techniquesi.e. logos, slogans, synthetic personalisation and the ideal-real constructionseem to blur the lines between traditional promotional and advertising discourse and (self-proclaimed) noncommercial enterprises, such as public health campaigns and charities (see also Waltz 2013). Nevertheless, it is becoming increasingly common for publicly funded health promotion agencies and charities to form alliances with private enterprises, and to utilise commercial advertising strategies to promote their causes. Yet, the increasing use of commercial discourse raises issues about the ethical considerations surrounding the development and application of public health communication (Guttman and Salmon 2004, 533), and the Diabetes UK/Tesco alliance is a telling case, calling into question the appropriateness of a private partner whose raison d'être is to maximise its profits. Although the alliance professes to raise public awareness of diabetes, the offered solutions through which people assess and respond to their risk of the disease are commercial, with the supermarket supplying access to a range of its goods and services, such as self-regulatory testing and its branded Diabetes Support diet plans.
Moreover, and more urgently, given that supermarkets have actively encouraged consumers' dependence on high-calorie, low-grade processed foods (Blythman 2007(Blythman , 2012), it is morally questionable that Tesco (the UK's largest grocery chain) should advocate diabetes awareness while at the same time promoting an extensive, prominent and relatively cheap range of less nutritious foods (biscuits, sweets, ready meals, sugary drinks and so forth) to consumers, while marking up more healthy produce such as fruit and vegetables (Blythman 2012). The ease of physical access to low-grade, hyper-palatable and nutrient-poor food supports unhealthy lifestyles: the link between diabetes and the over-consumption of such food is well-documented (Butland et al. 2007;van Dam et al. 2002). However, these are important factors to which the Diabetes UK/ Tesco campaign gives relatively little attention in its health promotion materials. In solely emphasising personal responsibility for diabetes risk assessment and management, the campaign downplays the situational factors associated with increasing rates of diabetes and obesity, instead preferring a neoliberal discourse which responsibilises individual viewers in relation to their health management and reducing diabetes risk. Indeed, the very notion of the obesogenic environmentlet alone how individuals can actually, practically tackle it and thereby help reduce the prevalence of diabetesis, quite remarkably, though conveniently, side-stepped altogether.

Concluding remarks
We have argued that the health promotion campaign examined in this study is ethically problematic in terms of its use of visual and linguistic "scare tactics", particularly when these scare tactics also appear to be commercially motivated. The three themes we have discussed throughout this paperthe depiction of grief and amplification of diabetesrelated danger, the promotion of diabetes risk and localisation of individuals' responsibility for their health, and the commercial framing of the Diabetes UK/Tesco partnershipoperate, we argue, in a unitary fashion, geared towards promoting access to specific, tailored goods and services offered by these organisations. Specifically, by first generating fear about diabetes as a devastating condition, and then instructing the general public that they are eminently at risk of the disease, and responsible for evaluating their own and families' risk, these images "set the stage", as it were, for offering the Diabetes UK/Tesco partnership as the solution to which the public should turn to check, and manage, their diabetes risk and indeed health more generally.
Health promotion texts are essentially multimodal, harnessing in their designs not only language but also visual elements, thereby making meaning over more than one level of semiosis. The multimodal, critical approach to discourse analysis that we have adopted in this study is well-suited to this task, enabling us to interpret the seemingly banal and taken-for-granted visual and linguistic aspects of these campaign images in terms of their deeper or more "hidden" meaning which reveal the neoliberal and commercial underpinnings of the design choices. Accordingly, we hope to have demonstrated how a multimodal approach to analysing health promotion discourse can illustrate the subtle workingsthe tensions, ambiguities and contradictionspresent in a contemporary health campaign, workings that a purely mono-modal analysis might fail to apprehend or interrogate so explicitly. However, as with much social semiotic research, the interpretation of these campaign images that we have presented is our own, and cannot claim to be representative of how other people will necessarily interact with and understand them. The ways in which audiences make meaning are highly subjective, deeply contextualised and impossible to predict with any absolute certainty. Nevertheless for those researchers who wish to appreciate how discourses are expressed, rendered natural and legitimised (Machin 2007, 1) in health promotion texts, it behoves them to embrace (or at least recognise) a critical multimodal perspective.
2. Practitioners distinguish two types of diabetes: Type 1 and Type 2. Type 1 diabetes usually develops before the age of 40, and occurs when the body fails to produce insulin, which results in unhealthy increase in glucose levels. Type 2 diabetesthe more common strainusually develops later in life, and comes about because the body either does not produce enough insulin or does not utilise the insulin that is produced (insulin resistance). Type 2 diabetes is often associated with obesity and a poor body mass index (NHS Choices webpage 2014). 3. We are aware that our initial description of the campaign images here already constitutes, as Barthes (1977, 33) observes, a metacommentary. Moreover, due to copyright restrictions, all the images reproduced in this article appear as illustrations. Accordingly, we refer readers to the reproductions of the originals which can be viewed directly at: www.pinterest.com/nhsinmanchester/talking-diabetes/ 4. Hyper-palatable foods are those foods which require little chewing, are easy to swallow and typically low in nutritional value (http://www.nytimes.com/2009/06/23/health/23well.html?_r=0) 5. Interestingly, Wallis and Nerlich (2005, 11) observe that the metaphor "hit" is the most common term used to describe the action and effect of severe acute respiratory syndrome (SARS) in a corpus of media coverage of the disease. The authors note the way in which the infection is accorded an active role (SARS a killer), commonly constructed as a single, unified, lethal entity, a disease very much a "free agent responsible for its actions", while those it affects are typically depicted as "passive, blameless victims" (Wallis and Nerlich 2005, 13). 6. This warning symbol is widely understood in Western cultures to communicate a warning to the viewer that something in their immediate environment is dangerous to them (Finan 2002, 17-18), in this case, evidently, their risk of developing Type 2 diabetes. 7. Synthetic personalisation is that communication (written or otherwise) which, although addressing a mass audience, is designed to make individual members of an audience feel as though they are being contacted on a personal and private level (Fairclough 1989), achieved, primarily, through the extensive use of first-person address and conversational linguistic style (Scannell and Cardiff 1991;Bell and van Leeuwen 1994;Garrett and Bell 1998). In reality, this implied personal, private channel of communication does not usually exist, least of all in public health awareness-raising campaigns and advertisements which are directed at mass audiences unknown, of course, to the text creator(s) on a personal level. 8. When going online to complete the risk assessment, as is advised in these posters, one is presented with yet more neoliberal rhetoric about diabetes-related health management. The questions in the risk assessment are concerned with mainly fixed biological factors, such as sex, age, ethnicity, height and, to some degree, weight. Yet, there are no questions about the availability, accessibility (and affordability!) of more wholesome food such as fresh fruit and vegetables where people live. These assessment criteria, by which risk is measured, serve to further deflect attention away from the environmental factors which contribute to the increasing prevalence of diabetes, firmly placing responsibility on the individual respecting risk management. Incidentally, even if one achieves a "low-risk" score, the user is still emphatically reminded, amid bursts of glib approbatory rhetoric, to keep up his/her standards, and that risk reduction is an on-going personal project: "You're doing a great job staying healthy! Remember to keep an eye on the factors that contribute to Type 2 diabetes, such as your weight and waist size". It seems that, no matter who you are, no matter how low your "risk score" is, diabetes is still constructed as an ever-present threat to our health and hence, as neoliberal subjects, we are exhorted to be healthier. Yet the message is a contradictory one: the variables about which we are questioned in this risk assessment are largely out of our control, so while the responsibility of achieving and maintaining good health is formulated as an individual endeavour, the goals to which we are instructed to personally strive are ultimately unattainable.

Notes on contributors
Gavin Brookes is a doctoral researcher in health communication at the School of English, University of Nottingham. He is currently investigating the discursive constructions of diabetes and diabulimia in online health fora.
Kevin Harvey is a lecturer in sociolinguistics at the School of English, University of Nottingham. His specific research interest is in the area of discourse-based health communication, which includes corpus linguistic and multimodal approaches to health-related discourse.