Seeing climate change: the visual construction of global ...blogs.ubc.ca/difrancesco/files/2011/04/Seeing... · construction of global warming in Canadian national print media Darryn

Corresponding author:Nathan YoungEmail: [email protected]

cultural geographies18(4) 517–536

© The Author(s) 2010Reprints and permission: sagepub.

co.uk/journalsPermissions.navDOI: 10.1177/1474474010382072

cgj.sagepub.com

Seeing climate change: the visual construction of global warming in Canadian national print media

Darryn Anne DiFrancesco and Nathan YoungDepartment of Sociology and Anthropology, University of Ottawa, Canada

Abstract

Visual communication is a critical but frequently under-estimated contributor to the ‘social and cultural life’ of environmental issues. This paper uses both content and discourse analysis to examine how visual communication is deployed in print media coverage of climate change issues in Canada. The Canadian case is internationally significant, given that Canada ratified the Kyoto Protocol but has since become obstructionist on the global stage. Our analysis, which focuses on image-language interactions, leads us to conclude that climate change is being inconsistently narrated to Canadians in this regard. While the power of visual communication comes from its ability to blend fact and emotion, to engage audiences, and to add narrative complexity to linguistic claims (and vice versa), we find instead a profound disjuncture between images and text in climate change coverage. In this case, visual and linguistic communication tend to pull in different narrative directions, advancing unrelated and sometimes contradictory claims that frequently confuse different aspects and positions on climate change.

KeywordsCanada, climate change, environment, imagery, media, visual communication

‘It is pictures rather than propositions, metaphors rather than statements, which determine most of our philosophical convictions’ – Richard Rorty1

Introduction

Judging by the amount of political, public, and media attention it has received over the past decade and more, it is now fair to say that global climate change is one of the definitive environmental issues of our time. Despite its high profile, however, climate change remains a technical and abstract problem, far removed from the direct experience of most people.2 Some observers have gone so far as to describe it as an ‘elitist’ issue, given the complexities of climate science and that political debate and conversation about climate change are dominated by highly-motivated actors and powerful interests.3 At the same time, however, it is increasingly clear that climate policy will go nowhere without public support and involvement. Yet bringing climate change ‘to the masses’

518 cultural geographies 18(4)

has proven challenging: while awareness of the issue among the general public is high, the urgency of the problem is not always recognized, particularly in North America.4

The powerful effect of Al Gore’s 2006 activist documentary An Inconvenient Truth suggests that visual imagery is essential for making climate change ‘consumable’ for a sizable fraction of the popu-lation. While images are often said to embody complexity (being worth the proverbial thousand words), media theory tells us that they also reduce complexity by providing interpretive frames or nar-ratives that selectively blend fact and emotion.5 The power of imagery to reduce complexity and gal-vanize public opinion has been well-documented with environmental issues such as ozone depletion, deforestation, drought, and endangered species.6 Unlike many other environmental problems, how-ever, global climate change is an issue that has ‘no ready-made metaphors’.7 As Ungar points out, linguistic metaphors such as the ‘greenhouse effect’ have proven too benign to resonate in the public mind,8 which suggests that visual imagery potentially has a strong role to play in public understandings of climate change. Despite this, few studies have been completed to date that empirically investigate the kinds of images that are being attached to the climate change issue,9 and even fewer about how images interact with text and language to present discursive or narrative packages to the general public.

In this paper, we analyse the types of imagery that are being used in print media communica-tions about climate change in two major national newspapers in Canada, and examine how these images fit into broader (textual) claims-making about the issue. The Canadian case is important for several reasons. First, Canada has demonstrated deep political ambivalence to the climate change issue. While the country has ratified the Kyoto Protocol, Canadian greenhouse gas (GHG) emis-sions continue to rise as political foot-dragging on climate policy continues. Internationally, the Canadian government has recently behaved in a recalcitrant and even obstructionist manner at several global events, including the critical Copenhagen Conference held in December 2009. Public opinion surveys since the mid 1990s have shown that a majority of Canadians desire strong action on climate change, although the most recent polling (following the ‘Climate-gate’ and ‘Glacier-gate’ scandals) shows that half now prefer a ‘go-slow approach’ to the issue.10 Second, Canada is a country rich with potential climate change images, from its vast Arctic spaces to its carbon-intensive resource sectors (including controversial oil sands development) to its lengthy marine and freshwater coastlines. In short, this is a place where we would expect to see rich and varied climate change-themed imagery. Third, cultural geographers and other social scientists have long noted the central role of wilderness, landscape, and climate in Canadian identity and sense of place.11 While the Canadian population is highly urbanized and largely grouped along the Canada-US border, cold weather is a major cultural trope in the country, along with winter icons such as icebergs, glaciers, outdoor hockey rinks, and charismatic fauna such as caribou and polar bears.12 Indeed, Canadian environmental groups frequently frame climate change as a threat not only to the Canadian economy and people but to its cultural heritage as well.13

Generally speaking, newspapers are neither the only nor the primary source of visual imagery for the general public. Nevertheless, research has consistently shown that newspaper reporting (in print or online) is a critical source of in-depth information for the public on climate change issues.14 Newspaper articles are also sites of intense interactions between imagery and language, wherein images are used to both communicate ideas directly to readers and to draw their attention to and contextualize accompanying text. This article therefore seeks to answer the following ques-tions: How is climate change being visualized in the Canadian context? What kinds of images and metaphors (if any) dominate discourses surrounding climate change? Is there a difference in the way in which climate change is talked about compared to how it is seen?

To answer these questions, we draw on a broadly constructionist approach to the climate change issue.15 While climate change is certainly a real-world phenomenon that can and must be investi-gated and addressed by the natural sciences, it is also clear that the climate change issue has a

DiFrancesco and Young 519

‘social and cultural life’ that is influenced by competing claims, values, priorities, and narratives. Drawing on insights from environmental sociology, communication theory, and media theory, we seek to understand how the climate change issue is packaged using combinations of visuals and lan-guage to be consumable by a broader public not necessarily well-acquainted with climate science.

Our investigation is based on analysis of newspaper items about climate change published over a six month period (January to July, 2008) in Canada’s two national newspapers, The Globe and Mail and The National Post. These newspapers are both based in Toronto, but are distributed across the country and focus predominantly on national and international issues.16 They are Canada’s second- and third-largest English language dailies by circulation (behind the Toronto Star) with average distributions of 337,000 and 206,000 respectively. Generally speaking, The National Post is considered to be a right-of-centre publication, while The Globe and Mail is more centrist. Both papers, however, are strongly oriented to Canada’s business class. Analysis of newspaper items proceeded in three phases: (1) content analysis of all images appearing in or alongside articles concerning climate change; (2) content analysis of the text of all articles concerning climate change (including articles without accompanying images) and (3) intertextual discourse analysis of a selected number of image-language combinations in order to more deeply investigate themes uncovered in the first two stages of the research.

The role of images in the social construction of environmental issues The study of social problems from a constructionist perspective puts special emphasis on claims-making, particularly the process of generating ideas and narratives and their advancement or pro-motion into the public sphere.17 Claims-making is often politically strategic, as is frequently seen in communications by interested parties on all sides of the climate change issue, but it can also be subtle and even unintentional – emerging less from political strategy than from honest attempts to report the news and inform the public. According to Young and Matthews, media reporting on scientific and environmental issues is guided by three goals or ideals: investigation (uncover-ing scientific or environmental issues, problems, or arguments hitherto unknown to the public), communication (translating expert language and claims into publically-consumable form), and narration (explaining ‘why one should care’ about a given problem or issue).18 While columns and editorials are widely recognized as ideological or ‘interested’ content, ordinary science and environmental reporting and news items should also be seen as important exercises in claims-making, as they involve the simplification and selective representation of complex issues. According to some observers, even ordinary coverage of science has become more political in recent years as reporters are now less deferential to experts, more willing to seek out and publish alternative claims, and make frequent use of controversy as a narrative to sus-tain interest in a topic.19

While not always recognized as such, imagery is an important part of textual claims-making. According to Hannigan, images are critical in the social construction of environmental problems because they create an avenue for environmental issues to be ‘dramatized in highly symbolic and visual terms ... [which] provide a kind of cognitive short cut compressing a complex argument into one that is easily comprehensible and ethically stimulating’.20 Similarly, Smith and Joffe argue that ‘the visual [has a] particular ability to arouse emotion, making it an effective medium for the social construction of risk messages’.21 Images also serve to ‘bring an issue home’ for audiences, not only spurring an emotional response but also engendering a sense of engagement and commitment to an associated problem or event.22 As such, they have an important role to play in the journalistic goal of narration or explaining to audiences why a particular issue or problem is significant and worthy of attention.


Images and visual communication more generally are not only exclamatory (drawing attention to an issue) but also interpretive. The power of imagery to persuade has long been recognized by advertisers and marketers.23 According to Ungar, much of the persuasive power of imagery comes from the ease by which visuals communicate metaphor and analogy.24 Existing research suggests that visual metaphors and representations connect and communicate more profoundly with audi-ences than do linguistic metaphors.25 Joffe explains this resonance by arguing that ‘visuals are readily absorbed in an unmediated manner because viewers are not generally provoked to reflect on or deconstruct them in the way that occurs in relation to verbal [and print] material’.26 As Marshall McLuhan observed forty years ago, images have the power to ‘speak for themselves’ while simultaneously inviting viewers to imaginatively fill in the blanks of what is off camera (i.e., the context of the image) in a way that printed text, with its linguistic precision and sequential presentation, cannot.27 Therefore, we argue that visuals and text should be considered together as ‘co-constructors’ of environmental narratives that, in combination, convey complex and multi-dimensional messages to media consumers.

News media coverage of climate changeWhile not always labelled so, the North American news media have been covering climate change issues since at least the 1930s.28 Given that early scientific studies of climate change aimed to explain the recurrence of historical ice ages, it is not surprising that these early media accounts focused on the potentially devastating effects of future global cooling.29 As is well documented elsewhere, concern about global warming did not find regular mainstream media coverage until the 1980s after the emergence of the ozone depletion problem as well as the unusually warm summer of 1988 that brought about droughts, floods, and heat waves around the world.30 Media interest in climate change has ebbed and flowed since through several attention cycles, during which the content of coverage has changed substantially.31 Trumbo found that over time, scientific voices have become less prominent in media discussions of climate change, while politicians and political considerations are increasingly dominant.32 As Demeritt points out, climate change migrated into the policy sphere with astonishing quickness given the complexity of the problem and (at the time) high levels of scientific uncertainty.33 The Canadian Government, for example, drafted an initial climate change policy in 1989 called ‘The Green Plan’ that focused on efficiency standards and consumer education but shied away from controversial topics such as industrial emissions from Canada’s politically powerful resource industries.34 The climate change issue, while always scien-tifically complex, also became politically complex early on, and this has clearly driven additional media interest in the topic.

As argued by Boykoff, the news media have inherent difficulty covering slow-developing or ‘creeping’ scientific or environmental issues, which means that news reporting of climate change has focused strongly on events in nature (extreme weather and natural disasters), science (publica-tions, reports, and other discoveries) and politics (summits, meetings, and policy announcements).35 In between such events, however, the climate change issue is maintained through a heavy reliance on narrative.36 Current research indicates that narrative is created in several ways, including a focus on pseudo-events such as celebrity involvement in activism.37 For the most part, however, media items on climate change use conflict as their primary narrative vehicle, although this occurs differ-ently in different regions. Studies of US media coverage invariably point to the ‘creation’ of con-troversy on the science of climate change itself, particularly as encouraged by (or excused by) the journalistic ‘norm of balance’ of giving equal time and space to competing opinions on a topic.38 While Smith and Joffe argue that ‘this trend is no longer observable’ in most media and that ‘more


recent coverage focuses almost exclusively on human causes’,39 our Canadian data refute that finding (see below). Studies of European media, in contrast, have found a predominant narrative of diplo-matic or intergovernmental conflict over the climate change issue, particularly between the US and EU.40 Available evidence suggests that the Canadian media thematically fall somewhere in between European and American tendencies – variously emphasizing scientific controversy and diplomatic/political conflict over appropriate responses to climate change.41

As mentioned, most studies of climate change reporting have focused exclusively on language, making it difficult to anticipate how images affect and colour coverage of events and interregnum periods of narrative-making. Smith and Joffe have conducted the most thorough investigation to date of climate change images in newspaper reporting in a study focused on the UK. Their research found that most images focused on consequences, with ‘over 50% of images illustrating one or more climate change impacts’, the most frequent being melting ice or glaciers, followed by flood-ing and polar bears.42 According to the authors, these images, along with photographs of people affected by climate events (in the UK), had the effect of ‘bringing the issue closer to home’.43 More than this, the predominance of impacts images leads Smith and Joffe to conclude that, overall, imagery is making a one-sided argument. While the text of the articles may contain ‘discussions of controversy, unknowns, and debate … the concretization of impacts in newspaper images speaks against the argument that climate change science is uncertain’.44 Their conclusions, however, should be qualified by research into audience reception of climate change imagery conducted by O’Neill and Nicholson-Cole.45 Using controlled-experiment and focus-group methods, these auth-ors found that fearful or hard-hitting images of climate change impacts do elicit strong emotional reactions, but are also ‘likely to distance or disengage individuals from climate change, tending to render them feeling helpless or overwhelmed when they try to comprehend their own relationship with the issue’.46 This leads them to conclude that images of impacts may in fact be counter-productive compared to more ‘positive’ representations of green technologies.

In summary, existing theoretical and empirical work suggests that images are critical means of simplifying and communicating complex messages about the environment. While it is unclear whether the lack of a central or defining image to represent global climate change has hindered public commitment to the issue,47 this lack of a default representation means that the visual medium is flexible in this case – many different images can be mobilized to add emotive and interpretive dimensions to (linguistic) climate change narratives. In the remainder of this paper, we present findings from a study of media coverage of climate change in two major national newspapers in Canada to analyse how text and imagery combine in communications about climate change issues.

The studyAs mentioned previously, this article uses methods from both content analysis and critical discourse analysis to investigate image-language interactions in climate change coverage in Canada’s two predominant national publications, The Globe and Mail and The National Post. Content analysis represents a more traditional approach to the study of communications that emphasizes denotative rather than connotative or latent content. In this case, it is used to present a broad overview of the-matic trends based on standardized content codes that are applied uniformly to each image and article – a method that is common in media studies and sociology.48 In contrast, our use of critical discourse analysis on selected image-language combinations allows us to more thoroughly examine how these interact (or fail to) in the construction of complex narratives. This approach, which is more common in the humanities and culturally-minded social sciences, looks beyond surface con-tent and seeks to interpret communications within broader social, political, and cultural contexts.49


Media items for analysis were gathered using the Factiva search engine’s ‘full text’ search option employing the keywords ‘climate change’ OR ‘global warming’ OR ‘greenhouse effect’, AND ‘images’, OR ‘photos’ OR ‘illustrations’. This yielded a total N of 375 articles for the six month period January 2008 to July 2008 (253 articles in The Globe and Mail and 122 in The National Post).50 The images in these articles were then subjected to a preliminary content analysis on their ‘surface’ themes, i.e., the subject(s) of the image, its thematic context (if discernable) and, if applicable, its geographic location.51

The linguistic components of the articles were also thematically coded according to a predeter-mined set of content codes. In so doing, it became clear that many articles containing images in fact mentioned climate change issues only in passing. After excluding these ‘passing reference only articles’, we were left with an N of 100 articles that substantively consider one or more issues related to climate change.52 These articles are the object of our analysis of image-language inter-actions presented below. Finally, we created a comparison group of articles without images pub-lished in the two newspapers during the same time period (this group was similarly screened). Content analysis of these non-image articles allows us to compare how climate change is discussed in the presence and absence of visuals.

Content analysis of imagesAs mentioned, one of the most basic ways to analyse an image is to identify the subject and its ‘surface’ or denotative content.53 Findings from this most preliminary level of analysis are pre-sented in Table 1.

Table 1 shows that the climate change story in Canada is being primarily told using human imagery (note that the categories are not mutually exclusive). While visuals of the natural environ-

Table 1. Content analysis of images

N %

Image subject Theme Globe & Mail National Post Total

Human (all themes) 169 (67%) 81(66%) 250 (66%) Political 58 (23%) 22 (18%) 80 (21%) Citizen 46 (18%) 13 (11%) 59 (16%)

Business/Industry 27 (11%) 21 (17%) 48 (13%) Scientist/Expert 8 (3%) 6 (5%) 15 (4%)

Environmentalist 8 (3%) 6 (5%) 14 (4%)Celebrity 6 (2%) 7 (6%) 12 (3%)

Nature (all themes) 115 (45%) 43 (25%) 158 (42%)Urban landscape 21 (8%) 15 (12%) 36 (10%)Natural landscape 22 (8%) 1 (1%) 24 (6%)Ocean/coast 8 (3%) 0 8 (2%)Polar bear 7 (3%) 3 (2%) 10 (3%)Other plant/animal 22 (9%) 3 (2%) 25 (7%)Snow/ice 12 (5%) 5 (4%) 17 (5%)

Industry/Technology (all themes) 66 (26%) 28 (23%) 94 (29%) Transportation 24 (10%) 8 (7%) 32 (9%)

Oil sands/refinery 19 (8%) 13 (11%) 32 (9%)Green technology 18 (7%) 2 (2%) 20 (5%)


ment are frequently deployed (42% of all images include an animal, plant, or landscape), two-thirds of all images focus identifiably on one or more human beings. The least frequent subject is that of industry or technology, although these still appear in roughly one-third of all images.

These findings are both consistent with and divergent from existing research. Smith and Joffe’s study of images in UK print media also found that climate change was being visually framed in predominantly human rather than natural terms.54 But whereas they argue that the focus on people ‘puts a human face’ on climate change and thus helps to ‘personify the issue’ by presenting images of an affected public,55 we find that most images of people in Canadian coverage are emotionally benign. As demonstrated in Table 1, the modal category for images of people involves depiction of a politician, and the third-ranked category is of people in business or industrial roles. These images are typically posed portraits or ‘action’ images from media scrums, parliamentary performance, or press conferences. A significant minority of images do depict ordinary citizens (16%), and many of these are presented as representative of a concerned or ‘affected public’ in the manner described by Smith and Joffe. However, Table 1 shows that the Canadian national media are notably averse to images of scientists, environmentalists, and celebrities. While the US and UK media have embraced these themes – to the point that Boykoff and Goodman argue that the climate change narrative is increasingly dominated by a ‘science-policy-celebrity complex’ that is used to connect climate change to people’s everyday experiences56 – this is not a major theme in Canada.

The images of nature that appear in the Canadian media are notable in their aversion to expected tropes. For instance, images of polar bears, which are frequently employed in activist campaigns by environmental groups in Canada, are largely absent from the print media (appearing in only 3% of cases). Slocum argues that images of polar bears are key ways of ‘localizing’ climate change issues for Canadians,57 and their absence suggests that this is not a media priority. The same may be said for images of snow and ice (5%) and oceans or coastlines (2%), which are important Canadian symbols that are expected to be directly affected by changing climate. Instead, the modal theme in this category is in fact presentation of an urban landscape, including city skylines, specific buildings (such as federal and provincial houses of parliament), and streetscapes. This tendency is not readily explainable, although it may reflect the general bias of national media towards urban issues. Interestingly, natural landscapes such as forests, mountains, oceans, glaciers, and rivers/lakes appear in only 6% of cases, and are virtually absent from The National Post (which may be due to the paper’s aforementioned conservative orientation).

Images of industry and technology are predominantly focused on transportation and representa-tions of Canada’s oil sands or ‘tar sands’ industry. The emphasis on transportation is notable (Canada is a large country) and the majority of these images depict gasoline stations and highways (remembering that the study period coincided with record-breaking global oil prices in the first half of 2008). The portrayal of the oil sands may also have been partially motivated by fuel prices, but, as we will discuss below, the oil sands have become a major symbol in critical claims-making about climate change in the Canadian context. We also note that the number of images of green technologies – which were identified by O’Neill and Nicholson-Cole as important ‘positive’ icons of climate change – is surprisingly low. Depictions of solar panels, wind turbines, and energy effi-cient light bulbs, all of which are prominent in activism, are rare in Canadian national media and almost completely absent from The National Post.58

Finally, when it was possible to identify geographic context of the images, the vast majority (75%) were of Canadian subjects (human, natural, or industrial). Another seven percent came from the United States or Europe, and eight percent were of unidentified Arctic or Antarctic regions. Only six percent of identifiable locations were ‘exotic’ (non-North American, European, or polar).


In sum, this preliminary level of analysis finds no evidence of a unifying or central image or visual thematic for representing climate change issues. Expected tropes, particularly from the nat-ural world, appear infrequently (although we will later argue that they are discursively powerful when they do appear). Overall, our findings are consistent with Trumbo’s observation that as the climate change issue has matured, it has become less focused.59 In Canadian coverage at least, the visual representation of impacts is relatively rare (contra Smith and Joffe’s UK findings) compared to depictions of everyday political and business affairs that are only marginally connected to cli-mate change problems and themes.

Image-language comparisonsImages alone, however, fail to tell the whole story. As we noted earlier, very little research has been done to date on how images and language interact to present thematic packages to readers. In this section, we examine the ways that images and language do and do not relate to one another. To do so, we subjected the accompanying print text of the articles to traditional thematic content analysis based on a pre-determined coding scheme intended to capture specific claims, such as assertions about causation, blame, and the risks and benefits of climate change, as well as key themes and metaphors such as generational rights, green or eco-capitalism, economic growth, and justice.60 As mentioned previously, this content analysis was performed on a subset of the original sample (N= 100) including only those articles whose accompanying text is predominantly concerned with climate change issues. We also introduce a comparison group of similarly-screened articles appearing without images in the two newspapers during this time period (N=391) to see whether articles containing visuals demonstrate particular tendencies.

Comparing articles with images to articles without imagesTo begin, the image-language comparison demonstrates that images are most likely to appear alongside editorials (59% of all editorials were accompanied by an image), followed by in-depth features (42%), and least frequently with news items (23%) and letters to the editor (4%). Also, images tend to accompany longer articles, averaging 959 words, compared to the non-image arti-cles, which averaged 524 words. Given the strength of these associations, all of the findings described below control for type of article and word count.

Table 2 provides a summary of the findings from our comparison of articles appearing with and without images. These comparisons were achieved using logistic regression, whereby each vari-able in Table 2 was entered as the dependent variable along with the independent variables of word count, editorial (as a 0–1 or ‘dummy variable’), news item (as a dummy variable), and presence or absence of an image. This method allowed us to control for word count and type of article and thus isolate the independent effect of the presence or absence of the image (using the standardized coefficients (Beta scores) generated by the regression models).61

The first column in Table 2 shows that articles that appear with images are significantly more likely to include, among other things, attributions of blame for climate change, identification of a victim (human or non-human), use of a ‘crisis’ metaphor, mention of rights (human or non-human), and mention of the future. This combination of themes suggests that images tend to appear along-side articles that are moral or emotional in tone. This conclusion is supported by our parallel find-ing that attributions of causation (be it anthropogenic or non-anthropogenic) are less likely to appear in articles with images (see column 2). In other words, articles with images are more likely


to attribute blame and discuss victimization and rights, while at the same time avoiding more tech-nical discussions of causation. This is a notable reversal of Smith and Joffe’s conclusion that ‘imagery plays an important role in fostering emotional reactions to climate change information’.62 By looking at visuals and text together, we find that most images are in fact unemotional (see pre-vious section), but are accompanying more emotionally- or morally-oriented print text.

As mentioned earlier, images are also presumed to be important means of ‘bringing an issue home’ or making the link between abstract problems and everyday symbols and experiences.63 Here, we find mixed evidence. Table 2 shows that articles with images are more likely to use Canadian sources or voices, and that they are more likely to mention the carbon-intensive oil sands, which is a uniquely Canadian issue. On the other hand, no difference was found between the groups with respect to international orientation, measured by mentions of other countries in the article (an original effect is present, but it is negated by word count as longer articles tend to mention other countries regardless of the presence of images). However, given that the vast majority of visuals are identifiable as Canadian (see above), we conclude that articles with images do indeed serve to ‘localize’ the issue.

Finally, we note that the presence or absence of an image does not add to the narrative complex-ity of an article. To measure this, we constructed a ‘narrative complexity index’ that summed all possible content codes relating to metaphors, themes, claims (for and against the reality of anthropo-genic climate change) and issues (such as ecological issues, economic issues, health issues, and so on). The absence of any effect on this composite variable suggests that images are not driving the article. That is, the presence of an image does not indicate a more complicated or nuanced text than found otherwise. Stated differently, articles without images are just as thematically sophisticated (or limited) as articles that are accompanied by visuals. This, along with our finding that unemotional images frequently accompany emotionally- or morally-themed text, suggests a disjuncture between image and article. Rather than being complementary, many image-language combinations appear to be dissociated and pulling the narrative in different directions (we revisit this point in the dis-course analysis section below).

Table 2. Differences in linguistic content between articles with images and articles without images appearing during the test period

Images are more likely to appear with …

Images are less likely to appear with …

Variables that show no effect (no differences)

Attributions of blame* Discussion or attribution of causation**

Domestic vs. international orientation

Identification of a victim of climate change**

Mention of ecological issue

Use of “crisis” metaphor* Mention of economic issueMention of the future** Mention of Aboriginal issueMention of rights (human or non-human)*

Narrative complexity index

Use of Canadian sources/voices*Mention of the oil sands**

* Statistically significant to <.05** Statistically significant to <.01


Comparing articles with different visual themes

Table 3 shows that different categories of images are only modestly associated with different lin-guistic themes.64 The variables in the top half of Table 3 list variables that do not vary across visual themes. For instance, images of people, nature, and industry or technology are all associated with longer word counts and certain types of articles (editorials). Similarly, all categories of image are associated with the moral-emotional themes of blame, victimhood, discussion of rights, and men-tions of the future. These variables are associated with the presence of an image, rather than the content of that image, which again reinforces our conclusion about the disjuncture that frequently exists between image and text.

In contrast, the variables listed in the bottom half of Table 3 do differ according to visual theme.65 For instance, expert knowledge claims are more strongly related to images of nature than of people or industry, which means that nature visuals are frequently attached to articles that discuss scientific issues (as well as to articles dealing with Aboriginal issues). Several other notable tendencies are revealed. Mentions of the oil sands coincide with visuals of people and industry, but not nature, which may reflect a preoccupation with the oil sands as a political and economic issue first, and an ecological one second. Finally, we note that images of people tend to accompany the most thematic-ally bland articles, while images of nature and industry touch on a wider range of themes (although these latter categories still do not differ from non-image articles in overall narrative complexity).

In-depth discourse analysisNext, we turn to a more in-depth critical discursive analysis of selected image-language inter-actions. As discussed earlier, the particular strength of imagery is its ability to dramatize issues, generate emotional responses, and ‘create cognitive shortcuts that compress complex arguments’.66 Thus far, our findings suggest that this is not really being done in Canadian print media coverage of climate change issues. As discussed earlier, however, the total impact of visual communication is difficult to assess using traditional content coding alone. Images typically have a high degree of latent content that requires a measure of ‘decoding’ by viewers, who must call upon tacit knowledge

Table 3. Comparisons across type of image

Variable Is the variable associated with … ?

Image of person Image of nature Image of industry or technology

Word count Yes Yes YesType of article Yes Yes YesNarrative complexity No No NoAttribution of blame Yes Yes YesIdentification of victim Yes Yes YesMention of rights Yes Yes Yes

Mention of the future Yes Yes YesExpert knowledge claim No Yes NoMention of ecological issue No Yes YesMention of oil sands Yes No YesMention of Aboriginal issues No Yes NoUse of “crisis” metaphor No No Yes


to adequately interpret them.67 Latent content can include juxtapositioning within the image, facial expressions, lighting and colour, angle of presentation, and the spatial centrality or peripherality of key subjects. Interpretation of latent content is personal but also culturally-informed (thus opening it up to discursive analysis) and is often explicitly guided by prompts such as titles, captions, and headings.68 In the following, we subject four images that represent the range of our findings to inter-textual discourse analysis, which looks at ‘how genres, messages, and narratives may be mixed’ across different media (image and language) to present interpretations of a particular issue or event.69

The PoliticianPerhaps most representative of the findings from the preliminary content analysis is an image attached to an article from The Globe and Mail that depicts Canada’s environment minister, John Baird, inside the House of Parliament. In the photograph, Baird stands facing a crowd of reporters, who are thrusting their microphones toward him expectantly. One journalist is asking a question, while the others uniformly wear sceptical expressions. A key government document on climate change policy appears in the foreground. Baird himself faces the scrum with a solemn expression, appearing as if he is looking past them.

Figure 1.Reprinted with permission of The Canadian Press.


While most of the images in the original sample were less loaded with latent meaning (for example, a snapshot of a politician, cropped around his or her face), this image actively constructs a narrative. First, a viewer of the image will immediately perceive contrast. Most of the image is dark, against which the politician’s and journalists’ faces are starkly white and prominent. Baird’s facial expression, body language, and body positioning (opposite the media) suggest confrontation. Overall, the image communicates a narrative of a politician being held accountable. The opposition of Baird and the media represents a visual trope of authority (the minister) versus the critical and expectant public (the media). According to the interpretive methods suggested by Loseke, the cen-tral character in this montage (Baird) is cast as somewhat of a ‘villain’, who is stand-offish, stone-walling, or hiding something, while the reporters are cast as investigators or interrogators diligently exposing him to the light.70

The linguistic prompts support this narrative. The title of the article reads: ‘National carbon market faces opposition’ while the subtitle explains: ‘As Ottawa [i.e., the federal government] moves to establish an emissions trading system, some provinces are already going their own way’.71 Interestingly, however, the main text of the article is not thematically consistent with the image or the caption. While the image presents a conflict narrative, the main text focuses on the potential to earn revenues by means of climate change mitigation, specifically the establishment of a national market for carbon emissions trading. Thus, this is a clear case of narrative disjuncture, where image and language are telling different stories about climate change.

The RefineryThis example from The National Post (‘Greenhouse ‘intensity’ to rise’) features a large image of a refinery in the province of Alberta’s oil sands region. The image depicts the refinery from afar, viewing industrial activity in the centre of a large and open landscape. The focal point of the image is a stream of white vapour exuding from the chimneys (or smokestacks, as they are commonly called, although in many instances, such as this example, the emission is in fact steam): three opaque clouds drifting upwards and across to cover the upper portions of the image. Depictions of oil refineries or their smokestacks were common within the Industry/Technology subject categor-ies, and were often used to represent industrial activities of all kinds. This image is particularly salient in its latent or connotative aspect – the distance from which the refinery is pictured presents the emissions as an assault to the landscape.

The text of the article describes the depicted oil refinery’s projected greenhouse gas emissions over the next several years, noting that ‘emission ‘intensity’ will jump dramatically over the next five years, eventually nearing levels recorded 18 years ago’ as a result of new refining practices currently being implemented.72 True to the journalistic norm of balance, the author references opposing positions on the issue at hand. First, the company’s president is quoted; he defends it on the basis that ‘yes, the greenhouse intensity is pushed up, but the other environmental impacts are actually minimized … it is a bit of dilemma’.73 Directly following this quotation is a statement by an environmental institute’s program director, who states that the noted increase in emissions is ‘alarming’ and that he is ‘disappointed’.74

In this case, the text of the article works to critically frame the issue as complicated and contro-versial, quoting both sides of the story, but finishing on a note of ‘disappointment’ on the part of the environmental institute. This reading affirms the negative connotative element of the image.

However, the image caption presents a frame almost converse to this, reading: ‘Oil sands refin-ing makes a smaller “surface impact” than mining operations, Suncor says’.75 This caption presents oil refining as an environmentally positive practice, and in effect positions the practice outside of


the controversy surrounding increased emissions altogether. Also notable is that directly below the image (and caption) are three different graphs, each indicating the company’s increase in green-house gas emissions. The graphs correspond more closely to the article content than to the image caption. The reader is thus left with the task of negotiating these distinct frames in order to con-struct an understanding of the issue.

The Polar BearsThe image accompanying The Globe and Mail article ‘Manitoba recognizes threat to polar bears’76 is classically emotive. The image depicts two polar bears, a mother and her young, nestled together. The cub has its face turned into his mother, as he licks and nuzzles her chin. Both have their eyes closed in what appears to be contentment or serenity.

The aesthetic and personified nature of the image invites the viewer to feel rather than internal-ize information.77 As mentioned earlier, these types of emotive images were in fact quite rare in our sample. This goes against the findings of Smith and Joffe in their study of UK media,78 and is also surprising given that ‘claims encouraging audience members to feel in particular ways are less likely to be challenged’.79 The rarity of this method of claims-making leads us to conclude that the national print media in Canada are not interested in playing up emotional aspects of the climate

Figure 2.Reprinted with permission of the Calgary Herald (image is as it appeared in The National Post).


change debate using images (keeping in mind of course our earlier finding that articles appearing with any image are thematically more emotive than those without images).

In this exceptional case, however, the polar bear imagery is deeply intertwined with the article text, which casts the animals as victims of climate change at risk due to dwindling food supplies. The article draws on traditional legitimacy-building methods to make the case, quoting multiple concur-ring sources (a government official and an environmental activist), and presenting statistics to prove the population decline of polar bears in the province of Manitoba. In short, this is one of the few examples in our sample of intertextual harmony, whereby image and article establish and reinforce the same narrative. While polar bears appear rarely in the sample (in three percent of cases), they are, as Slocum argues, ‘charismatic … symbols that do not need text with it – the bear can stand alone’. This charisma allows the image to seize hold of the print text in a way that most other images do not.80

Climate Change RefugeesFinally, we consider an extreme example of narrative disconnection between image and print text. An article from The National Post (‘Refugees another side of climate change’) presents a collage of three images alongside its text: a picture of an asteroid hitting the earth, a photograph of a

Figure 3.Reprinted with permission of The Canadian Press.


flooded street in New Orleans following Hurricane Katrina, and a photograph of a forest fire (unfortunately, we were unable to secure permission to reproduce these visuals as they appeared in the article). These images, which are themselves not obviously connected, also have little-to-no direct relevance to the text of the article, which discusses the potential implications of legally acknowledging persons displaced by global warming as refugees. In this respect, the image of New Orleans is the most thematically consistent with the article text, although Hurricane Katrina as an isolated event cannot be directly linked to climate change. By contrast, the image of the forest fire is geographically unattributed and entirely de-contextualized from the issue of human migration. The asteroid is the most notable of the images, given that it is an artist’s rendition of the way an asteroid would look as it pummels from outer space towards Earth. The caption, which does not directly corre-spond to the article’s content, does however provide readers with a frame by which to view the images:

The flooded streets of New Orleans after Hurricane Katrina in 2005 are a foretaste of what could happen to coastal cities as ocean levels rise, climate-change advocates warn. Droughts could also lead to more forest fires. Over the eons, the Earth has had its share of cataclysmic events, such as the one 150 million years ago, when an asteroid struck the planet in Mexico’s Yucatan peninsula.81

There are two plausible interpretations of the image-language interactions in this case. The first is more critical: while the forest fire and New Orleans images are used to provide real-world refer-ents for the discussion of climate change refugees and encourage emotional commitment among readers, the asteroid image and caption text then serve as a subtle assertion that global climate change may be (as the asteroid was) beyond human control – thus framing the narrative as one about consequences rather than causation. The second interpretation is not incompatible with the first, but stresses the narrative difficulties posed by ‘creeping problems’ such as climate change:82 without the benefit of a central metaphor or set of widely-accepted visual tropes, news media are both more free in their choice of visual representation and more constrained by the lack of recog-nizably ‘relevant’ images to attach to stories – leading to situations of narrative ‘scatter’, where visual and printed text pull in different directions and contribute to confusion rather than clarity.

Conclusions Visual communication has played a critical role in many environmental issues, from ozone layer depletion to deforestation to nuclear energy and industrial pollution. Global climate change, despite its status as one of the most talked-about and pivotal environmental challenges of our time, appears to lack key visual symbols or metaphors. While there are many potential candidates to fill this void (melting snow and ice, polar bears, changing coastlines, parched earth, colour-coded maps, hurri-canes), the lack of a central unifying visual trope is affecting the ‘social and cultural life’ of this issue. In our view, the dearth of clear imagery around global climate change makes it more difficult for ordinary citizens to visualize potential impacts and consequences, and to link (often) abstract language claims to the real world and to everyday life.

As we have discussed, the power of images comes from their ability to blend fact and emotion; to engage viewers as ‘witnesses’ rather than as detached consumers of information and claims. Given the different strengths of images and language, we have argued that these ought to be con-sidered as ‘co-constructors’ of environmental messages. While newspaper reporting is not the pri-mary source of visual communication in most people’s lives, it remains the most important source of information (in print and online formats) for the general public on environmental issues.83 They are also sites of intense interaction between imagery and language, where both are packaged together with the intention (presumably) of presenting cogent narratives to audiences.


Our research has addressed three main questions. First, we asked: How is climate change being visualized in the Canadian context? ‘Surface’ content coding showed that most images were focused on human subjects, followed by the natural world (although the modal subcategory here was an urban landscape), and industry and technology. Smith and Joffe have argued that images of people are an important means of localizing global climate change and bringing the issue closer to their lived space.84 Given that the great majority of images across all categories could be identified as Canadian, we agree with this conclusion. However, we also note that the images in our sample were predominantly unemotional in both their denotative and connotative (latent) themes. For instance, very few images sought to directly capture or represent impacts of climate change, and expected visual tropes of polar bears, coastlines, and snow and ice were rarely seen (see Table 1). Therefore, we argue while the images do localize the climate change issue, this localization is limited – embedding climate change in everyday and frequently unexceptional discussions about Canadian politics, business, and land and resource management.

Our second guiding question was: What kinds of images and metaphors (if any) dominate dis-courses surrounding climate change in Canada? Surprisingly, the only prototypically Canadian images to emerge from our research involve the oil sands, and even these appeared in only nine percent of cases. Given that the oil sands issue is more about causation than consequences, this is further evidence that impacts are not being visualized in the Canadian context. It is also notable that, in comparison to US and European media,85 the Canadian newspapers are not giving much visual representation to celebrities or environmentalists. In short, the Canadian media is notable for its lack of expected patterns in visual representation of climate change.

Finally, our third research question was: Are there differences in the way that climate is ‘talked about’ compared to how it is ‘seen’? The answer is certainly in the affirmative, although not in the way that was anticipated. Joffe and Smith’s study of images in UK coverage of climate change found that imagery was generally one-sided; that while the text of an article may contain ‘discus-sions of controversy, unknowns, and debate, … the concretization of impacts in newspaper images speaks against the argument that climate change science is uncertain’.86 In other words, they argue that images have ‘the last word’ in delivering an ultimately one-sided climate change narrative to audiences. Our study, which involves a more rigorous analysis of image-language interactions, leads us to conclude that images and text are having markedly ‘different words’ on the climate change issue. Overall, we found that image and article frequently refer to completely different dimensions of the climate change issue, thus presenting multiple and sometimes even competing narratives to readers. Inter-media coherence, where image and language present a mutually rein-forcing interpretive framework, was rare (see The Polar Bears example above). Most telling in this regard was our finding that the presence or absence of any image had a greater impact on the the-matic content of the article than did the type of image. Even with key controls in place, our statisti-cal tests showed that articles accompanied by an image tended to have more moral and emotional themes than articles without images, while being less likely to present technical arguments about causation (see Table 2). In contrast, comparisons across different types of image (people, nature, industry/technology) showed notable consistency on key moral dimensions, particularly attribu-tions of blame, discussions of victims and victimhood, and mentions of the future (see Table 3).

This finding suggests that images are definitely not driving the content of articles, nor is it likely that articles are driving the content of images. Rather, it appears that in many cases journalists and editors are attaching images post-facto to articles that tend to be morally or emotionally edgy, regardless of the content of those images. As argued by Boykoff, ‘creeping’ environmental prob-lems are difficult to cover because of the lack of abrupt events to spur interest.87 Our research shows that this problem applies to the visual realm as well – a lack of central visual tropes means


that news media have more choice of imagery, but ultimately less focus, in their representation of the issue. In the Canadian context at least, image-language disjuncture is the norm. In our view, this is not a case of misleading the public (in most cases), but is indicative rather of an important ‘blank spot’ in current claims-making about climate change. The issue is still awaiting its iconic image(s) with the power to steer arguments and ‘speak for themselves’ in the public sphere.

Acknowledgements

Research for this paper was conducted with the support of the Social Sciences and Humanities Research Council of Canada. We would like to thank Eric Dugas for his assistance in compiling the data. Thanks also to the Editor of cultural geographies and to three anonymous reviewers for their valuable comments and suggestions.

Notes

1 R. Rorty, Philosophy and the Mirror of Nature (Princeton: Princeton University Press, 1979), p. 12. 2 J. Durfee and J. Corbett, ‘Context and Controversy: Global Warming Coverage’, Nieman Reports, 59(4),

2005, p. 88. 3 D. Levy and D. Egan, ‘A Neo-Gramscian Approach to Corporate Political Strategy: Conflict and

Accommodation in the Climate Change Negotations’, Journal of Management Studies, 40(4), 2003, pp. 803–29; M. Lahsen, ‘Technocracy, Democracy, and US Climate Politics: The Need for Demarcations’, Science, Technology and Human Values, 30(1), 2005, pp. 137–69.

4 L. Antilla, ‘Self-Censorship and Science: A Geographical Review of Media Coverage of Climate Tipping Points’, Public Understanding of Science, 19(2), 2010, pp. 240–56.

5 W. Gamson and A. Modigliani, ‘Media Discourse and Public Opinion on Nuclear Power’, American Journal of Sociology, 95(1), 1989, pp. 1–37; D. Loseke, Thinking about Social Problems: An Introduction to Constructionist Perspectives (New York: Aldine de Gruyter, 2003); N.W. Smith and H. Joffe, ‘Climate Change in the British Press: The Role of the Visual’, Journal of Risk Research, 12(5), 2009, pp. 647–63.

6 K. Litfin, Ozone Discourses: Science and Politics in Global Environmental Cooperation (New York: Columbia University Press, 1994); S. Ungar, ‘Knowledge, Ignorance and the Popular Culture: Climate Change versus the Ozone Hole’, Public Understanding of Science, 9, 2000, pp. 297–312; R. Cox, Environmental Communication and the Public Sphere (Thousand Oaks, CA: Sage, 2006).

7 Ungar, ‘Knowledge and Ignorance’, p. 305. 8 Ungar, ‘Knowledge and Ignorance’, p. 305. 9 R. Slocum, ‘Polar Bears and Energy-Efficient Light Bulbs: Strategies to Bring Climate Change Home’,

Environment and Planning D, 22, 2004, pp. 413–38; J. Doyle, ‘Picturing the Clima(c)tic: Greenpeace and the Representational Politics of Climate Change Communication’, Science as Culture, 16(2), 2007, pp. 129–50; S. Bronniman, ‘Picturing Climate Change’, Climate Research, 22, 2002, pp. 87–95; Smith and Joffe, ‘Role of the Visual’.

10 S. Chase, ‘More Canadians Leaning Right, Poll Finds’, Globe and Mail, 12 March 2010. Article # 1499051.

11 C. Sandilands, ‘Multiculturalism, Wilderness and the Desire for Canada’, Space and Culture, 2(4–5), 2000, pp. 169–77; E. Mackey, ‘Death by Landscape: Race, Nature, and Gender in Canadian Nationalist Mythology’, Canadian Woman Studies, 20(2), 2000, pp. 125–30; S. Sorlin, ‘The Articulation of Territory: Landscape and the Constitution of Regional and National Identity’, Norwegian Journal of Geography, 53(2), 1999, pp. 103–12.

12 Slocum, ‘Polar Bears and Energy-Efficient Light Bulbs’.


13 For instance, several environmental groups collaborated on a campaign entitled ‘Save Hockey, Stop Climate Change!’ A shortlived website was constructed (inactive at the time of writing) entitled www.savehockey.ca

14 Antilla, ‘Self-Censorship’, p. 245.15 D. Demeritt, ‘The Construction of Global Warming and the Politics of Science’, Annals of the Association

of American Geographers, 91(2), 2001, p. 311.16 The National Post has recently experienced financial difficulties, which means that it is no longer

distributed in all regions of the country.17 Loseke, Social Problems; J. Hannigan, Environmental Sociology (London: Routledge, 2006).18 N. Young and R. Matthews, The Aquaculture Controversy in Canada: Activism, Policy, and Contested

Science (Vancouver: UBC Press, 2010), pp. 159–60.19 M. Schaefer, ‘From Public Understanding to Public Engagement: An Empirical Assessment of Changes

in Science Coverage’, Science Communication, 30(4), 2009, pp. 475–505.20 Hannigan, Environmental Sociology, pp. 77–8.21 Smith and Joffe, ‘The Role of the Visual’, p. 648.22 A. Iyer and J. Oldmeadow, ‘Picture This: Emotional and Political Responses to Photographs of the

Kenneth Bigley Kidnapping’, European Journal of Social Psychology, 36(5), 2006, pp. 635–47; H. Joffe, ‘The Power of Visual Material: Persuasion, Emotion and Identification’, Diogenes, 217, 2008, pp. 84–93; S. O’Neill and S. Nicholson-Cole, ‘Fear Won’t Do It: Promoting Positive Engagement with Climate Change through Visual and Iconic Representations’, Science Communication, 30(3), 2009, pp. 355–79.

23 A. Hansen and D. Machin, ‘Visually Branding the Environment: Climate Change as a Marketing Opportunity’, Discourse Studies, 10(6), 2008, pp. 777–94.

24 Ungar, ‘Knowledge and Ignorance’.25 D. Graber, ‘Say it with Pictures’, Annals of the Academy of Political and Social Science, 541(1), 1996,

pp. 85–96; A. Boholm, ‘Visual Images and Risk Messages: Commemorating Chernobyl’ Risk, Decision and Policy, 3(2), 1998, pp. 125–43.

26 Joffe, ‘Power of Visual Material’, p. 85.27 M. McLuhan, Understanding Media: The Extensions of Man (New York: McGraw-Hill, 1964).28 M. Boykoff, ‘We Speak for the Trees: Media Reporting on the Environment’, Annual Review of

Environment and Resources, 34, 2009, p. 437.29 S. Weart, The Discovery of Global Warming (Cambridge, MA: Harvard University Press, 2008);

Bronnimann, ‘Picture This’, p. 90.30 S. Ungar, ‘The Rise and (Relative) Decline of Global Warming as a Social Problem’, Sociological

Quarterly, 33(4), 1992, pp. 483–501; M. Hulme, Why We Disagree about Climate Change (New York: Cambridge University Press, 2009).

31 Ungar, ‘Rise and (Relative) Decline’; Boykoff, ‘We Speak for the Trees’.32 C. Trumbo, ‘Constructing Climate Change: Claims and Frames in US News Coverage of an

Environmental Issue’, Public Understanding of Science, 5, 1996, pp. 269–83; see also A. Carvalho, ‘Ideological Cultures and Media Discourses on Scientific Knowledge,’ Public Understanding of Science, 16, 2007, pp. 223–43.

33 Demeritt, ‘Construction of Global Warming’, p. 307.34 J. Simpson, M. Jaccard and N. Rivers, Hot Air: Meeting Canada’s Climate Change Challenge (Toronto:

McClelland and Stewart, 2007), p. 49.35 M. Boykoff, ‘From Convergence to Contention: United States Mass Media Representations of

Anthropogenic Climate Change Science’, Transactions of the Institute of British Geographers, 32, 2007, p. 485.


36 K. McComas and J. Shanahan, ‘Telling Stories about Global Climate Change: Measuring the Impact of Narratives on Issue Cycles’ Communication Research, 26, 1999, pp. 30–57.

37 M. Boykoff and M. Goodman, ‘Conspicuous Redemption? Reflections on the Promises and Perils of the Celebritization of Climate Change’, Geoforum, 40, 2009, pp. 395–406.

38 A. McCright and R. Dunlap, ‘Defeating Kyoto: The Conservative Movement’s Impact on US Climate Change Policy’, Social Forces, 50(3), 2003, pp. 348–73; L. Antilla, ‘Climate of Scepticism: US Newspaper Coverage of the Science of Climate Change’, Global Environmental Change, 15, 2005, pp. 338–52; M. Boykoff and J. Boykoff, ‘Balance as Bias: Global Warming and the US Prestige Press’, Global Environmental Change, 14, 2004, pp. 125–36.

39 Smith and Joffe, ‘The Role of the Visual’, p. 649.40 D. Brossard, J. Shanahan and K. McComas, ‘Are Issue-Cycles Culturally Constructed? A Comparison

of French and American Coverage of Global Climate Change’, Mass Communication & Society, 7(3), 2004, pp. 359–77.; U. Olausson, ‘Global Warming – Global Responsibility? Media Frames of Collective Action and Scientific Uncertainty’, Public Understanding of Science, 18, 2009, pp. 421–36.

41 L. Antilla, ‘Climate of Scepticism’; N. Young and E. Dugas, ‘Representations of Climate Change in Canadian National Print Media’, Canadian Review of Sociology (forthcoming).

42 Smith and Joffe, ‘The Role of the Visual’, p. 652.43 Smith and Joffe, ‘The Role of the Visual’, p. 658.44 Smith and Joffe, ‘The Role of the Visual’, p. 658.45 O’Neill and Nicholson-Cole, ‘Fear Won’t Do It’.46 O’Neill and Nicholson-Cole, ‘Fear Won’t Do It’, p. 375.47 Ungar, ‘Knowledge and Ignorance’; O’Neill and Nicholson-Cole, ‘Fear Won’t Do It’.48 K. Krippendorff, Content Analysis: An Introduction to its Methodology (London: Sage, 2004).49 N. Fairclough, Discourse and Social Change (Cambridge: Polity Press, 1992); L. Lees, ‘Urban

Geography: Discourse Analysis and Urban Research’, Progress in Human Geography, 28(1), 2004, p. 105.

50 It is unclear why The Globe and Mail makes more use of images related to climate change than does The National Post, although we note that the former does publish more articles on the subject more generally (371 versus 290 for the year 2008 as a whole).

51 H. Jamieson, Visual Communication: More than Meets the Eye (Chicago: University of Chicago Press, 2007), p. 70.

52 While it may appear that N=100 is a conveniently round number, it is the legitimate outcome of our screening process.

53 Jamieson, Visual Communication, p. 70.54 Smith and Joffe, ‘The Role of the Visual’.55 Smith and Joffe, ‘The Role of the Visual’, p. 659.56 Boykoff and Goodman, ‘Conspicuous Redemption’, p. 395.57 Slocum, ‘Polar Bears and Energy-Efficient Light Bulbs’.58 Slocum, ‘Polar Bears and Energy-Efficient Light Bulbs’.59 Trumbo, ‘Constructing Climate Change’.60 Contact the second author for a copy of the content coding framework used.61 Table 2 only presents findings that achieve the <.05 level of statistical significance, with the exception of

column 3.62 Smith and Joffe, ‘The Role of the Visual’, p. 660.63 Slocum, ‘Polar Bears and Energy-Efficient Light Bulbs’.64 In order to perform these statistical tests, the image categories were made mutually exclusive. Images

were originally coded across multiple categories in 28 of the 100 cases (for instance, a person standing


in a natural landscape). To make the categories mutually exclusive, we prioritized in this order: nature, industry, people, yielding the following: images of people, N=38; images of nature, N=25; images of industry and technology, N=35.

65 Given that word count and type of article are consistent across all types of image, there is no need to introduce these as controls. Therefore, chi-square tests are used in these comparisons instead of regression analysis.

66 Hannigan, Environmental Sociology, p. 77.67 Loseke, Social Problems.68 Jameison, Visual Communication, p. 70.69 N. Fairclough, ‘Discourse and Text: Linguistic and Intertextual Analysis within Discourse Analysis’,

Discourse & Society, 3(2), 1992, p. 195.70 Loseke, Social Problems, p. 86.71 S. McCarthy, ‘National Carbon Market Faces Opposition’, The Globe and Mail, 12 March 2008), p. B.3.72 C. Tait, ‘Greenhouse “Intensity” to Rise’, The National Post, 19 June 2008, p. FP.4.73 ‘Greenhouse “Intensity” to Rise’.74 ‘Greenhouse “Intensity” to Rise’.75 ‘Greenhouse “Intensity” to Rise’.76 M. Mittelstaedt, ‘Manitoba Recognizes Threat to Polar Bears’, The Globe and Mail, 8 February 2008, p. A. 1.77 C. Dorman, ‘A Taxonomy of Visual Metaphors’, in R. Paton and I. Neilson (eds) Visual Representations

and Interpretations (New York, Springer, 1999), pp. 279–88.78 Smith and Joffe, ‘The Role of the Visual’.79 Loseke, Social Problems, p. 76.80 Slocum, ‘Polar Bears and Energy-Efficient Light Bulbs’, p. 15.81 C. Offman, ‘Refugees Another Side Effect of Climate Change’, The National Post, 21 June 2008, p. A. 9.82 Boykoff, ‘From Convergence to Contention’.83 Antilla, ‘Self-Censorship’, p. 245.84 Smith and Joffe, ‘The Role of the Visual’.85 Boykoff and Goodman, ‘Conspicuous Redemption’.86 Smith Smith and Joffe, ‘The Role of the Visual’, p. 658.87 Boykoff, ‘From Convergence to Contention’.

Biographical notes

Darryn Anne DiFrancesco was a graduate student at the University of Ottawa’s School of International Development and Global Studies at the time of writing. Her Master’s thesis was a New Institutional analysis of an Aboriginal community-owned enterprise in Western Canada. Additional research interests include cultural studies, discourse analysis, and social understandings of environment, place, and health. She begins her PhD in Sociology at the University of British Columbia in fall 2010. She can be contacted at: [email protected]

Nathan Young is Assistant Professor of Sociology at the University of Ottawa. His research interests include environmental sociology, economic sociology, and risk communication. Current research projects include Public Discourses on Climate Change in Canada and The Socio-Economics of Climate Change in Coastal Canada. His most recent book is The Aquaculture Controversy in Canada (University of British Columbia Press). He can be contacted at: [email protected]

Documents

Seeing climate change: the visual construction of global ...blogs.ubc.ca/difrancesco/files/2011/04/Seeing... · construction of global warming in Canadian national print media Darryn