News media and food scares: the case of contaminated salmon

Abstract

In 2004, Science published a study on organic contaminants in farmed salmon. The study had a clear normative message and worked strategically and successfully to gain worldwide media attention. In this article, we investigate global media coverage of the study. The varying types of attention and different framings of selected national broadsheets in 14 countries are analysed. (Framing is where a complex and often uncertain reality is simplified in order to support a specific understanding of the issue and/or push an agenda.) The results show that even if the scientists and the sponsor of the study had a clear ambition to publicize and disseminate their results and normative proposals to the wider society, the newspapers did not act as a passive medium for distributing the original message. Instead, diverging understandings and framings were developed.

By way of conclusion, it is stated that ambitious strategies for attracting media attention may be successful in terms of media coverage; this does not, however, mean that the message is passively transmitted. The national context and the logic of media cause issues to be framed in specific ways with the aim of telling stories and catching the attention of the reader.

Keywords:

• Food scares
• mass media
• science communication
• health risk
• organic contaminants
• salmon

1. Introduction

During the last couple of decades we have witnessed a dramatic increase in reporting about the different kinds of risks that human beings face, and through mass media people around the world are told stories about a wide range of risks to the environment and human health. Some researchers even claim that we in the West live in a ‘culture of fear’, in which fear, based on secondary mediated experiences, is becoming widespread (Furedi 2002; cf. Beck 1992; Giddens 1991). The empirical evidence of such fear among populations is, however, scanty (Gouldson et al. 2007).

Without doubt, however, there are intimate links between science, politics, media and public cognitions of risks. The media feeds us with different types of scare stories, reporting on global epidemics, natural disasters, climate change, international terrorism, everyday violence and different types of food and health risks. Regarding the latter we constantly hear about such risks as salmonella in chicken, pesticide residues in fruit and vegetable, acrylamide in cooked food, dioxins in fish, BSE (bovine spongiform encephalopathy) in beef, hormone-disrupting chemicals in animal products and risks involving genetically modified vegetables.

As laypersons we rarely have any direct experience of these kinds of risks, since they often lie beyond the range of human sense perception. Instead, science defines the risks and the scale of the problems (Lidskog et al. 2005). However, scientific statements and results have to be translated in ways that make them understandable and meaningful for non-scientific actors within industry, politics and the public. A space is opened up for interpreting and disseminating expert knowledge; and over the last couple of decades there has been a drastic increase in the number and size of the organizations that populate this space (Held et al. 1999). Several researchers have emphasized the decisive role organizations such as environmental movements, public agencies, industrial companies and interest organizations play in the creation of people's consciousness and understanding of environmental risks (Hansen 1991; Lidskog 1996; Jamison 2001). All these organizations are, however, dependent on the media. It is primarily through the reporting and framing in the media that citizens have become acquainted with buzzwords such as acidification, ozone depletion and the greenhouse effect, as well as a number of health risks associated with food products. (Framing is where a complex and often uncertain reality is simplified in order to support a specific understanding of the issue and/or push an agenda.)

This is an important reason underlying the increased importance that the governmental and scientific communities attach to science communication (Lidskog & Sundqvist 2004). Science communication through the media has become important in two ways. First, there is political pressure on science demanding more open knowledge production and an obligation to inform society in general about ongoing research and results (Nowotny et al. 2001). Today, universities and academic institutions – apart from conducting research and educating students – have a mission of research communication: to provide information to people and organizations outside the academic world. In order to reach the large and diverse general public with information about scientific work and new findings, science has to attract media attention and generate reporting on scientific findings. Thus, apart from their personal wish and right to inform and warn the public about health risks and other problems, researchers now have a formal obligation to inform actors about current research findings.

Secondly, the politicization of scientific funding in which certain areas are prioritized and receive more governmental funding at national and international levels, leads to a struggle for attention in the intersection between the scientific and the political communities. One important way for science to get this attention is via the media. Politicians are extremely sensitive to media reporting, and media attention will increase the chances for research questions to be put on the political agenda and become prioritized areas. One example of this is how the Swedish ‘acrylamide alert’ was given attention by the European press, which led to immediate reactions by different authorities. As Löfstedt (2003: 412 – 413) has found:

The British Food Standards Agency announced that it would monitor the situation and after the scare allocated funding for independent UK tests […] The World Health Organization (WHO) announced on the 26th of April that it would hold an urgent expert consultation on acrylamide because of the Swedish announcement.

The increased competition for industrial and private funding further pushes science to strive for public attention via the media. Due to modern communication technology, scientific results can now be made public almost instantaneously throughout the world. Through satellite news services, cable distributors and the Internet, news bulletins are delivered 24 hours a day. Journalists worldwide continuously check BBC World, CNN International, BBC Online, Reuters Internet portal, CNN Interactive, and so on, to find interesting stories. CNN International, for example, uses 23 satellites to reach all corners of the world (Chalaby 2003).

The media is, however, anything but a passive and neutral part of a transmission process in which scientific results are processed for political and public consumption. Instead, with the general aim of producing news, the media rather unreservedly adds new meanings to the message originally formulated by the scientists. This is not to say that the news reporting in the media necessarily biases research results (Wåhlberg & Sjöberg 2000). Scientific results may be correctly referred to, although told in ways that fit into media logic and media interests. This is not least the case when it comes to scientific findings that can be constructed as urgent threats to human health, which have proven to be of central importance in seizing the public's attention.

Furthermore, the creation of news takes place in specific social settings, which means that news items are converted to fit specific media interests and national audiences. Thus, even if there is a global spread of news, an issue is not reported and framed in a uniform way. The social, cultural and political contexts, as well as what kinds of events that are addressed, are still significant for how global news is framed, as well as for how citizens interpret and make sense of it (Hjarvard 2001). Some even claim that the national setting is always vital in news reporting, and that global news items are almost always the object of domestication processes (Clausen 2004).

Since the media is an increasingly important mediating actor for science, as well as for political and public interest in science, it is of vital importance to study how the media covers research on environmental risks. In this article, we do not have the ambition to provide a full-fledged analysis of national differences in the media reporting on risks. Instead, we merely wish to discuss the media's role in reporting on environmental risk; and to make some preliminary observations on the interaction between science reporting, the media's news creation and national context. The article is of exploratory character, focusing on the broadsheet press coverage in various national contexts of a particular scientific report on a food risk.

Empirically, the article presents a brief analysis of how a study published early in 2004 in Science was covered by certain broadsheets in a number of different countries. The article concerned organic contaminants in farmed salmon, and raised warnings that consumption of farmed Atlantic salmon may pose health risks. What makes this case relevant is that ambitious and professional efforts were made to disseminate the results to the wider society, and this ambition was successful in the sense that the study was reported worldwide. In particular, we investigate how national dailies in 14 countries made news of this study. Guiding questions are: how was this study reported and framed in different national newspapers? If there are national differences, how can they be explained? The exploratory character of this study makes it impossible to draw any firm empirical conclusions. However, it provides an interesting input to the wider discussion on how the media produces meaning on scientific reports on risks to human health and the environment. The presentation of preliminary results from our study at an international conference on organic pollutants, where natural scientists were eager to discuss the role of the media in science communication, further strengthens this. 1

The paper comprises five parts, this introduction being the first. The second part consists of a brief review of the discussion on ‘food scares’, which are a specific kind of risk popular with the media. The third part consists of the empirical study: a content analysis of quality newspaper coverage of the article on contaminants in farmed salmon. It is found that especially the UK and Norway had divergent media images of the food scare. The fourth part analyses these differences. In the fifth and concluding section, we return to the issue of food scares and the media. We state that even if science is important in triggering news reporting, it is the media that is pivotal in the way an issue is framed and disseminated in different national settings.

2 Food scares and the media

Food-related risks are today seen as ‘hot’ news items, to which the media willingly pays attention (Seale 2002). When the media transmits these kinds of risks they may become potential, dramatic and urgent dangers to human health; ‘food scares’ are constructed (Kitinger & Reilly 1997; Reilly 1999).

Food scares are also interesting for the media, since they may involve powerful interests; sometimes they concern national economics or national cultural values. A recent example of this is the Russian import ban on Norwegian farmed fresh salmon in early 2006, which in Norway resulted in political mobilization, scientific investigations and intensive diplomatic activities with the aim to convince Russia to withdraw its ban. 2

According to Adam (1998) the BSE food scare has been the greatest news event in Great Britain since the Second World War. Often, however, food scares flare up in the media for short periods and then disappear from the agenda, such as was the case with acrylamide in cooked food (Löfstedt 2003). Sometimes they are recurrent themes in the media reporting for years, with or without a few dramatic peak alerts.

News reporting does not influence the public's understanding in any simple way, but instead, items of news are actively taken in and interpreted by the public. Media reception research in general has shown that people in complex ways contextualize, interpret and find meaning in media reports by relating them to their previous knowledge, socio-cultural experiences, perspectives and values (Morley 1980; Höijer 1992, 1998). However, at the same time this should not be seen as meaning that the media reports have no effect at all on people's understanding and behaviour. There are numerous examples of media campaigns that have strongly influenced public opinions and consumer habits. The most well known is probably the Greenpeace media strategy against Shell's proposal to sink the oil platform Brent Spar in the North Sea (Löfstedt & Renn 1997; Tsoukas 1999).

Food scares may thus influence the citizens' risk consciousness and affect their consumption habits. To give some further examples of this, egg sales dropped 20% in England after a media alert about salmonella in eggs in 1989, and beef sales fell 37% after the BSE alert in 1996 (Miller & Reilly 1995; Young & Morris 2001: 170). Potato chip sales in Sweden went down after the acrylamide scare in 2002 (Löfstedt 2003). However, these changes may often have been of temporary nature and when the media alarm has diminished and the public discussion has faded away, risk consciousness dissolves and previous habits are re-activated.

Food scares may be related to the broader discourses of health promotion and health risks, which, as pointed out by several researchers, have become a predominant concern in our society (Petersen & Lupton 1996). In line with general processes of privatization and individualization in society (Bauman 2001; Beck & Beck-Gernsheim 2002; Höijer et al. 2006) health issues are transformed into personal responsibilities. Citizens are reduced to individual consumers, and it is up to each and every one of them to avoid risks. Meanwhile, the mass media have been keen to exploit health issues from such an individualized perspective (Kamin 2005). The newspapers may, for example, present lists of food dangers as guidelines for the individual consumer, as they did in Sweden after the acrylamide alert in Spring 2002 (Löfstedt 2003). Indeed, processes of individualization provide many opportunities for the media to make stories about food scares.

It is not just the threat food risks pose to the health of individuals that makes them attractive stories for the media. Another important aspect is that food risks often involve different and contrasting expert opinions, where different researchers have come to very different conclusions. The media is often quick to emphasize scientific controversies, since they fit well into the media logic of polarization (Väliverronen 2001). Focusing on conflicts and controversies is a narrative strategy believed to make stories more appealing. It is typical for environmental news that experts are arguing about what to do and how dangerous the situation is (Darley 2000).

3 The study: Organic contaminants in farmed salmon

3.1 The article in Science

On 9 January 2004, Science published an article on organic contaminants in farmed salmon (Hites et al. 2004). The article, authored by five scientists and a representative of a citizens group from the USA and Canada, investigated the amount of organochlorine contaminants in 700 farmed and wild salmon collected from around the world. Farmed Atlantic salmon from eight major production regions in the Northern and Southern hemispheres were purchased from wholesalers; and farmed Atlantic salmon were purchased from supermarkets in 16 large cities in North America and Europe. All the samples were homogenized and analysed by gas chromatographic high-resolution mass spectrometry. Fourteen different organochlorine contaminants were analysed, and 13 of these contaminants were found to be significantly more concentrated in farmed salmon than in wild salmon. A proposed explanation of this difference was their different diets. A deeper analysis was then made of four of the contaminants – PCBs, dioxins, toxaphene and dieldrin.

Based on their findings, the authors calculated the risk to human health of salmon consumption. Using a cumulative risk assessment approach developed by the US Environmental Protection Agency, the comparative health risks of consuming farmed and wild salmon were assessed. The highest levels of risk were found in the farmed salmon fillets purchased at stores in Frankfurt, Edinburgh, Paris, London, and Oslo. According to the US Environmental Protection Agency (EPA), the concentration of the contaminants indicates a recommendation of no more than one to two meals per month of farmed salmon. The authors state (Hites et al. 2004: 229):

Our data indicate that farmed salmon have significantly higher contaminant burdens than wild salmon and that farmed salmon from Europe are significantly more contaminated than farmed salmon from South and North America.

The article states that fish is healthy food, but that the result of the study suggests that consumption of farmed salmon may result in exposure to a variety of persistent bio accumulative contaminants with the potential for health risks. ‘Although the risk/benefit computation is complicated, consumption of farmed Atlantic salmon may pose risks that detract from the beneficial effects of fish consumption’ (Hites et al. 2004: 229).

Despite the fact that the authors are using the word ‘indicates’ about their results, they end with a firm prescriptive conclusion, claiming that the study demonstrates the importance of labelling salmon as farmed and identifying the country of origin.

Thus, the article does more than just report on its discovery of higher levels of organic contaminants in farmed as opposed to wild salmon, and European as opposed to North and South American farmed salmon. It also recommends a course of action by proposing a labelling system to make it possible for consumers to choose wild salmon or at least farmed salmon from North or South America. The recommendation can be contested, because a number of different recommendations are possible to derive from this study, such as strict tolerance levels for organic contaminants in salmon. However, by putting consumer choice at the centre, this kind of regulation is more dependent on media reporting. In contrast to state regulation through legislation, this kind of voluntary regulation requires that information about a substance or product's environmental harmfulness be widely distributed in society.

Furthermore, the proposed recommendation frames the problem as lying with European farmed salmon, which may overshadow the fact that even if American farmed salmon has a lower level of contaminants than the European, it still has a very high level. Applying the US EPA's recommendation, the salmon farmed in North and South America should not be eaten more than twice a month (compared to about once for the European). Thus, a successful introduction of the proposed labelling strategy will create a number of winners (e.g. Chile, Canada) and losers (e.g. Norway, Britain, Germany).

The Pew Charitable Trust, which sponsored the study with 2.5 million USD, actively worked to get media attention as a means of distributing the results to a wider audience than the research community. 3 Besides the press conference, arranged by Science a few days before it published the salmon study, a number of initiatives were made to disseminate the research results. The main author of the salmon article states:

The Pew Trust hired a NGO to generate a press release and a web-site; this NGO hired a media communications person for a couple of weeks to pitch the story. It was all timed to come out in the week when the paper was published. […] Certainly if they had not paid for the NGO then the story would have received less attention. 4

3.2 Content analysis of the press coverage

Broadsheets from 14 countries covering a period of 2 weeks were studied, starting with the day after the press conference (i.e. 8 to 21 January 2004). Guiding criteria in the selection were to cover nation states from different regions in the world, with rather large coastal areas and whose fisheries are of economic importance for the country. Furthermore, the selection covered countries that were investigated in the study and which from the point of view of the result became winners (Chile, USA, Canada) and losers (European countries, especially Britain, Norway and Germany). The selection also included China, Japan and South Africa; that is countries from other regions than the salmon study pinpointed (Europe, North America and South America).

Two morning papers were selected from the USA, one from the west coast and one from the east coast. The reason for this was that the study was sponsored by a US trust with the objective of promoting positive change in America. Also, two newspapers were selected from Great Britain, one English and one Scottish. The reason for this was that the salmon study found that farmed salmon from Scotland had the highest concentration of the contaminants, and it therefore was of interest to include a Scottish newspaper. In Table I, the names of the newspapers are given.

Table I. Salmon-related articles found in the newspapers for 8 – 21 January 2004.

Newspapers	Articles	Size (in characters)
The Scotsman (Scotland)∗	24	77 600
The Guardian (England)^†	13	55 800
Vancouver Sun (Canada)	9	29 700
Aftenposten (Norway)	7	11 200
Le Figaro (France)	5	9900
New York Times (USA, east coast)	4	15 200
La razón (Spain)	4	9800
San Francisco Chronicle (USA, west coast)	2	8800
Dagens Nyheter (Sweden)	2	4900
The Santiago Times (Chile)	1	4400
Mail & Guardian (South Africa)	1	1300
Clarín (Argentina)	1	400
Frankfurter Allgemeine (Germany)	1	200
Volkstrant (Holland)	–	–
Yomiuri shinbun (Japan)	–	–
China Daily (China)	–	–
∗The Scotsman contained 19 articles, and Scotland on Sunday (The Scotsman's Sunday edition) contained five articles.
^†The Sunday edition, The Observer, is included. The Guardian contained nine articles and The Observer contained four articles.

The press material in English, French, German, Norwegian and Swedish was collected from the archives of the newspapers' web pages and coded by us (in total 11 newspapers). 5 The Net editions of the morning newspapers from Japan, China, Argentina, and Spain, and a paper edition from Holland were selected and coded by native speakers in each country.6 These five persons were selected through a research network of dioxin researchers to which we had access. The participating ‘coders’ received instructions and a coding schema, which they used in their data gathering of the selected newspapers' coverage of the salmon study. Of course, this way of coding the data has its disadvantages, such as the risk of a systematically biased coding. However, as will be shown below, the study was sparsely reported in the selected newspapers of these countries, and therefore this kind of methodological problem is not relevant to the discussion in this article.

Frames and framing mean that a complex and often uncertain reality is simplified in order to support a specific understanding of the issue and/or push an agenda (Entman 1993; Schön & Rein 1994; Fischer 2003). By selecting and omitting aspects of a phenomenon, a complex issue is simplified and then packaged in a way that guides our thoughts in certain directions. Through framing, opportunities for action – or inaction – are created, for example through political mobilization or green consumerism (Klintman & Boström 2004).

All issues are possible to frame in different ways. It is a general process valid for all types of communication including scientific ones. Through framing, journalism offers its readers a specific way to interpret and understand reality (Väliverronen 2001: 23). Thus, a framing not only has implications for how a problem is defined but also for which solution is made relevant (Entman 1993). Frames may work as ideological instruments related to political, industrial, economic, cultural or national interests.

In our analysis, we have used a coding schema to grasp different framings of newspaper articles on the salmon study. The coding schema covered six variables and a manual coding was performed of the investigated articles:

	number of articles
	article size
	main subject of each article in five categories: health (focus on health aspects related to eating salmon), economic questions (focus on economic issues, such as job loss or sales drop), political questions (focus on how politicians handle the topic), research (focus on what the study and/or other related research stated), and mixed subjects (no obvious main subject but mixes different subjects).
	sources used in nine categories: public authority, politicians, researchers, fishermen, journalists, company spokesmen, citizens, environmental groups, and others.
	connections to other food scares (such as BSE, dioxins in other food, etc.). The variable dichotomized into yes/no.
	constituting a health risk (four categories: yes, no, mixed and not discussed, see Table II for clarification).

Table II. Main subject of the articles and the focus of the stories (number of articles given in parentheses).

	The Scotsman (24)	The Guardian (13)	Aftenposten (7)	Vancouver Sun (9)
Main subject
Health	11	7	3	5
Economic questions	5	3	4	–
Political questions	3	1	–	1
Research	4	1	–	3
Mixed	1	1	–	–
Main sources∗
Public authority	4	6	3	5
Politicians	7	–	–	–
Researchers	6	7	4	3
Fishermen	2	1	–	1
Journalists	–	1	–	–
Company spokesmen	10	6	4	2
Citizens	1	1	1	–
Environmental groups	–	3	–	2
Others	1	–	–	–
Connections to other food scares	3	6	–	4
Does the salmon scare constitute a health risk? ^†
Yes	5	4	1	2
No	8	1	4	4
Mixed	7	6	–	2
Not discussed	4	2	2	1
∗The main sources have been double coded; that is, more than one main source can be found in each article.
^†‘Yes’ means that the article explicitly argues that it is dangerous to eat farmed salmon, and ‘no’ means that the article puts forth arguments against it. If both positive and negative arguments are mentioned in relation to the danger of eating farmed salmon, the article is coded as mixed. ‘Not discussed’ means that the article does not mention any health risk at all.

4 Results

As Table I shows, the study of organic contaminants in farmed salmon was widely distributed in the world and in that respect it can be seen as a global news item. However, in the Asian and Dutch newspapers studied there were no articles about the issue. 7 The salmon story was given varying degrees of attention by different national newspapers. Of all the morning papers, the ones from Great Britain paid the most attention to the story. The Canadian and Norwegian newspapers also covered the story in several articles, while it, to some extent, received lesser attention in other countries. The Swedish morning paper, for example, had only two shorter articles about the new scientific findings.

The four newspapers with the highest number of articles about the salmon scare were singled out for deeper analysis. A number of different subjects were brought up in the newspapers, but health risks were the dominating subject; in all four newspapers about half of the articles framed the subject as a health issue. Questions were raised about the dangers of eating farmed salmon and how organic contaminants affect humans. In some articles it was explicitly claimed that farmed salmon is a health risk, in other articles that it is not.

Besides these kinds of articles, there were articles that did not take any firm standpoint, but provided arguments in favour of, as well as against the claim that farmed salmon constitutes a risk to health. As shown in Table II, the Norwegian Aftenposten mainly rejected the claim that eating farmed salmon is a risk to health. To pick some examples from this newspaper:

Norwegian farmed salmon is clearly below the limit value established by the EU, according to Anne-Katrine Lundebye Haldorsen from the National Institute of Nutrition and Seafood Research. (Aftenposten, 10 January 2004, our translation)

‘I will continue to eat salmon. It's a fantastic fish which can be used for so many things, and the price is good’, says René Hellerström who has just bought a half kilo fillet of salmon … (Aftenposten, 10 January 2004, our translation)

‘I've never felt as good as since I began to eat salmon’, says another customer. (Aftenposten, 10 January 2004, our translation)

Aftenposten also framed the issue as an economic question to a greater extent than did the other newspapers. Economic consequences, such as job losses and export reduction, were emphasized.

More diverging views were expressed in The Scotsman, The Guardian and Vancouver Sun. The English newspaper The Guardian, however, had just one article clearly claiming that salmon does not constitute any risk, and compared to the other newspapers it emphasized risk somewhat more than the other newspapers. This was also apparent when quoting ordinary people, as in the following example from the Sunday edition of The Observer:

‘There used to be fresh salmon on the table when I was a boy’, he told The Observer. ‘But that was a different fish to what we get now. I wouldn't feed my children with the stuff that comes out of these farms.’ (The Observer, 11 January 2004)

From Table II it can also be noted that almost half of the articles in The Guardian and The Vancouver Sun brought up other food scares and made connections to them, while this was only occasionally done in The Scotsman and not at all in Aftenposten. The Guardian and the Vancouver Sun, for example, interviewed environmental groups that criticized the farming of salmon. The Guardian in particular also brought up critical views on the industrialized farming culture, the scientists and the politicians:

But farmed salmon are fed ‘fish chow’, a feed which, among other things, contains ground-up fish. The same concerns which triggered the BSE scare – that beef by-products were being fed to cows – are in evidence here. (The Guardian, 9 January 2004)

Just as BSE research prompted an EU ban and shoppers' boycott of beef almost eight years ago, now public confidence was being rocked in the very foodstuff nutritionists have been telling us we must eat more of. (The Guardian, 11 January 2004)

Aftenposten, The Scotsman, and also the Vancouver Sun published some articles, mostly columns and debate articles, which were quite critical of the study reported in Science (the ‘no’category in Table II). These articles point out the weaknesses of the Science study, as the headlines indicate: ‘Norwegian and British researchers dismiss the risk for cancer’ (Aftenposten, 10 January 2004, translated to English), ‘Salmon is safe says US food expert’ (Scotland on Sunday, 11 January 2004), ‘Salmon scare was flawed and biased’ (The Scotsman, 16 January 2004), ‘Claims of unsafe fish run contrary to the facts, say scientists’ (The Scotsman, 16 January 2004), and ‘Salmon, both farmed and wild, is safe to eat’ (Vancouver Sun, 11 January 2004). A common standpoint taken in the articles referred to people defending the farmed salmon, such as company spokesmen and public authorities. It is claimed that the benefits of eating salmon are well known, or that the Science study has used an altogether too low allowed level of concentration of potentially cancer-causing toxins. The first citation below is from a spokesperson of the Norwegian public authorities, the National Institute of Nutrition and Seafood Research (NIFES), and the second from the British television chef Jamie Oliver who was interviewed about these new findings.

‘Our objections are tied to the way the measures of the risk estimations are made. It is controversial, and it is also not accepted by the American health authority’, she says. (Aftenposten, 10 January 2004, our translation)

‘I personally think that the more Omega-3 and fats that we can get the population to consume, the healthier we would be. The reality is that we do not eat enough fish.’ (The Scotsman, 10 January 2004)

5 Discussion

This brief analysis shows that the findings reported in Science and promoted by a press release and a web site did receive global broadsheet attention, albeit not in the newspapers included in the study from the Asian countries. The findings were in part reported differently and framed by different national dailies.

However, it is not possible to draw far-reaching and firm conclusions from this rather small body of empirical data. However, keeping in mind that the empirical foundation is small, we would like to discuss some tentative conclusions that can be drawn from this study.

The empirical material reveals that there are differences in the reporting, not least between the British and Norwegian newspapers' ways of reporting and framing the issue. Firstly, the study seems to be more reported in Britain than in Norway; this despite the more central role that the fishing industry plays in Norway than in Britain. Secondly, in Britain (as well as in Canada) the salmon study was to a somewhat greater extent framed as a health issue, whereas in Norway more emphasis was placed on economic consequences. Thirdly, the articles in the British press covered a broader range of standpoints – for, against or mixed standpoints – compared to the Norwegian paper, whose articles mostly emphasized that the warning about contaminants in salmon should not be taken seriously. Only one article clearly brought up negative health aspects.

How might this difference in framing between Norway and Britain be explained? We would like to propose some contextual factors that were in effect in this case. The higher level of media coverage in Britain may partly be due to the fact that the study presented in Science showed that ‘PCB, dioxin, toxaphene, and dieldrin concentrations were highest in farmed salmon from Scotland and the Faroe Islands’ (Hites et al. 2004: 227). In contrast to Norway, Britain has experienced a number of additional food scares (Seale 2002: 70f.), which has increased the media's willingness to pay attention to food warnings. In particular, the BSE catastrophe provided the media with a sounding board – i.e. a possibility to attach to an earlier kind of food scare and thereby create a story that resonates with the public – to convert issues to food scares. In the reporting on the salmon study, the British press emphasized health, made connections to other food scares and used many sources. The result was a more ambivalent view of the study, where different voices were arranged in a way that invited the reader to become more open to different interpretations of the issue. This can be contrasted with the Norwegian reporting, where it was emphasized that the study may result in an economic threat to Norwegian fishery, due to seafood being its second largest export industry (Nygård & Storstad 1998). Thus, whereas the issue in Britain was framed as an internal threat to its citizens (in terms of health risks), in Norway it was rather framed as an external threat to the domestic economy.

Furthermore, Britain has experienced a number of cases of domestic food contamination, whereas most of Norway's experiences of food contamination have come from imported food. 8 Whereas Britain has a food market of international character (with food from all over the world), Norway still has more domestically oriented food consumption and a national-symbolic meaning is often explicitly attached to the food produced in Norway. Until 1998, the domestic food market in Norway was protected from foreign competition through quantitative regulation. It was forbidden to import food products as long as the country was self-sufficient in these products. Nygård and Storstad (1998) emphasized that food in Norway – not least seafood – has emerged as a strong symbol of Norwegian identity. No doubt British food is also connected to national identity. It has, however, been tarnished by BSE and other domestic food contaminations.

Lastly, we would like to suggest that differences in political culture fuelled the framing processes in Britain and Norway. In Norway, environmental activism and sub-politics are relatively rare compared to Britain (Dryzek et al. 2003). Norway has been a driving force in international environmental negotiations and has developed a far-reaching environmental policy. Its former prime minister Gro Harlem Brundlandt was the chairman of the UN committee on Peace and Environment, whose final report (WCED 1987) proposes ‘sustainable development’ as the goal for the global community. Norway was also one of the first countries to introduce green taxes (which were introduced in 1991). Norwegian environmental groups are to a large extent included in the state sphere, receiving state subsidies. In Britain, environmental groups are more excluded from the policy process, which has meant that they have had to remain relatively autonomous from the state. They are more apt to create media-based campaigns that put environmental issues on the political agenda. This might explain why the British dailies allowed more voices to be heard, and included more perspectives and definitions in their reporting than the Norwegian newspaper did.

Thus, these four factors – the historical experience of food scares in Britain, the economic importance of the fishery industry in Norway, the strong symbolic meaning of domestically produced food in Norway and the more exclusive policy process in Britain – can explain the differences in the national framings in Norway and Britain. Obviously, as stressed above, these are tentative explanations and more systematic and deeper studies are needed to elaborate on the question of how scientific results on environmental risks are framed in the media in different countries. Nevertheless, this pilot study illustrates that the media is not prepared to be a megaphone for science, reporting scientific results in the way researchers may wish. Instead, the media brings in multiple voices from different social spheres and interests and is especially willing to pay attention to critical voices. Polarization is certainly common when the media reports on science (Väliverronen 2001).

5.1 Domestic media and global environmental problems

The media is a key institution in attracting public attention to environmental and risk issues (Beck 2000). A growing awareness of this has emerged within science, industry and among governmental agencies. A number of actors have developed strategies for obtaining media attention to disseminate their message to the wider community. In some cases, these strategies are developed to create public inattention or avert public outrage, as we can see in the case of the Bhopal catastrophe, where the company responsible systematically worked to spread its specific understanding of the cause and consequences of this chemical disaster (Rowell 1996; Beder 1997).

In other cases, strategies are developed to mobilize people and change their consumer behaviour. It is possible to see this in the work of governmental bodies set up to combat climate change, where information campaigns are undertaken to encourage people to reduce their private transportation. Science, however, seems to be far behind, with a quite naive view of the media as spokesmen for researchers and scientific findings (Allan 2002). The Swedish acrylamide ‘alert’ is an illustrative example where the research community was taken by surprise, and did not recognize how media and journalism work (Larsson 2005). The heated debate in the media led Löfstedt (2003: 431) to conclude that, ‘the way that the findings were portrayed in the media will only lead to greater public distrust of science as a whole and toxicologists in particular’.

As emphasized in the Introduction, the media is an active producer of meaning. Domestic media are embedded in national contexts and combined with the logic and interest of media; issues are framed in specific ways with the aim of telling stories and catching the attention of the readers. At the same time, it should be noted that the media does not merely adapt its message to the national culture. To a large extent the media is also a resource for the creation of national symbols and cultural meanings. That which in the public sphere is regarded as a national cultural symbol or an economic interest is the result of a dynamic and dialectical process involving a number of actors, among which media plays an important role (Dahlgren 1995; Lash & Urry 1995; Thompson 1995).

As shown in this case study, even if the scientists and the sponsor of the study had a clear ambition to publicize and disseminate their results and normative proposals to the wider society, the newspapers did not act as a medium for distributing the original messages. Instead, diverging understandings and framings developed, which to some extent can be explained by the different national contexts of the newspapers. What we see here is in accordance with theories of domestication of global news (Hjarvard 2001; Clausen 2004). Thus, science can serve as a trigger for global news reporting of scientific findings and its implications, but the original scientific meaning is transformed when moved from its scientific context to other social contexts. A number of studies have shown that news about risks follows a general journalistic logic, such as dramatization, sensationalism, polarization, personification and stereotyping (Anderson 1997; Friedman et al. 1999; Allan et al. 2000; Smith 2000; Weingart et al. 2000). This logic is always embedded in specific social frameworks, where national contexts are important when producing news (Clausen 2004). It is even the case that national contexts seem to be necessary ingredients in all sorts of meaning-making (Billig 1995). Even in a globalized world, the national context is ‘a particular logic among others that organize economic, political, technological and cultural territories and flows’ (Wiley 2004: 78).

However, the media should not be interpreted as being a uniform phenomenon, not even within a single country. There are different types of news media, such as television, popular science magazines, the tabloid press, and so on, which differ quite strongly in their coverage and framing of scientific issues. Furthermore, the reporting and framings provided by the media do not determine the interpretations made by the public. As remarked by Hall (1996) and Morley (1980) a dominant message in the media may be interpreted in oppositional ways by some groups of the audience, and many studies show that the public can be quite critical of the media's reporting on risks (Reilly 1999; Wilson 2000).

5.2 Striving for media attention

Food scares will probably be a growing phenomenon in the future, and scientists as well as industry and governmental agencies will use the media with the aim to distribute messages and influence consumer behaviour as well as public regulation. When science meets the media, the journalistic logic and national contexts will, however, continue to shape the original message of the scientists.

The case of contaminants in farmed salmon did not end with the study in Science and the media coverage of that study. Recently, the debate on the health risks has been fuelled by additional studies. The empirical material presented in Science in early 2004 has been further analysed by the researchers and presented in scientific journals during 2005 (Carlson & Hites 2005; Foran et al. 2005a,b). Through a cost-benefit analysis, they discuss the net benefits of eating salmon and find that the benefits of eating wild Pacific salmon outweigh those of eating farmed Atlantic salmon. In these articles, the proposal is repeated that salmon should be labelled so that it is possible to identify whether it is farmed or wild, but also to identify its country of origin. Thereby, the consumer could make informed choice with the aim of reducing exposure to dioxin-like compounds (Foran et al. 2005a: 554).

In an interview, one of the researchers recommends that consumers choose wild salmon from Chile and that farmed salmon from North America would be a better second choice than European farmed salmon (Lange 2005). The researchers also recommend that consumption of Norwegian salmon be restricted to three meals a year. Norwegian authorities have tried to counteract this recommendation by stating that Norwegian salmon poses no risk to health because the found levels of contaminants are below acceptable levels. Thus, on the basis of the findings, the researchers behind the salmon study try to influence consumer behaviours, and national governments and industry may respond to this as well. In this interaction, the media has a pivotal role, because – as emphasized in our article – it does not serve as a passive channel for distributing messages, but as an active player in the creation of news.

As emphasized in this article, there is a current and growing interest in how science tries to get media attention, how it succeeds or fails in this and how its original message is converted and framed through media logic and national contexts. This issue is not only relevant for media researchers and governmental agencies, but also for all environmental scientists who strive to spread their findings to the wider society.

Acknowledgements

We would like to thank Bert van Bavel, Professor in Environmental Chemistry at the MTM research centre, Örebro University, for his help in finding dioxin researchers willing to collect and code some of the national newspapers. We would also like to thank four anonymous reviewers for constructive comments on an earlier version of this paper. The Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas) funded the study.

Notes

1. International Symposium on Halogenated Environmental Organic Pollutants and Persistent Organic Pollutants (POP:s), Berlin, 6 – 10 September 2004.

2. A scientific report on increased levels of lead and cadmium in salmon led Russia, in January 2006, to impose a total import ban on Norwegian farmed fresh salmon. Since Russia is the largest export market for Norwegian seafood, this ban was immediately met by strong reactions from Norwegian agencies, which declared that it had comprehensive scientific documentation showing that Norwegian farmed salmon did not constitute any risk to human health. Norway started to negotiate with the Russian authorities in the beginning of January 2006, and in April of this year it succeeded in convincing Russia to gradually withdraw its ban.

3. The trust is located in Philadelphia and currently grants about 200 million USD a year to provide organizations and citizens with fact-based research and practical solutions to challenging issues. The objective of the trust is to advance policy solutions on important issues facing the USA public, inform this public on key issues through research and polling; and to support civic life in the USA (see www.pewtrust.com).

4. Personal communication with the main author of the article, Professor Ron Hites (6 December 2004). He was contacted by us via e-mail and kindly answered a few questions about the media involvement.

5. A limitation of using Net editions is that the strategic placing of an article, such as front page headlines and other attention-seeking means will be missed in the analysis.

6. The Santiago Times of Chile is written in English and was coded by us.

7. In these countries, we complemented our data collection by investigating some other national newspapers. In the second, smaller Dutch paper (NRC Handelsblad) that we looked into we found two articles. Four papers were checked in Japan and three in China, and no articles were found in any of them.

8. In Britain, a number of contaminations and diseases in animals and food products have been found, for example listeria and salmonella in egg (Seale 2002). Norway, on the other hand, is one of the countries with the lowest levels of food-related contamination. In the case of salmonella in eggs, 93% of the known cases, from 1983 to 96, were found in imported eggs (Nygård & Storstad 1998).