Communicating environmental science beyond academia: Stylistic patterns of newsworthiness in popular science journalism

Approach

As stated in the ‘Introduction’ section, Bednarek and Caple’s (2012, 2014), newsworthiness framework (discursive news values analysis) is the starting point in this project. The specific aim of this study is to identify the main strategies of making science ‘discursively constructed as newsworthy’ (Bednarek and Caple, 2012: 44), with special attention paid to lexico-grammatical features that can be identified as having been used to present science-related information to appear as interesting, attractive or relevant. However, we selectively draw from their typology of news values by focusing on novelty and superlativeness, timeliness, negativity/positivity and, to some extent, prominence. This section defines these news values and links them to their typical linguistic realizations in general news based on Bednarek and Caple (2012, 2014; cf. Bednarek, 2015). This is followed by an explanation of other discursive strategies instrumental for generating engagement in science journalism, and of their operationalization. These include rationalization and speculation, direct address and conversationalization.

To attract readers in the media-saturated reality of science journalism, the news value of novelty, including not only newness but also unexpectedness, and described as ‘deviance/unusuality/rarity/surprise’ (Bednarek and Caple, 2012: 43), is notable. Bednarek and Caple (2014) enumerate the following linguistic realizations of this value: evaluations of unexpectedness through modifiers, comparisons that indicate rarity; references to surprise as an emotion displayed by social actors involved in the event; and references to unusual happenings and extraordinary events. In our study of environmental science articles, novelty is expected to be mainly realized with respect to scientific results and discoveries that apparently exceeded scientists’ expectations. Novelty is also likely to coexist with superlativeness, that is the discursive ‘maximizing or intensifying of particular aspects of an event’ (Bednarek and Caple, 2012: 44, 2015), which can be reflected in the usages of ‘large’ quantifiers, adjectival intensifiers, intensified lexis and references to growth or escalation (also through repetitions, comparisons and figurative expressions). We hypothesize that science journalism, to retain credibility, may have a preference for constructing superlativeness through numerical data, comparison and the narrative organization of the report.

Timeliness, according to Bednarek and Caple (2014), is built by establishing a close relation between the event and the publication through indications of recency, explicit and implicit time references, or references to an ongoing process (e.g. through verb tense and aspect). It can be assumed that considering the need to make readers pay attention to less popular scientific domains and to technical information, some of the science-oriented coverage will be represented as ‘the latest developments’ that either endorse or contradict previous findings, while the outlet’s build-up of recent information in one thematic thread will be represented as ‘tracing current scientific progress’. Some of the timely news items may additionally be marked for their possible impact (cf. the notion of ‘breakthrough’). Although Bednarek and Caple (2014, cf. Bednarek, 2015) see realizations of the news value of impact in terms of (speakers’) evaluations of significance or references to possible (relevant) consequences, scientific journalism may well resort to other patterns of projecting relevance and impact through conditionals, modality or narrative.

Negative aspects of events, which tend to effectively build newsworthiness, may be realized through negative evaluative modifiers, reference to negative emotion and attitude, or lexical items that refer to undesirable states and actions (Bednarek, 2015; Bednarek and Caple, 2012, 2014). Given the current environmental crises, we may expect a predominance of negative information in environmental science articles, which might be responsible for a sense of fatigue or alienation with respect to this thematic domain. However, even if the overall findings are worrisome, the nature of scientific coverage is such that new research reports may give a sense of security by offering knowledge, explanation, warning or remedy, which can be evaluated as positive. The oscillation between negativity and positivity in science coverage is an important stylistic pattern that characterizes science journalism (cf. Molek-Kozakowska, 2016), which is why in this project the news value of negativity will be analysed vis-a-vis its counterpart – positivity.

The news value of prominence (Bednarek and Caple, 2012), also known as eliteness (Bednarek and Caple, 2014; cf. Bednarek, 2015), should be briefly mentioned in the context of science coverage. Science, and thus academia, is a privileged and prestigious domain of activity. The discursive implementation of eliteness (status or acclaim) can be achieved through various institutional status markers, including labels, titles and recognized names, as well as evaluations of importance and descriptions of achievement. Even though these are often used indirectly in academic discourse, popular science journalism’s credibility rests on representing information as emanating from elites, as reliable and thus worthy of notice. We thus hypothesize that eliteness will be a prominent news value in our data.

This takes us to the next level of our approach (beyond Bednarek and Caple’s framework), namely to other stylistic aspects that seem to be particularly significant in the case of science journalism: rationalization and conversationalization/direct address.

In order not to alienate readers who are interested in (and want to be knowledgeable about) science-related issues, science journalism needs to offer descriptions, explanations and details collected from authoritative sources, as well as to spell out cause–effect relations, exemplify general rules with particular examples, indicate problem–solution sequences and argue for/against various alternatives (Perez-Llantada, 2012). These logical and rhetorical strategies characteristic of expository discourses, including academic texts, can be regarded as rationalization strategies (Fahnestock, 2005; Moirand, 2003). The claim made here is that rationalization (a substitute for exposition) is a feature of popular science journalism in which science-related issues are to be presented as relevant, attractive and entertaining to non-specialists without compromising the outlet’s credibility. Rationalization is also achieved through intertextuality and attribution of information, as well as by illustrating the applicability of science results through proximization. However, since implications and consequences of various scientific discoveries are not yet known, science coverage is likely to involve various degrees of epistemic modality and commitment to truth, thus mitigating the uncertainty inherent in scientific research (cf. Molek-Kozakowska, 2015). In this way, appeal to the public can be enhanced by obscuring the speculatory nature of some claims.

While rationalization strategies and tentativeness are also seen in academic discourse, conversational features of discourse and direct audience address are typically avoided in academia (Perez-Llantada, 2012). This, however, is what makes science coverage much more ‘dialogical’ or ‘conversational’ than most science materials, even science textbooks (cf. Halliday and Martin, 1993). Science articles also rely on quoting the words of scholars and experts; however, such a ‘quote’, or mostly free indirect discourse, is often sourced from spoken interaction (interview), not publication, and includes expressions of beliefs, hopes or feelings that would be absent in an academic report. This personal stance of the expert is mirrored by the personal appeal to the reader, with the linguistic realizations of direct address (personal pronouns, imperative structures), combined with easy-to-follow formats and human-centred story lines. Such stylistic features could be appreciated when effective popularization of science beyond academia is at stake, but they could also impress as signs of infotainment that displaces responsible science communication.

Data and sampling

Assuming that science journalism constructs science news items as newsworthy through its style, we have selected only articles that appear in the ‘most-read’ ranking list on the NS’ homepage, rather than a random sample, in order to ensure that the textual material studied here actually generated readers’ engagement. Out of over 200 ‘most-read’ articles collected between October 2013 and December 2014, we chose the ones that related to a thematic category that seems to have caused some readers’ fatigue (Hansen, 2011) – the environment (i.e. climate change, carbon emissions, renewable energy, air pollution, depletion of resources, restoring biodiversity) – because arguably, it takes some effort to attract readers to such an extensively covered domain. ³ It seems justified that (in order to study instances of real and current stylistic patterns responsible for generating engagement) one should use a sample of articles which have been frequently chosen by broad audiences. This informed our choice of the ‘most-read’ articles from a free website of one of the most popular and established international English-language middle market science popularization outlets. The sampling yielded 38 articles (of overall word size 22,150), which are listed in Appendix 1 and will be referred to in the subsequent analysis with the numbers they are assigned in that list.

The sampled texts have been subjected to multiple close readings and analysed using an open-coding procedure, with special attention given to the format of the whole article, the phrasing of the headline and lead, the recurring linguistic resources used for representing the issue, evaluating its significance, explaining the scientific procedure, building argument and referring to sources. Each pattern has been identified by comparing linguistic features across the sample. The ‘pervasive’ patterns were subsequently labelled and grouped according to how they realized selected news values and other identifiable strategies operationalized earlier. Representative examples have been selected to illustrate both the consistency and the variability of stylistic patterning in the sample. Examples are discussed in the following with explanations of how the accumulation of certain infotaining stylistic devices works to reproduce the ideology of newsworthiness and impinges on the NS readers’ understanding of environmental science.

The discursive construction of news values

Rationalization and mitigation of speculation

So far we have defined rationalization as logical and rhetorical strategies reminiscent of expository discourses, including academic texts (which are by no means rhetoric-free; cf. Fahnestock, 2005; Perez-Llantada, 2012). Rationalization is the realization of rhetorical logos, as opposed to the evaluative (ethos) and emotive (pathos) features realizing news values, as discussed earlier. In the NS sample, 35 articles involve rationalization instantiated via various stylistic patterns. For example, as many as 30 articles introduce numbers, cite statistics, calculate ratios, trace increases/decreases, or provide quantitative data in specific measurements (e.g. of temperature, wind velocity, electric output, surface, volume, productivity, costs and savings) to give factual evidence to claims made in the heading. Sometimes, however, the use of numerals is less precise and more likely to be rhetorically charged: ‘Earth’s tectonic plates have doubled their speed’ (26), ‘The Arctic is warming twice as fast as the rest of the planet’ (29), ‘5 reasons to worry about a quagga mussel invasion’ (31). Jargon (e.g. ‘polar vortex’, ‘extratropical storm’, ‘demographic momentum’, ‘extreme weather events’, ‘acidified oceans’, ‘hiatus’), technical specifications and concrete data are used to strengthen the credibility of the reports.

Another way to rationalize coverage is to articulate logical relations between propositions. The NS sample as a whole (38 articles) includes 41 uses of ‘because’, 16 of ‘due to’, 14 of ‘since’ and 8 of ‘as’ to indicate reasons/causes, as well as 14 cases of the use of ‘[this] mean[s]’ to explain a consequence. Note, for example, the following: ‘humans may have played a role, because they have been farming the area for 2500 years’ (20), ‘the remaining difference between models and observations is probably due to increased heat uptake by the oceans’ (6), ‘it’s a machine with a dual role, as excess humidity can lead to problems like leaf mould’ (7), ‘the climate is highly sensitive to carbon dioxide, according to several new studies, which means that our greenhouse gas emissions will lead to strong warming’ (14), ‘using waste heat from a coal plant to power the heat pump can mean 80 per cent of the energy put into the slush can be retrieved’ (13). Additionally, time and conditional clauses, if–then clause complexes, as well as infinitives of purpose are found in all articles. Some articles are indeed designed to explain a phenomenon at issue from a given perspective (6, 8, 16, 19, 23, 29, 34, 36, 38). As such characteristics of explanatory discourse are indeed a primary property of science communication as a register (Hyland, 2000), there is no wonder that they are also pervasive within NS’ stylistic repertoire.

Yet another way of rationalizing the coverage is the appeal to expertise: ⁵ quotes from experts (scientists, research and development (R&D) managers, developers, project leaders, environmentalists, officials) and institutional bodies (academic, non-governmental organizations (NGOs), administrative) are frequent and pervasive, although most citations come from just the authors of the study reported on (sometimes hyperlinked to the original journal article or press release), and usually not from independent authoritative sources (except for 27, 29, 36, where the plurality of sources confirming the result is touted). A typical example of this pattern would be in the form of inserted and attributed words of the main author: ‘“We’re developing a glass that you can pump like a liquid,” says Halotechnics CEO Justin Raade. “It has a viscosity order of magnitude below traditional window glass, and at 400 °C is about the viscosity of honey”’ (13). References to experts are missing only in four feature articles (10, 11, 19, 38), while the interview (1) is conducted with an expert, so all arguments are to be taken as elicited from an informed source.

The sample is also pervaded by the use of scientific ‘metalanguage’ referring to how research is done and reported. These include the following nouns: ‘findings’ (n = 14), ‘results’ (n = 15), ‘solutions’ (n = 18), ‘technologies’ (n = 20) and ‘projects’ (n = 30), as well as a number of verb phrases denoting researchers’ activities. Typical of reporting in free indirect discourse, there are dozens of instances of verbs of thinking and saying. In addition, there are some verbs that connote scientific work: ‘demonstrate’ (n = 3), ‘understand’ (n = 4), ‘present’ (n = 4), ‘develop’ (n = 4), ‘cope’ (n = 5), ‘report’ (n = 8), ‘calculate’ (n = 11), ‘prove’ (n = 15), ‘model’ (n = 20), ‘show’ (n = 21), ‘find’ (n = 29). On the other hand, there are even higher numbers of verbs that refer to what environmental researchers do not know for sure and have to still verify: ‘suggest’ (n = 20), ‘predict’ (n = 22), ‘expect’ (n = 23), ‘seem’ (n = 24), ‘estimate’ (n = 26). ⁶ With a larger sample, it could be compared if the frequency/keyness of references to scientists’ work differs from academic discourse in which data are sometimes shown as emergent and conclusions as self-evident. Nevertheless, the stylistic pattern of using science-related metalanguage contributes to a rationalization strategy that projects the world of environmental science as one in which new findings and solutions are to be constantly expected, which gives the readers a reason to follow NS’ coverage, despite the fatigue.

Uncertainty is an indispensable aspect of doing science, but it can be mitigated or even backgrounded with various stylistic devices (Molek-Kozakowska, 2015; Perez-Llantada, 2012). Since many of the sampled articles focus on the future consequences of present activities (particularly of climate change), scientific knowledge is used therein to underpin various ‘versions’ of reality and to promote specific solutions. As speculative discussions might be more engaging than mere reports on the findings (Hansen, 2016), we can observe high numbers of hedges (probably, likely, certain/ly) and expressions featuring modal verbs that signal various degrees of certainty, ability, probability and expectation (e.g. ‘will’, n = 175; ‘can’, n = 88; ‘could’, n = 56; ‘need’, n = 42; ‘may’, n = 28; ‘might’, n = 10; ‘should’/‘ought to’, n = 6). It can be observed that the stylistic patterns realizing rationalization obscure the fact that some of the coverage is a projection of an envisioned version of the future rather than a report on concrete facts.

Conversationalization and direct address

Apart from the consideration of stylistic patterns of newsworthiness and rationalization versus speculation in science journalism, one should also note the interactional generic design of this kind of journalism. If the sample collected here is treated as representative of popularity-driven environmental coverage in science journalism, one should note that among the ‘most-read’ articles one is an interview (1), one a ‘staged’ conversation (14), one a personal account of diving in the ocean (22), one a review and critique of three publications (38), and all, save three features, extensively rely on the voices of various experts and officials. ⁷ What is notable, thus (especially when set beside academic discourse), is the pervasiveness of stylistic characteristics of spoken language/conversational register (Biber and Conrad, 2009), such as contractions/elisions, clauses with initial ‘and’ or ‘but’, sentence fragments, personal pronouns and deictics, emphatic phrases that correspond to intonation units, interrogatives or colloquial idiomatic collocations. The following could be regarded as representative examples of strategically deployed ‘conversationalization’: ‘If you thought shale gas was a nightmare, you ain’t seen nothing yet’ (10), ‘But while they are great for electric sports cars, the batteries aren’t cheap and aren’t flexible enough to cope with wildly different loads for power grid storage’ (13), ‘But that doesn’t mean that the project won’t damage the reef. Far from it’ (19), ‘“It’s looking to us that it’s probably going to be the last one [hiatus] that we’ll see in the foreseeable future,” says England’ (27), ‘What would it take to defuse this population time bomb?’ (34), ‘The immediate cause of this craziness was super typhoon Nuri’ (36).

Direct appeal is manifested in NS through imperative constructions or the use of the pronoun ‘you’ to address readers: ‘take the escalator down to platform level at any deep London Underground station and a wave of warm air wafts over you’ (5), ‘remove these wobbles, and you are left with an additional signal’ (6), ‘urban warehouses, derelict buildings and high-rises are the last places you’d expect to find the seeds of a green revolution’ (7), ‘A compelling candidate for renewable energy in a post-apocalyptic future? Tick that box as well’ (11). In fact, many leads, captions or transitional paragraphs feature similar informal direct appeals that break the stylistic monotony of expository, descriptive or narrative discourse to position the reader as a participant (not only a receiver) within the discourse frame. Needless to say, this stylistic pattern is very hard to find in serious academic prose (on various clines of detachment and impersonalization, cf. Perez-Llantada, 2012: 62; cf. also Biber, 2006).

One case deserves specific attention, namely, article 14, which is stylistically manipulated to resemble a conversation between an expert on global warming and an average lay person. The unusual format is introduced with a justification: ‘It’s a complicated story. So New Scientist has broken it down.’ The questions asked by the non-specialist tend to be very colloquial: ‘What is this climate sensitivity business about anyway?’, ‘Thanks, that’s very cheering. Why has there been so much uncertainty?’, ‘But that doesn’t mean sensitivity is higher, does it?’, while the answers, even though featuring precise data, details and fine terminological distinctions, tend to be short and to the point. For example, the answer to the last of the above questions is ‘It means we can no longer dismiss models with high sensitivity on the basis that their projections don’t match reality. On the contrary, the most sensitive models may be the most realistic’. Alternatively, in this conversation, there are definitions, comparisons, examples, enumerations, transitions and analogies (e.g. ‘If you kicked your best friend in the teeth, how would they react? […] Climate sensitivity is a measure of how strongly the planet will react to the kicking we are giving it’) that are to facilitate the understanding of contemporary dynamics of climate change. Article 31, in a similar vein, presents its findings as a list of five reasons to worry about quagga mussle ‘invasion’ using colloquial idioms and ‘an inner-city ghetto’ analogy: ‘It keeps bad company: The quagga mussel’s faeces provides food for other invasive organisms’, ‘It shows no mercy: Quagga mussels literally suffocate other mussel species’. ⁸