Responsible Use of Language in Scientific Writing and Science Communication

Scientists are increasingly advised to use marketing strategies and strong metaphors or dominant news frames to get their message across (Bubela et al. 2009, Larson 2011, Kuchner 2012, Nelder 2013), because there is growing competition for funding, public attention, and the ability to influence decisionmaking (e.g., Fischer et al. 2012). We think that this development creates problems to the extent that it invites scientists to undermine their objectivity by making assertions that are not based entirely on data. Instead, when scientists make public statements that build on their legitimacy as neutral and knowledgeable experts, their appropriate role is to act as “honest brokers of policy alternatives” (Pielke 2007). They can contribute to controversial debates and express their personal opinion, but, to the extent that this is possible, they should explain the degree of subjectivity in their opinion and provide alternative ways to interpret the evidence.

Decisionmaking in our time depends on reliable scientific evidence, so neglecting or misinterpreting such evidence may impair decisionmaking. However, knowledge is often uncertain or incomplete. In these instances, incipient knowledge can be used to substantiate multiple conflicting views on an issue, thereby contributing to policy impasses (Sarewitz 2004, Kahan et al. 2012). As biologists, we are increasingly confronted with such situations, especially in an era of biotechnological innovation and global environmental change in which much knowledge that once seemed solid is fundamentally changing. This puts a great burden of responsibility on science communication, so media coverage of scientific findings is rightly expected to follow strict guidelines of good practice and ethical behavior. Trustworthy science journalists focus on peer-reviewed literature, carefully fact check articles, and report not only primary research results but also critical assessments by peers in the research field. Scientists are equally concerned that the interpretations and limitations of their science are presented in a correct and evidence-based way.

In this article, however, we highlight an aspect of responsible reporting of scientific evidence that has received less attention. It appears that, although scientists and science journalists carefully evaluate scientific results and ways to interpret them, they more freely choose metaphors to communicate to their audience (see Nerlich et al. 2009). In this manner, they may communicate interpretations and weightings of scientific results that are not supported by data. Although there is no such thing as value-free communication (Weber and Word 2001), we argue that scientists and journalists should reflect carefully on their communication choices to minimize confusion and contested interpretations of uncertain knowledge.

The performative quality of language at the research frontier

We focus on metaphors here because they are linguistic choices that have been shown to shape the thinking of scientists and, therefore, the direction of research fields (Keller 2002, Brown 2003, Larson 2011). They are, in this sense, performative—meaning that they lead to actual outcomes in the world. Accordingly, a narrow choice of metaphorical tools can restrict the diversity of research approaches and creative thinking, especially at emerging research frontiers (Larson 2009). This is of particular concern when research outputs can have huge consequences for society.

For instance, scientific and popular articles about research in synthetic biology often build on metaphors taken from the worlds of machines, computers, or consumerism (Cserer and Seiringer 2009, Keller 2009, Hellsten and Nerlich 2011). The titles of recent articles in New Scientist, for example, proclaim “Evolution machine: Genetic engineering on fast forward,” “Genetic code 2.0: Life gets a new operating system,” “Synthetic biology's Ikea,” and “Synthetic ‘upgrade’ for fruit fly's DNA.” These titles often encapsulate the core metaphor of an article, but such metaphors are usually prevalent throughout a text. Elsewhere, for example, individuals affiliated with syntheticbiology.org, which was initiated by students, faculty, and staff from Massachusetts Institute of Technology and Harvard, seek to “specify and populate a set of standard parts that have well-defined performance characteristics and can be used (and re-used) to build biological systems,” “reverse engineer and re-design pre-existing biological parts and devices in order to expand the set of functions that we can access and program” (http://syntheticbiology.org/FAQ.html), and “construct… a genetic code with an enlarged alphabet of base pairs” (http://syntheticbiology.org/Who_we_are.html). The use, here, of mechanistic language implies an ease of manipulation and a certainty of understanding that misrepresents the knowledge of these biological systems (Boudry and Pigliucci 2013), as well as what synthetic biologists actually do in their research and how they think about the associated risks (Cserer and Seiringer 2009, Marris C and Rose 2012). Furthermore, the implied ability to control biological systems prioritizes an instrumental way of valuing life, in contrast to an intrinsic or even sacred one (Deplazes-Zemp 2012). Given the holistic and dynamic nature of the genome and the cell (Boudry and Pigliucci 2013), we recommend following ecology and evolutionary biology and, instead, using metaphors such as those of tangled genome (compare tangled bank in Darwin 1869) or tinkering instead of design (Jacob 1977), because they more fully acknowledge organismal complexity (Proctor and Larson 2005). Using such alternative metaphors could change the way research is done in synthetic biology and how it is taken up by industry and the general public.

There are many other prominent metaphors in biology that have gained disproportionate importance and that shape both research and social outcomes (see Larson 2011 for a review). The metaphor of ecosystem services, for example, currently dominates ecological research on human–
environment relationships, thereby reinforcing a perspective from economic production that emphasizes the direct benefits that ecosystems provide to humans while neglecting alternative conceptions that are focused on stewardship and moral duty, long-term sustainability, and the cultural values of ecosystems (Raymond et al. 2013). Similar examples can be given for other biological research fields, such as health (e.g., Sontag 1989) or neuroscience (e.g., Slaney and Maraun 2005). The problem here is not so much that a metaphor is wrong but that it is misleading: It encourages the interpretation of a partial view as the whole truth or the attribution of too much importance to the view provided by one metaphor as opposed to the different insights provided by a plurality of them.

Such a tunneling of our intellectual engagement with an emerging topic is problematic for a number of reasons. It weakens creativity and threatens our ability to foresee unknown problems and risks (“unknown unknowns”), and it also excludes members of society who do not share the implicit values of the dominant narrative, which can lead to public opposition and controversy (Marris C and Rose 2012). Excluding people in this way undermines responsible scientific conduct, which requires scientists to ensure that their research contributes to democratic deliberations that serve the common good (Kitcher 2004).

Inclusive and effective communication with the public

Effective communication with the public is a rapidly growing concern for scientific fields ranging from biotechnology to ecology (Beardsley 2006, Bubela et al. 2009, Groffman et al. 2010, Nerlich et al. 2010), and metaphors are often used to help readers connect to scientific results. Although we do not question the usefulness of metaphors in science communication, we turn to the example of invasive species to highlight how value-laden messaging can also hinder effective communication and action.

Metaphors of war are often used when reporting about nonnative invasive species (Larson 2005, 2011), as well as in other environmental sciences (Hamblin 2013). Popular and scientific articles abound that include terms such as enemy, eradication, fight, and war or that liken biological invasions to bioterrorism. However, there is a growing consensus in the literature about invasive species that such language hinders effective engagement with diverse stakeholders and nature conservation practitioners (Young and Larson 2011), especially in situations in which the scientific information is ambiguous, uncertain, or incomplete (Larson et al. 2013). Indeed, researchers in the field think about the problem in diverse ways, which leads to differing assessments of the negative effects of nonnative species and the urgency of the problem (Humair et al. 2014). In particular, the positive effects of nonnative species are increasingly acknowledged (e.g., Schlaepfer et al. 2011), so popular scientific articles portray them as beneficial and encourage readers to “embrace invasives” (Vince 2011) or “welcome weeds… [that] could save the Earth” (Hamilton 2011). Trapped within an either–or duality, these authors overstate this alternative view so that it becomes as misleading as the view that all nonnative species are bad. The extreme of battling invasives may encourage managers to take rigorous action under all circumstances without considering alternatives and balancing implementation costs, opportunity costs, and benefits (Larson 2005), just as the other extreme of embracing invasives may encourage them to ignore the problem rather than to seek solutions. Using such hyperbole to report on an issue that is actually quite nuanced can lead to confusion and may hinder the responsible and effective use of scientific evidence in decisionmaking and the media. In contrast, neutral and balanced language will help ensure that policymakers and scientific advisers are not in the difficult situation of needing to overcome stereotypes when seeking solutions to the problem that adequately address its sociopolitical and biological complexities.

The metaphors in scientific texts should be used to help readers understand scientific findings, not to convince them without explaining the reasons. The International Union for Conservation of Nature (IUCN) has, for instance, recently listed the giant salvinia (Salvinia molesta) as one of the world's 100 worst invasive species and reported this in Nature in a letter entitled “Alien species: Monster fern makes IUCN invader list” (Courchamp 2013). We consider this choice of words to be undesirable, because it merely expresses a value judgment of the authors (i.e., that the species is like a monster because it is bad) rather than illustrating the science. The metaphor devalues this plant species in its entirety (like a monster that is always bad) rather than specifying which aspects of its behavior are problematic. In short, the monster metaphor implies an absolute value judgment, which may undermine the public's perception of science as a critical and balanced form of inquiry and, ultimately, their trust in it. Given their societal role, scientists should make value statements such as this one in a more explicit way—for example, by calling the fern a problem rather than a monster. Or they might use a metaphor to explain the underlying science rather than to express an opinion—for example, by calling this aquatic fern that rapidly proliferates in water polluted by nutrients a gourmand fern.

Strong messaging can undermine the credibility of ­science—in particular, when contradictory facts become available and, therefore, scientific uncertainties become visible to a broader public. For instance, ecologists adopted apocalyptic warnings in the Waldsterben (dying forest) debate in the 1970s and 1980s in Europe, and this choice still undermines their expert status several decades later (Horeis 2009). It might seem that engagement with uncertainties will reduce the capacity for science to contribute to decisionmaking, and there is undoubtedly political demand for clear and fixed answers. In the long run, however, ­scientists and science journalists need to reiterate that science is not so much a collection of established facts as a process of inquiry; therefore, especially in the case of contested and complex societal issues, responsible science communication entails stating the facts in neutral language and acknowledging areas of uncertainty (Pielke 2007). Conversely, a major concern for scientists is that journalists tend to portray a scientifically undisputed case—such as the anthropogenic causes of climate change—as an unresolved issue by giving unjustified space to extremists’ views (e.g., Boykoff and Boykoff 2004). Using loaded language may invite such media coverage by giving the impression that the inclusion of an opposing view is needed for a well-balanced story.

Another problem with promotional communication by scientists is that they are seldom trained in how to appropriately or effectively communicate. Accordingly, their choices may be counterproductive to their intended outcomes. Invasive species and climate change scientists still tend to resort to fear-based language about negative impacts of nonnative species or climate disasters, even though there is evidence that such language is ineffective or even counterproductive (Nerlich et al. 2010, Larson 2011, Kahan et al. 2012). To change environmental attitudes, a focus on solutions and cultural values—as opposed to danger and loss—is often more effective. As another example, Pigliucci and Boudry (2011) argued that the use of machine-information metaphors in biology, such as was discussed above for synthetic biology, can play into the hand of proponents of intelligent design, because they “bolster design-like misconceptions about living systems” (p. 454) that are pivotal to the intelligent design argumentation of creationists.

Engage early with audiences

There is now widespread recognition among communication scholars that the limiting step for greater scientific understanding is not simply more science communication to inform the public (Bubela et al. 2009, Kahan 2010, Kahan et al. 2012), as many scientists tend to assume (Peters 2013). This view is known as the deficit model of the public understanding of science, and it has been discounted for numerous reasons, not least that knowledge is only one element of how people come to decisions. This realization does not devalue effective science communication but, instead, places it in a more appropriate context, as part of a broader suite of engagement with the public about the production and use of scientific knowledge. Evolving models of science communication incorporate citizens much earlier in the process, even insofar as designing a research program itself (Bubela et al. 2009, Kueffer et al. 2012). Interactions between scientists and stakeholders can help improve our understanding of a problem and help determine how it should be framed, formulated, and integrated into the sociocultural environment. Scientists should therefore engage earlier with diverse audiences through participatory processes to facilitate the effective use of metaphors, whether in the context of basic or applied research (see Larson 2011), and to reduce the risk of pervasive outcomes in science communication such as those mentioned above.

Many terms used in ecology and environmental studies, such as disturbance, ecosystem health, restoration, and wildlife, imply a preferred state of nature and are, therefore, implicitly normative (Carolan 2006, Larson 2011). Recently, new terms such as novel ecosystem and Anthropocene have gained prominence in biodiversity conservation to denote the dominant role of humans in nature. These new terms shift the normative stance within the field and, indeed, divide experts and stakeholders who feel that they are at opposing ends of a contested issue: Is the pervasive human presence in nature bad or good? Some conservationists fear that these terms could endorse human domination of the planet (Crist 2013) and could exacerbate the shifting cognitive baseline whereby humans tend to become accustomed to new and often degraded ecosystems and thus forget the nature of the past (Caro et al. 2012). Conversely, others propose that associating widespread anthropogenic ecosystems with positive values will be the only way to ensure that people will still care for nature and the ecosystems around them (Marris E 2011). But this is not really about choosing between two options: It is, instead, a complex and ambiguous question that touches on fundamental worldviews. It should, therefore, be seen as an issue requiring the engagement of the whole conservation community (Kueffer and Kaiser-Bunbury 2014). Such questions about how we put our scientific concerns into a cultural context cannot, therefore, be left to biologists alone but must be elaborated with the assistance of partners from the social sciences and humanities, including ethicists (Rose et al. 2012, Sörlin 2012), as well as the engaged public (Yung et al. 2013).

Toward responsible and balanced science communication

We have argued that scientists should—with the help of journalists and social scientists—find ways to communicate (often uncertain) results in an engaging way that explicitly attends to its metaphors and values. How can the situation be improved in practice? We certainly do not suggest that scientists can or should abandon metaphorical and engaging language in science communication. Instead, we propose that they adopt criteria to help distinguish problematic from fruitful ways of using such language (box 1), although this distinction will be sensitive to the scientific and societal context, must evolve over time, and will never be perfect. Therefore, we also propose procedural guidelines and training for scientists, science journalists, and the editors of scientific journals to work collaboratively to report scientific results using frames and metaphors that speak to diverse experts and stakeholders (box 2). Scientists and journalists must continuously strive to get both the facts and the story straight, because it is as unscientific to communicate with inappropriate language as it is to ­present bad data.

We thank Brigitte Nerlich, Erle Ellis, Franziska Humair, Juanita Schlaepfer-Miller, and several anonymous reviewers for comments and suggestions on earlier versions of this article.

References cited