2. Some neglected challenges in building trust in science

When members of the public encounter a scientist, they often encounter an individual person and hear their testimony, which can induce them to take up certain beliefs if they trust this person. But the reason why they should trust this person as a scientist is not that they trust this particular person, but rather that this scientist is a member of the system of scientific knowledge production. This system, however, consists of highly divided labor (e.g., Gerken Reference Gerken2022, Reference Gerken2023; Hardwig Reference Hardwig1985, Reference Hardwig1991; Longino Reference Longino1990). As Gerken puts it metaphorically: “Scientists no longer stand on the shoulders of giants as much as they stand within an edifice built by a myriad of their predecessors and contemporary peers” (2022: 28). One indication of this phenomenon is the high number of co-authors in many fields (ibid.). These collaborators rely on what Gerken (Reference Gerken2022) calls “intra-scientific testimony” (between scientists who directly collaborate) and “inter-scientific testimony” (without direct collaboration, but still crucial for sharing data, analyses, etc. through publications). Through such testimony, different insights and results can be brought together; where appropriate, it can also be used for integrating other forms of knowledge, e.g., indigenous knowledge.Footnote ⁵ At least as important as direct collaboration is the role of scientists in double-checking each other’s work: as peer-reviewers, discussants, respondents, initiators of replication studies, etc. While these mechanisms are not perfect, they provide a form of quality control that serves to weed out error and fraud.Footnote ⁶

This “social view of science” sees it as a community-based endeavor with ethical and epistemic norms that its members need to follow (e.g., Longino Reference Longino1990: chap. 4). Feminist philosophers of science have added an important dimension to this view: to fulfill its function as well as possible, the scientific community needs to fight against biases, e.g., along lines of gender or race, because, as Rolin puts it, “inclusive and responsive dialogue based on shared standards of argumentation is a background condition that makes it more reasonable to trust the community’s ability to detect and eliminate error and bias than otherwise” (Rolin Reference Rolin2002: 101, see also Rolin Reference Rolin2017). The public can trust science – as an institution – insofar as the members of this community, together, come up with results that have been checked and vetted, following the right social norms (e.g., responsive dialog) while attempting to minimize bias (e.g., Oreskes et al. Reference Oreskes2019). I take this rough sketch to be uncontroversial on the descriptive level. Insofar as non-scientists are expected to “trust science,” they therefore need to place trust in this system as a whole.

It might be objected that one cannot trust a system, because trust is a moral relationship, the violation of which induces blameworthiness (Baier Reference Baier1986). It is not immediately clear how such a moral relation could hold towards a system of divided labor. But one can make sense of this claim if one takes into account that the role of a scientist comes with epistemic responsibilities that have a moral dimension. As Rolin puts it:

The self-interest or the personal epistemic goals of scientists may deviate from what they need to do for scientific cooperation to go well, epistemically speaking; therefore, an additional moral motive is needed (ibid., p. 475, see also Frost-Arnold Reference Frost-Arnold2013). One can spell out this additional motive in terms of commitment to the institution of science and its norms (Fricker Reference Fricker2002) or commitment to specific epistemic values such as competence, conscientiousness, and honesty (e.g., Margitay Reference Margitay2021).

In this way, trust in science can be understood as trust in a social system of divided labor. Of course, this does not mean that one should neglect formal mechanisms of checks and balances. This has been emphasized in particular by those who see the scientific system more in terms of reliability than in terms of trust. Scientists can base their beliefs on the testimony of other scientists because the latter are embedded in institutions in which false or misleading testimony is discovered and sanctioned (e.g., Fricker Reference Fricker2002: 383). Such an account relies on strictly enforced standards, not only when it comes to the responsibilities of scientists, but also concerning structural questions, such as preventing problematic forms of commercialization (on the latter see, e.g., de Melo-Martín and Intemann 2018: chap. 8; Fernández Pinto Reference Fernández Pinto2020). From this perspective, it is particularly important that there are no distorting incentivesFootnote ⁷ that would tempt scientists to deviate from their role responsibilities.

Sociological research on collaborating scientists confirms that trust is not only a matter of interpersonal relations but also of structures of reliability. It shows that scientists’ mutual trust is not “blind” (as Hardwig (Reference Hardwig1991) had once put it), but that they use an “ongoing probing process” and other strategies, e.g., indirect monitoring, for building mutual trust (Wagenknecht Reference Wagenknecht2015). The ability to collaborate is itself a capacity that scientists need to learn. As Gerken (Reference Gerken2022: 127) puts it: “part of scientific competence consists in a meta-competence that includes at least two aspects. The first aspect consists in some ability to recognize a competent scientific collaborator. The second aspect consists in some sensitivity to defeaters of intra-scientific testimony.” Nonetheless, in many cases, Hardwig’s (Reference Hardwig1991) basic point still stands: the sheer complexity of the division of academic labor makes it unfeasible to double-check everything and therefore makes trust within the scientific system unavoidable.

What does this fact of divided labor, and the need for trust within the scientific system, mean for the scientists who want to “build trust” with nonscientific audiences through science communication? The scenario I have in mind, in what follows, is a scientist who communicates with non-scientists, whether in classic science communication scenarios, e.g., giving public talks, or in science-policy contexts, e.g., in policy advisory boards. In such situations, a speaker invites trust through speech-acts such as accepting a public role or giving advice (Barimah Reference Barimah2024, drawing on Faulkner’s Reference Faulkner2007 view of affective trust, see also Kelsall Reference Kelsall2021).

This individual scientist speaks to non-scientists as a member of the institution of science – and the message he or she brings to the non-scientists will almost never be based only on his or her own work or that of close collaborators. To put it metaphorically, this person often becomes a spokesperson for a whole social system – if not “the academic world” as a whole, then at least the relevant disciplines and lines of research they draw on in their work.Footnote ⁸ Even if invited to speak only about their own research, they will typically need to draw on other scientists’ research.

In the scenario I have described – and which will be the focus of my discussion from now on –, it seems almost unavoidable for communicating scientists to say things like “This has been confirmed in a neighboring field of research” or “I here draw on peer-reviewed research from field x,” without having the expertise to judge the quality of that research (e.g., they would never themselves be invited as peer-reviewers for these studies). Often, the best response would in fact be: “I can give you the rough outlines, but in our present context of discussions, some of the details really matter, and unfortunately I am not sufficiently steeped into the material to give them to you, we need to find someone else,” or “Please give me two hours, I first need to talk to a colleague who really knows about this area.” In some circumstances, this may be possible, and it would indeed be the right thing to do. However, in many communicative settings, such strategies are not practically feasible – think about public podiums, life interviews, or committee meetings with policy-makers under time pressure. It may then be almost unavoidable for scientists to “epistemically trespass.”

The concept of “epistemic trespassing” (Ballantyne Reference Ballantyne2019; Gerken Reference Gerken2018; DiPaolo Reference DiPaolo2022; Hardwig Reference Hardwig and Wueste1994) describes the phenomenon of experts testifying about a domain in which they are not experts. As DiPaolo argues, this is deeply problematic: non-experts depend on experts, and this makes them vulnerable, “by requiring partial abandonment of their intellectual autonomy” (2022: 236). Often, “[p]ractical dependence accompanies epistemic dependence” (2022: 236), when non-experts need to act on the beliefs they form on the basis of trust in science. Trespassing experts therefore “irresponsibly abuse their authority and neglect the responsibilities they have toward epistemic dependents” (Di Paolo 2022: 238).

Nonetheless, as has been conceded from the beginning, a complete avoidance of trespassing would be problematic as well. Gerken (Reference Gerken2022: 166–7; Reference Gerken2023), for example, discusses this as a problem for intra-scientific testimony, and briefly mentions the challenge for testimony to non-scientists as well:

More recent accounts of trespassing have, indeed, painted the phenomenon in a somewhat more positive light. Pavličić et al. (Reference Pavličić, Dimitrijević, Vučković, Đorđević, Nedeljković and Tešić2024) point out that many scientific breakthroughs happened because scientists trespassed into another area and brought new perspectives (although they admit that trespassing can also lead to real harm). They also emphasize that many interdisciplinary forms of research, e.g., in nutrition science, would be impossible without trespassing, and therefore suggest calling it “hospesism” instead (ibid., 259ff.). Watson (Reference Watson2022) suggests replacing the metaphor of “trespassing” by that of “epistemic neighbourliness”: it describes “the claim that domains can be nested inside one another or they can overlap, and that epistemic authority is better analyzed as a function of the degree of overlap and nature of the shared epistemic project than by whether experts stay within their respective boundaries” (2022: 15).

Nonetheless, this does not mean that scientists engaging in science communication would not face normative challenges here. For the sake of argument, we can exclude notorious cases such as Nobel-prize-winning chemist Linus Pauling, who ended up publicly promoting vitamin C as a cure for cancer (Ballantyne Reference Ballantyne2019: 195–96). Such cases are, arguably, normatively overdetermined in their wrongness. But even well-intending, epistemically humble scientists who are invited to speak about their own research to a nonscientific audience can be faced with questions about neighboring areas, or may need to use results from other areas as input into their own research, without having produced them themselves. Of course, every academic in such a situation should do the obvious “homework” before: reading up on relevant research, double-checking claims they vaguely remember from earlier projects that might be of interest to the audience, etc. But there are practical limitations, stemming from at least two directions: one is that it is not always possible to predict what one will be asked about and which turn the discussions will take; another is the sheer amount of time that is needed for such preparation, and which the hectic life of academia often hardly admits.

This problem is particularly urgent for scholarship that cuts across disciplines. While it is often feasible to acquire some “interactional expertise” (Collins and Evans Reference Collins and Evans2007: 14), in the sense of being able to understand the methodology and results of studies in neighboring fields that one relies on, interdisciplinary researchers cannot fully grasp all details from other fields on which their research draws. The challenges multiply if one needs to draw on work from fields that are known to have problems, e.g., the replication crisis in social psychology. Scientists in, say, organization science, arguably need to be extremely careful when their research is connected to such fields and they communicate about it to the public.

Ironically, these problems are particularly pronounced in the more dialogical forms of science communication that are seen as most suitable for building trust in science. In “sending” models of science communication – e.g., TED talks with a lone scientist in the spotlight – careful preparation can minimize the risk of harmful trespassing, e.g., by carefully double-checking every study from a neighboring field one draws on, running one’s claims by colleagues from different fields, etc. But such forms of science communication have rightly been criticized for relying on an outdated “deficit model” (the public has a knowledge deficit that the communicating scientist needs to fill). Instead, more “dialogical practices” of public engagement are seen as the gold standard of developing trust in science among laypeople (see e.g., Ivani and Dutilh Novaes Reference Ivani and Dutilh Novaes2022 for a critical discussion and defense). But in such settings, it is much more likely that scientists are confronted with unexpected questions, or that the discussion turns to topics in which one has some “interactional expertise” – which may still be greater than the expertise of one’s interlocutors – but not the deep expertise of one’s own research. Hence, the risks of “epistemic trespassing” and the need to carefully avoid harmful forms of it are particularly great.

This systemic dimension of “building trust in science” has not been discussed much in the literature on trust in science, where the focus is mostly on the trustworthiness of individual scientists communicating with the public. For example, Intemann (Reference Intemann2024: 3, building on Goldenberg Reference Goldenberg2023: 370) argues that “epistemic trustworthiness” involves “four key features”:

A scientist who shows these features gives “laypersons good reasons to think the conditions of epistemic trustworthiness are met” (Intemann Reference Intemann2024: 4; see also Irzik and Kurtulmus Reference Irzik and Kurtulmus2019: 1146Footnote ⁹ ). Of course, there may be questions about the ease with which non-scientists can evaluate whether communicating scientists show these features (e.g., Anderson, Reference Anderson2011; Goldman, Reference Goldman2001). It also depends on one’s conception of science communication how the division of labor between experts and laypersons is imagined in detail. Gerken, for example, argues for a model of “justification expert testimony,” according to which communicating scientists should, “whenever feasible, include appropriate aspects of the nature and strength of scientific justification, or lack thereof, for the scientific hypothesis in question” (2022: 158, see also pp. 185–7 for a parallel norm for science reporters such as journalists).

My point is not to criticize these conditions of trustworthiness (or to defend this specific view of science communication). Rather, it is that whatever conditions and task descriptions one defines, one misses part of the picture if one focuses only on one individual, namely, the communicating scientist. Trust needs to extend to the whole group of scientists involved in the relevant research (see also Wilholt Reference Wilholt, Brady and Fricker2016).Footnote ¹⁰ Therefore, the conditions of trustworthiness carry over to the scientists on whose work the communicating scientist draws: if they have not been epistemically competent, for example, this may undermine the trustworthiness of the communicating scientist as well. Or if the justifications they provide for a certain result – and which the communicating scientist passes on, either explicitly or implicitly as part of a more complex justificatory narrative – do not stand up to scrutiny, the justification that non-scientists receive is incomplete.

This makes the communicating scientist vulnerable to any form of negligence, incompetence, let alone fraud, by colleagues on whose work they draw. They might end up not being trustworthy, or failing to provide proper justification, not because of a fault of their own, but because something went wrong within the wider social system of science. Intemann acknowledges this point in a footnote. She writes:

However, what is here assumed, namely, that “these criteria can also be applied to groups or institutions,” is not something we can take for granted. It may be relatively unproblematic in the case of clearly delineated and organized groups – Intemann refers to the IPCC or CDC as examples – that make a collaborative effort to share their results among each other and to relate different forms of evidence to each other, agreeing on the specific ways in which they make inductive risk judgments and communicating these publicly. But even these groups draw on many studies done by nonmembers.

Moreover, many forms of science communication do not rely on such group processes of mutual scrutiny and integration of insights. Often, it is a single scientist who communicates and who draws on results they have taken over from others, as part of the overall narrative that they convey to non-scientists. In such situations, in which one scientist acts – voluntarily or not – as the “spokesperson” for science, great ethical care is needed not only for how one communicates about one’s own research, but also in how one relates to the system of science of which one is part. One obvious challenge is not to take credit for work done by others. Another is to convey, as precisely as possible, the degrees of expertise one has for different claims: has one done a study oneself, or together with others, or does one draw on other people’s work? How well can one judge the soundness of the methodology? Has one seen just one study about a certain effect, or is there a broader array of evidence, maybe based on different methodological approaches?

This dependence may be less visible because everything has “gone” well in the research on which a communicating scientist depends. But consider the many ways in which biases or methodological problems can make the confidence in past research results brittle. For example, many scientific fields, especially in medicine, have only recently started to reckon with the fact that much past research has been done on male patients only, which makes many recommendations that were previously taken for granted problematic for female patients. Or take the many methodological criticisms that have been brought forward by psychologists and other social scientists against the applicability of rational choice methodology as often used by economists. Being aware that such criticisms exist, but no consensus has yet emerged on how to deal with them, can further complicate the position of scientists who want to draw on such research in their communication to the public.

In an ideal scenario, a scientist would have enough time to carefully lay out these challenges – and the audience enough patience to listen to them. There would indeed be an opportunity to say: “Let me check this with a colleague and get back to you later about it.” If the discussion turns to a topic that really requires a different expert, the communicating scientist would suggest inviting a colleague from another field to take over from them. In longer-term collaborations, e.g., in team-based medical care (e.g., Watson 2022: 8), such reflexive ways of dealing with the divided labor of scientific research are possible. However, in many situations in which scientists communicate, this is practically unfeasible. In such settings, there can be a temptation to trespass in a different way: to trade different forms of authority for each other. I turn to this issue next, drawing on Weber’s (1978) classic distinction of three forms of authority.

3. Trespassing between types of authority

To discuss this related, but different form of trespassing, which has not been considered in the philosophical literature on science communication so far, I draw on a classic tripart distinction from political theory. Max Weber’s distinction between traditional, bureaucratic, and charismatic authority may be well-worn, and yet it throws interesting light on the problem I analyze in this paper, especially that of science communication with a broader public.Footnote ¹¹ Here is a brief summary of Weber’s ideas (1978). The first type, traditional authority, claims legitimacy “by virtue of the sanctity of age-old rules and power” (ibid., p. 226). It can often be found in close-knit communities with “personal loyalty which results from common upbringing” (ibid., p. 227). The second type, “rational legal authority” (ibid., pp. 218–225), in contrast, is based on rule-based decision-making by bureaucrats with their specific offices, which they are supposed to execute in a “spirit of formalistic impersonality: ‘Sine ira et studio’” (ibid., p. 225). As such, it means “domination through knowledge,” which makes it “specifically rational” (ibid.). Thirdly, charismatic authority is exercised by someone who rules by help of “a certain quality of an individual personality by virtue of which he is considered extraordinary and treated as endowed with supernatural, superhuman, or at least specifically exceptional powers or qualities” (ibid., p. 241). Historically, those were often religious leaders, e.g., prophets, but in any case, it is a matter of “extra-ordinary” personalities (ibid., p. 244). Charisma is often experienced as “a ‘call,’ a ‘mission’ or a ‘spiritual duty’” (ibid.).

To be sure, this is a very rough distinction, which can be no more than a heuristic for thinking about certain problems. Like all heuristics, it has limitations, but it is useful insofar as it lets us see certain things more clearly. And while scientists who communicate with a broader audience or with policymakers do not have direct authority – only “epistemic authority”Footnote ¹² – it is interesting to think through the question of how they can build trust with the help of Weber’s three categories.

Traditional authority would mean that scientists are trusted because they have always been trusted. Descriptively, this may sometimes be the right way of understanding trust in science – think about the way in which doctors were, and sometimes still are, seen as authorities whose judgment lay persons simply have to accept, because it is part of their community’s mores that one “listens to the doctor.” As empirical work on science communication shows, the typical “highly visible scientist” is male, elderly, and good-looking, playing into clichés about traditional authority, maybe mixed with some charisma – a pattern that a recent study about highly visible “pandem-icons” during the COVID-19 pandemic confirmed (Joubert et al. Reference Joubert, Guenther, Metcalfe, Riedlinger, Chakraborty, Gascoigne, Schiele, Baram-Tsabari, Malkov, Fattorini, Revuelta, Barata, Riise, Schröder, Horst, Kaseje, Kirsten, Bauer, Bucchi, Flores, Wolfson and Chen2023; see also Fahy Reference Fahy2015). But from a normative perspective, this is a problematic model: it provides no reasons for justifying the trust that laypeople put into scientists. The fact that scientists have always (or rather, for a long time, in certain societies) been listened to does not, as such, explain why they should be listened to.

Bureaucratic authority seems to come closer to the view of science as a social system that I have described above. While the parallel is, of course, not exact, some features are comparable: the existence of clear rules and interlocking roles, the possessors of which are supposed to contribute to the overall functioning of the institution “sine ira et studio.” What is obviously different is that in the scientific system, it is not hierarchy that creates authority; as is well-known, social hierarchies introduce many problems in what should be a knowledge-oriented social system. Also, science as a system is not “value free,” as authors in the philosophy of science have shown against Weber’s use of that term (e.g., Douglas Reference Douglas2009; Elliot 2017). But it is the impersonality that is comparable: just as it should not make a difference which official a citizen speaks to when putting down a request at a public office, the messages that reach the audience from the scientific community, on a particular topic, should not depend on which scientist conveys them.

What about charismatic authority? Like traditional authority, it obviously cannot deliver a normative basis for public trust in science. But at a descriptive, and in particular at a psychological level, it is an interesting category. In a media environment dominated by celebrities, it is quite common to see particular scientists rise to prominence as the “public face” of their field.Footnote ¹³ Often, this has to do with features such as relatability, rhetorical skills, or even simply good looks (Joubert et al. Reference Joubert, Guenther, Metcalfe, Riedlinger, Chakraborty, Gascoigne, Schiele, Baram-Tsabari, Malkov, Fattorini, Revuelta, Barata, Riise, Schröder, Horst, Kaseje, Kirsten, Bauer, Bucchi, Flores, Wolfson and Chen2023). Some of these “science stars” are willing to open up their whole life to the public, e.g., doing home stories with gossip magazines. Their names become brands, which can open doors for lucrative book deals and consultancy gigs. This phenomenon can be seen against the context of not only a historiography of science that often misleadingly celebrates lone geniuses (see, e.g., Gerken Reference Gerken2022: 148), but also political developments that see more and more populist leaders who present themselves as the direct voice of the people (e.g., Mueller Reference Mueller2017). Social media facilitates the (seeming) closeness of “leaders” and “followers.”

It is here that individual scientists who communicate with an audience of non-people may run into a dilemma: effectiveness may call for traditional or charismatic communication, even though this wrongly places the reasons for trust on the character of a single individual, not on the system of science as a whole. Even though one may deplore this fact, on a purely descriptive level, the efforts of such “science stars” for building public trust in science may in fact be more important than many other strategies. One can grant that some forms of stardom may be put aside as unethical for independent reasons (e.g., because they reflect an idea of superiority that is in tension with a democratic understanding of moral equality). But apart from egregious cases, it is worth emphasizing that charisma may well have positive effects in terms of reaching audiences and building trust. Research by Fiske and Dupree (Reference Fiske and Dupree2014), on public communication by different professional groups, shows that scientists are perceived as having high competence but showing low levels of “warmth.” According to these authors (Fiske and Dupree Reference Fiske and Dupree2014: 13595), scientists “earn respect but not trust.” Being charismatic may offer a counterweight to this tendency of not being perceived as “warm.” Another study found that more personal communication (e.g., selfies, in contrast to pictures of apparatuses or diagrams) on social media leads to a greater perception of trustworthiness (Jarreau et al. Reference Jarreau, Cancellare, Carmichael, Porter, Toker and Yammine2019). Again, the personal dimension of communication seems positively correlated with trust in science.

Therefore, playing into the charismatic line may have large payoffs in terms of increased trust in science, especially for scientists who have good communicative skills and winning personalities.Footnote ¹⁴ Strategies that draw on personal charisma may have better chances to reach large audiences in today’s media environment. And to be sure, not all communication based on charismatic authority is morally wrong, all things considered. If important goods are at stake, worries about using the wrong means of communication may seem a luxury that needs to be set aside. And one should also grant that there can be many shades of gray in the ethical evaluation of charismatic communication. Using drastic or funny language while presenting a rational argument, for example, may be considered less problematic than mongering fear or other emotions in ways that overshadow the rational core of the argument.Footnote ¹⁵

And yet, drawing on charismatic authority is a form of trespassing in the sense that it neglects the reasons why individuals should trust scientists and replaces them with different reasons. The implicit message of science communication should not be: “trust me because it is me who says this,” but rather “trust me because we (!) did this according to the best possible scientific methods and standards.” Using charisma to build trust in science remains a problematic strategy in this sense, and it is made worse by the fact that untrustworthy actors often exploit charismatic strategies as well, sometimes even by playing precisely the role of charismatic scientist that is here under discussion. Especially when one thinks about the trust relation between science and society as a long-term, structural relation, it becomes clear that this relation should be based on the right reasons for trusting science, not on the fact of having charismatic people as the public faces of science.

Another way of putting this point is that different goals of science communication can here stand in tension with each other. Intemann (Reference Intemann2020: 282) suggests the following goals: “(1) accuracy, (2) understandability, (3) predictive relevancy, (4) generating enthusiasm/interest, and (5) facilitating trust.” These goals can conflict (ibid.), and here they indeed do: accuracy (especially about the process of knowledge generation and the reasons why science is trustworthy) and maybe also predictive relevancy pull toward an emphasis on the system of science, but enthusiasm/interest and trust pull in the charisma-based direction.

Is this a problem that scientists simply have to live with, at least in the current media environment? To be sure, they can try to find outlets for science communication that offer more favorable circumstances, in which there is a chance to be clear about the limitations of one’s own expertise and the systematic reliance of scientific knowledge on the scientific system as a whole. But this will not always be feasible. In the next section, I suggest another strategy, which has not received enough attention in the discussion about trust in science and the ethics of science communication: to think of it as a team task.

4. Building public trust in science as a team task

If one takes a broader historical and sociological perspective, it becomes clear that the communication between science and society has often been embedded in structures that included more people than just “the scientist” and “the audience.” In the 20^th century, star scientists with their own TV programs coexisted with many science journalists in less visible, but structurally important roles for regularly bringing updates about scientific insights to the public. Since the rise of the internet, however, with advertising money shifting to the online sphere, many newspapers had to reduce staff or completely shut down their science desks because of financial problems (Weingart and Guenther Reference Weingart and Guenther2016). Other channels of communication went via membership organizations, e.g., unions or NGOs that had ties to research institutions in areas relevant to them, and these continue to exist – but a lack of funds in many such organizations can also lead to a thinning out of the “intermediary experts” that can help translate and contextualize relevant scientific findings.

It is against this broader historical context that the calls for more science communication need to be understood. To some extent, PR departments of universities and free-lance science communication experts have filled the gaps, with science communication having “developed into an industry” (Weingart and Guenther Reference Weingart and Guenther2016: 1). However, while the older model of science journalism was guided by journalistic standards (and other intermediate experts probably by the interests of their audiences), in many of the new forms of science communication, the motives are not immediately clear. Weingart and Joubert (Reference Weingart and Joubert2019) distinguish between two types of science communication activities, each with several subcategories: the first type is “driven almost exclusively by educational objectives, is executed without an institutional interest influencing its contents, serving only the interests of the target audience” (2019: 8). In cases of type two, in contrast, the communicator (typically a political agency, a university, or an academic organization) entertains other interests: “in self-promotion, in augmenting institutional prominence, and in attracting attention” (ibid.). Weingart and Joubert call for clarity about motives, because a “conflation of motives … ultimately threatens the credibility of science as an institution, of scientific organisations and the mission of science communication” (ibid., 3).

This is an important warning, but it does not mean that scientists who want to build trust in science cannot benefit from looking at some of the historical practices and what worked well in them. Specifically, I want to argue that the fact that science journalism was often a team approach, with clearly defined roles, can help address the two types of trespassing that I have discussed above: trespassing between disciplinary fields and trespassing between forms of authority.

I hasten to add that I do not take this to be a one-size-fits-all solution; all answers to these challenges need to be highly sensitive to the different contexts and issues at stake. For example, in crisis situations in which human lives are at risk, prioritizing effectiveness through charisma may be morally legitimate and even required. This may bring short-term gains in trust; however, these need to be weighed against possible losses, namely, the perpetuation of a wrong picture of how science functions and of what justifies its authority, which may create risks of backfiring in the longer term.

What, then, do I mean by a team approach? Scientists can work in teams (e.g., at the level of research institutes or university, or research alliances across institutions), supported by communication experts who help them place their message in a way that takes into account the characteristics of different media and the best chances of reaching the target audience, all while paying an eye to the values of science communication, such as integrity, accessibility, sensitivity to the specific audience and its needs, etc. They can also, depending on the opportunity and the state of evidence in a certain field, invite colleagues who are critical of their own interpretations of the data and might suggest different ones, so that the audience sees the process of scientific deliberation in action. When it comes to policy issues, they can write reports together, carefully weighing the insights from different lines of research.

Such an approach would take up Watson’s (2022) notion of “epistemic neighbourliness” from the context of interdisciplinary research and transpose it to contexts of public engagement and “building trust in science.” Watson (2022: 16–18) uses the notion of “councils of trust” whose collaboration can create conditions of trustworthiness for lay people. They would consist of “several experts in overlapping fields” who together can vouch for certain information, as when the family physician can tell the family that vaccines do not cause autism because he is part of such networks. To function well, these experts must be “free to disagree” (ibid., p. 17), and they must be “from neighboring, or “linked” domains” (ibid., p. 18). My suggestion to see science communication as a team approach can be understood as a strategy that builds such “councils of trust” for the audience. Watson is fully aware, of course, that this approach can also run into problems, but nonetheless suggests that the insights of such a council can “justify novices in identifying and trusting relevant experts” (ibid., 20).

The division of labor within such a team approach does not have to stop at scientific personnel, however. By drawing the circle even more widely, the role of charismatic authority can, where necessary, be integrated into a team approach, without falling back into old-fashioned clichés of “science stars” (good-looking elderly males in laboratory coats). The model I here have in mind could be called “the charismatic presenter who explains the process of science.”Footnote ¹⁶ Having relatable, charismatic communicators is important for winning trust in science. But these need not be “science stars” themselves – they can be science journalists who explain to the audience how science works and how they get information from scientists, e.g., by checking peer-reviewed studies and meta-analyses and talking to different researchers. Of course, this is a demanding task, and it typically requires a whole team in the background. This is a luxury that few science journalists enjoy these days. But it may be the best way of preventing the perpetuation of a problematic view of science and yet harvesting the benefits of charisma in terms of building trust.

Rather than calling on all scientists to be communicators, scientific institutions might want to consider how they could support such models, as a way to addressing the problems of trespassing between disciplines and between types of authority.Footnote ¹⁷ This is likely to lead to better outcomes than if all scientists, individually, communicate about the results of their own studies in an unmediated and uncoordinated process (see also Slater and Scholfield Reference Slater and Scholfield2022). Such unmediated communication – which social media invitesFootnote ¹⁸ – can all too easily fail to provide context or to balance the results, where necessary, against other research results, explaining their significance and relevance. The audience might also all too easily get the impression that there is an unorganized cacophony of voices in science, none of which is, per se, particularly trustworthy, especially if their claims clash. Therefore, the best strategy for individual scientists may often be not to always try to communicate directly, on their own, but rather to find opportunities to participate in team-based approaches to science communication – or to help build them where necessary. They can, for example, help science journalists find the best experts to talk to, to understand the relation between new studies and the state of the art, or to judge the degree of certainty with which they should communicate certain claims to their audience.Footnote ¹⁹

Before concluding, let me respond to two objections. The first is that the “team approach” can lead to new ethical problems and dilemmas. For example, there might be a problematic distribution of credits, or the invisibilisation of certain groups already marginalized in science (e.g., researchers of color) might be continued. Or there might be genuine questions about how to think of the boundaries of the relevant “teams” – are those meant to be the same people over a longer period of time, or might the constellation vary from topic to topic? How, for example, should one deal with different research paradigms in these contexts, which differ not only in methodologies but also in the underlying ontologies and values?

In response, let me grant that such questions can arise and need to be handled with care and ethical responsibility. Sometimes, a team approach may require painful compromises, e.g., when deciding whom to “send out” to the media. In a public health emergency, a team may decide to send out the stereotypical authority figure (the elderly man in the lab coat) because it is really important that as many people as possible trust the message – even though they would otherwise prefer to send out a young person of color in order to break with precisely that stereotype. Such judgment calls will require contextual sensitivity and practical wisdom – but then, this is also true for science communication in general, at least in today’s difficult media environment. Teams are, arguably, in a better position to take these ethical challenges seriously and think through them from different perspectives than individual scientists on their own. In other words, the very fact that science communication raises various ethical issues is, in fact, another argument for doing it in teams rather than alone, and the teams can then also deal with the new ethical challenges that may arise in the team approach.

A second objection is that the team versus individual issue is ultimately less important than another issue: that of communicating about results or processes. As many commentators on trust in science have pointed out, there is a tension between communicating a (seemingly clear) result and communicating about the process of creating scientific knowledge, in all its messiness and provisionality. A focus on the latter can mean that scientists communicate with lots of caveats and caution. But as Naomi and Oreskes (2010: 687) once famously remarked, referring to such a style of communication: “Although this care and nuance is intellectually scrupulous and admirable, being so philosophical about the ‘factual’ nature of climate change doesn’t serve public communication.”

In response, let me suggest that these two challenges are both important, and they both require attention from the scientific community, but also from the community of science journalists and other mediating instances. Both groups need to, on the one hand, position themselves within the current very non-ideal media landscapes, but also, on the other, ask themselves how they can contribute to improvements over the status quo, i.e., developing ways of communicating in which “care and nuance” are not rejected by the public. Teams of scientists need to be aware of this challenge and deliberate about the best ways of dealing with it. Arguably, in some countries, the science communication around the coronavirus pandemic did contain some positive movements in this direction, for example, regular in-depth interviews with scientists discussing the latest state of the art and the most recent developments that gave the audience a vivid sense of results and processes of scientific enquiry. While scientists can play a role in this, it is essential that (science) journalists and the media also do their share in moving in this direction.