I appreciate the opportunity to continue this conversation on how non-experts might identify and, thereby, come to trust experts. While so much of contemporary philosophical discussion might be called destructive—attempts to defeat an “opponent’s” claims through counterexample—this forum has been a refreshing exercise in constructive philosophy, wherein we attempt to build on one another’s insights in order to both better articulate the novice-expert problem and formulate plausible solutions… [please read below the rest of the article].

Image credit: Bongkarn Thanyakij from Pexels

I appreciate the opportunity to continue this conversation on how non-experts might identify and, thereby, come to trust experts. While so much of contemporary philosophical discussion might be called destructive—attempts to defeat an “opponent’s” claims through counterexample—this forum has been a refreshing exercise in constructive philosophy, wherein we attempt to build on one another’s insights in order to both better articulate the novice-expert problem and formulate plausible solutions. In response to Brennan’s most recent contribution, I echo his sentiment that I find little to disagree with. So, in this piece, I summarize some of our central points of agreement and then build on his suggestion of employing a “communal, jigsaw-puzzle piece approach” (2020b, 55) to solving the novice-expert problem.

Plight of the Navigator: Snark Hunting in a Sea of Distortions

When novices need specialized skills or knowledge, they must correctly distinguish experts from the morass of frauds, hoaxers, pariahs, and epistemic trespassers. In my initial response to Brennan (2020a), I used the wayward seafarers from Lewis Carroll’s poem, “The Hunting of the Snark,” as a metaphor for the plight of novices as understood through contemporary expertise studies (Watson 2020). One way for a novice to know enough about a domain to adequately identify an expert is to become an expert themselves. But in that case, they would, through their efforts, cease to be a novice. This would not be so bad except that the number of topics for which novices need expertise far outstrips their time, interest, and abilities to become experts in them all. Of practical necessity, then, novices must rely on experts. The question is whether there are good enough epistemic grounds for doing so.

As I also explained, in attempting to identify experts, novices are beset by at least three kinds of epistemic obstacles. They face internal obstacles associated with their cognitive and motivational limitations. Novices must care that they are identifying real experts and not simply people who share their political or moral views. They face external obstacles associated with the direct and indirect indicators associated with experts’ competence. These include the length and quality of an expert’s experience or practice, their degrees and certifications, their contributions to their domain, and their respect among other experts in their domain. And they face ecological obstacles associated with the conditions under which experts attempt to communicate with novices. An issue may be high-stakes, such as whether to stay home from work to prevent the spread of a dangerous virus; it may be time-sensitive, such as whether people should sanitize their next order of groceries or cancel an upcoming vacation to avoid exposure to that virus; and it may be highly politicized, such as whether a government should mandate vaccinations for such a virus.

The ubiquity of these obstacles and their myriad combinations make the novice’s prospects for successfully navigating the sea of putative experts daunting, at best, and insurmountable, at worst. Despite this, Brennan maintains a modicum of optimism, and I am happy to say that our ongoing discussion leads me to believe he is right in doing so. The devil, though, is in the details as to how novices can, in good epistemic conscience, come to trust experts.

Elephants, Experts, and Enemies of the People: On Gaining the Right Perspective

Brennan reframes the debate slightly, turning away from my metaphor of hunting the Snark and toward an ancient Indian parable, according to which six blind men attempt to describe an elephant, though from different angles. After describing quite different experiences, the men proceed to argue over the “true” description of the elephant. (We aren’t sure why. Maybe they want to build one?) I appreciate this redirection because, while I think my metaphor better captures the current state of expertise studies, Brennan’s suggests a way out of the conundrums posed by mine. Rather than arguing about what an elephant is like from within the constraints of one observer’s perceptions, the solution—sadly, unachieved by the folks in the parable—is for each observer to realize that their perspective is limited and that the others may have information they lack, and then to collaborate on a description of the elephant. If experts are more like elephants than Snarks, novices might not be in the dire position painted by contemporary expertise studies. As Brennan puts it, “The ecological as well as the motivational obstacles facing novices can be overcome if they can build the right coalition” (2020b, 57).

How might an epistemic coalition accomplish this daunting epistemic task? Brennan and I agree that its feasibility depends, first, on giving up a common assumption in expertise studies, namely, that experts and novices occupy opposite ends of the epistemic spectrum in the domain at issue. Rather than starting from the assumption that all experts are highly competent and all novices are uniformly incompetent, I agree with Brennan that we should begin from the assumption that “both novices and experts vary widely in their competencies” (Brennan 2020b: 55). We also agree that its feasibility depends, second, on “a richer picture of the overlap in domains” including “how competence in those areas of overlap” (Watson 2020, 57; Brennan 2020b, 56). Finally, we agree that these assumptions make the problem “thornier” in one sense, but potentially more tractable in another. In what remains, I expand on these senses.

I’m not sure what Brennan would say, but I think one way these assumptions make the problem thornier is by implying that there is no God’s-eye-perspective from which novices can begin the search for experts. Consider Evan Selinger’s (2011) discussion of Henrik Ibsen’s play, An Enemy of the People (1882). In the play, Thomas Stockmann is a doctor at a public bathhouse in a small Norwegian tourist town. Stockmann discovers that a poison is contaminating the water of the baths and takes this information to the public, confident that they will rally together to solve the public health threat by changing the location from which the water for the baths is sourced. Surprisingly (for Stockmann, anyway—the title gives it away for the rest of us), the townspeople reject his advice and refuse to fix the problem. Addressing the pollution would disrupt their lives and livelihoods, and they perceive that such a disruption would be too great a cost for a problem they are not currently experiencing. They value their current way of life more than the disruption of fixing the problem. They might say: Sure, Stockmann is an expert on technical matters. But we are the experts on the life we want to live. His advice is not relevant to our domain of expertise. Stockmann thought he would be the people’s hero, but instead, he has become their enemy.

We in the audience are appalled at what, today, we would call their “science denial.” Of course Stockmann’s advice is relevant to the life they want to live, we say to ourselves. The people are simply ignoring the fact that their way of life will be disrupted one way or another, either now, by prophylactic measures, or later, by the poison.

Interestingly, the example strikes us as appalling primarily because we are sitting outside that world, with a God’s-eye perspective, in the audience. The play’s narrative makes it clear to us who has the “better” perspective on the matter. But our outsider perspective is misleading. As Selinger puts it: “The person in the audience is not standing anywhere, not situated with respect to this aspect” (2011, 35). In the real world, there is no audience. We are in the play, so to speak.

In any real controversy, … no one occupies the audience position; everyone is as it were “on stage” in a situated position, standing someplace with particular involvements which give rise to a particular understanding of the situation and, with it, an inclination to accept some people rather than others as authoritative (Selinger 2011, 35).

When there is no audience, we cannot be neutral about whom we trust. To see how this can play out in real life, consider the example of expert safety advice regarding herbicide 2,4,5-T (Collins and Evans 2002, 277). No less than eight times, farm workers in the UK raised concerns to the Advisory Committee on Pesticides (ACP) about the safety of the herbicide. Each time, experts assured them that the herbicide was safe when used as directed. It turns out that the workers had a perspective on the issue that the ACP did not. “Farm workers, on the other hand, argued that, because the appropriate precautions could not be taken in the day-to-day settings in which the chemical was actually used, then it was not safe” (277). The point is that, even if the farmers trusted the experts that 2,4,5-T is safe when used correctly, they could still reject their advice that it is safe for them because they had good reason to believe it could not be used “correctly.” Their “embodied experience” (as Collins and Evans call it) on the use of the herbicide put them in a better epistemic position regarding 2,4,5-T’s safety than the regulators’ scientific expertise because the regulators were evaluating the product from an abstract laboratory perspective. In the terms of Brennan’s metaphor, the two groups were looking at different parts of the elephant. They were experts on different aspects of the problem, and, consequently, they were novices on other aspects of it.

So, the problem of trust in experts is thornier when we acknowledge the variation in competence among experts and non-experts, but it is the situation in which we find ourselves. Happily, though, this acknowledgement also suggests a strategy for navigating out of the sea of distortions (or rather, for building a better elephant). Returning to our question, then: How might a coalition help?

Building a Better Elephant

In closing my initial response to Brennan, I pointed out that, in many cases, novices do pretty well at identifying and trusting experts. Clearly, things are more complicated when it comes to highly politicized or heavily value-laden issues. But there is little reason to think that a novice is not exhibiting epistemic responsibility when she chooses to trust someone as a medical expert who is in a hospital, wearing a white coat, has an “MD” badge, and claims to be a doctor. (To be sure, there are cases where this trust would not be warranted, but those are rare enough that it would not be a mark against her epistemic responsibility.) The relevant question is how novices can distinguish the epistemically relevant differences in this case from those, for instance, of someone on the internet who is wearing a white coat with an “MD” badge and claiming to be a doctor. Which better describes an elephant?

One key difference in the cases is how the evidence related to credibility in each case is linked to evidence the novice already has good reason to trust. Recall from my initial response to Brennan that novices’ prospects for accurately identifying experts depends partly on their ability to assess how close an expert’s domain is to what C. Thi Nguyen (2018) calls their “cognitive mainland,” that is, how closely their knowledge and skills are related to things they already understand well enough to assess.

To see how this is relevant, consider how experts trust other experts when building a skyscraper. Experts from various construction backgrounds come together to build something hundreds of feet tall that supports thousands of people while withstanding all manner of weather. Yet, the people who collaborate on the project have expertise in a wide range of very different specializations—electricians, plumbers, structural engineers, crane operators, welders, etc. And none of these becomes an expert in any of the other domains through the process. They simply have to trust that the others (especially the crane operators) are competent do their parts of the job.

How, then, does a construction manager, who is not a crane operator, come to trust a crane operator, who is not an expert construction manager, and vice versa? Importantly, each knows enough about the skills required that they can assess whether the other is sufficiently competent at their job. Their skills and knowledge overlap in relevant ways.

Now consider an expert once removed: the investor who wants to build the skyscraper. Surely, she knows less than the construction manager about whether the crane operator is competent. But, if she’s a good investor, she knows enough about construction management to choose a competent one. Finally, consider the expert corporate executive who needs the skyscraper for her company. While she may not be able to assess construction managers, she has the skills to assess investors. How does a corporate executive justifiably trust a crane operator? She asks the investor, who confers with the construction manager. With this chain of trust in experts, people who are not experts in the target domain can ride the elevator from the garage to the seventy-fifth floor and back day-in and day-out, confident that the folks who built the building knew what they were doing.

The skyscraper example suggests that different kinds of expertise are linked in ways that allow people who aren’t quite experts in a domain to recognize those who are, and this marks a path for those who are not even close to being experts in a domain to successfully identify those who are. And this, of course, is just one way that expertise can be linked, namely, an overlapping of shared knowledge regarding a specific project—what Stephen Turner calls a “coordination heuristic” (2014, 247).

There are other kinds of links, as well, such as an overlapping of shared training. Hematologists and nephrologists, for example, are both medical specialists, so they are trained to understand the basics of one another’s specialties. In this way, a hematologist understands why they can trust a nephrologist, and vice versa. Shared specialization further allows family practice doctors to trust hematologists and nephrologists, and, similarly, allows people with some basic understanding of biology and disease to trust family practice doctors (in person, not on the internet). If someone without a medical degree is concerned about a nephrologist’s competence, she might ask her family practice doctor to help her decide whether she needs a second, or even a third, opinion. Call these domain links a “nesting heuristic.”

These examples suggest that coalitions can help because they link domains of expertise in ways that support justified trust in experts. It is noteworthy, given so many appeals in the expertise literature to credentials and track records, that experts rarely conduct extensive background checks on other experts. A nephrologist does not try to find out what place the oncologist graduated in their class, their publication record, whether they have demonstrated trustworthiness, or even whether they have a track record of success. That’s not to say that these features do not figure into domain linkages, but simply that those features are not the most accessible or even the most relevant to adequately identifying experts.

There is an additional aspect of the examples above that is less evident but no less important. All the linked domains mentioned operate under accountability conditions—market forces, accreditation requirements, or easy access to litigation. It is more epistemically responsible to trust a roofer when you know that, if they make a mistake, you can hold them accountable to fix it. And if they know that, they are less likely to make the mistake in the first place. Turner calls these “bonding” conditions (2014, 188ff). Bonding conditions introduce transparency into the novice–expert transaction that is not present on, say, the internet or in political speech. Without a means of holding putative experts accountable, novices are less like elephant builders and more like Snark hunters. Their coalition may do more harm than good.

Brass Tacks: Can Novices Really Do It?

Are novices in a position to build the right team? Certainly not all. There are plenty of people who lack specialized training and whose lives are sufficiently isolated from those who have it that most claims to specialization will be subject to distortions along the lines of the three types of obstacles described earlier. Their cognitive mainlands are relatively small and poorly linked to other domains.

Further, some of the experts closest to these epistemically isolated folks share their idiosyncratic values. For example, some pharmacists still refuse to fill birth control prescriptions on the grounds that doing so is like participating in an abortion despite vast evidence that birth control is not an abortifacient. And some oil refinery engineers reject claims about anthropogenic climate change for much the same reason as the townsfolk in Ibsen’s play: because accepting them would negatively affect their livelihood. We are in the play with these experts, and most of us are not pharmacists or engineers. So, if we do not have access to just how idiosyncratic these views are, that is, if we lack the relevant links to a wider set of expert perspectives, then it would be hard to blame us for believing that these experts accurately represent their domains. And unless we are extraordinarily motivated to research consensus views among pharmacists in our free time, this sort of isolation puts identifying experts largely out of reach.

But all is not lost. Our discussion suggests a path forward, both for novices and for expertise studies. If we can identify cases where novices (real novices, not Snark hunters with blank maps) are justified in trusting experts (real experts, not mythical repositories of knowledge) and explain what features of those cases render trust justifiable, we might be able to replicate those features, for ourselves and others. I think a good start is understanding better how domains are linked, and to use that understanding to build coalitions of trust.

References

Brennan, Johnny. 2020a. “Can Novices Trust Themselves to Choose Trustworthy Experts? Reasons for Reserved Optimism.” Social Epistemology 34 (3): 227-240.

Brennan, Johnny. 2020b. “Finding the Snark Together: A Response to Watson and Hinton.” Social Epistemology Review and Reply Collective 9 (6): 54-59. https://wp.me/p1Bfg0-58a.

Collins, Harry and Robert Evans. 2002. “The Third Wave of Science Studies.” Social Studies of Science 32 (2), 235-296. https://doi.org/10.1177/0306312702032002003.

Nguyen, C. Thi 2018. “Cognitive Islands and Runaway Echo Chambers: Problems for Epistemic Dependence on Experts.” Synthese. 1-19. https://doi.org/10.1007/s11229-018-1692-0.

Selinger, Evan. 2011. Expertise: Philosophical Reflections. Automatic Press/VIP.

Turner, Stephen P. 2014. The Politics of Expertise. London: Routledge.

Watson, Jamie Carlin. 2020. “Hunting the Expert: The Precarious Epistemic Position of a Novice.” Social Epistemology Review and Reply Collective 9 (4): 51-58. https://wp.me/p1Bfg0-4YG.

Categories: Critical Replies, Uncategorized

Tags: C. Thi Nguyen, embodied, epistemology of expertise, Evan Selinger, expertise, experts, Harry Collins, healthcare, how novices identify experts, Jamie Carlin Watson, Johnny Brennan, medical expertise, parable of the blind men and the elephant, Robert Evans, Stephen Turner, trust