Epistemic trespassing, or epistemic squatting?
Who gets to define the boundaries of expertise?
The other day I saw someone use the phrase “epistemic trespassers” in a Twitter rant. I traced it back to this tweet, which links to an essay (paper?) by the Fordham University philosopher Nathan Ballantyne:
Ballantyne’s essay is basically a 24-page argument that people should stay in their intellectual lane. He starts out with some well-known examples of people who are respected experts in one field becoming quacks in another field — Linus Pauling hawking vitamin C, and so on. Ballantyne then cites other people who have complained about the same phenomenon (Plato!). Following anecdote and argument-from-authority, he then goes on to make a number of conjectures about the harms from epistemic trespassing. Finally, having argued to his own satisfaction that epistemic trespassing is a problem, he throws out some proposals for solutions — basically, more interdisciplinary collaboration and intellectual modesty.
Now, I’m all for interdisciplinary collaboration and intellectual modesty, as a general rule. But although Ballantyne raises interesting points and creates food for thought, he fails to make a conclusive case that what he calls “epistemic trespassing” is, on balance, bad for society. And his arguments raise uncomfortable questions that he doesn’t really wrestle with — most importantly, the question of who gets to decide who’s a trespasser.
Who gets to decide who’s a trespasser?
Identifying trespassers is not a trivial task. Ballantyne declares that “epistemic trespassing of the sort I’ve noted is easy to recognize,” but his own list of examples contains a glaring question mark — Richard Dawkins’ writings about religion.
Now, I am not a big fan of Richard Dawkins. But it’s highly questionable to declare that in writing about religion, he has trespassed on someone else’s field of expertise. For Dawkins, writing about religion is not a lark, or even an avocation — it has become the primary focus of his intellectual career for at least a decade. Dawkins writes books about religion; he no longer does biology. Who is to say that Dawkins is not a scholar of religion now?
Ballantyne’s answer is: Philosophers of religion. He writes that “experts in the philosophy of religion—atheists, agnostics, and theists alike—charge that Dawkins fails to engage with the genuine issues and sets up strawmen as his dialectical opponents.” But did Ballantyne take a poll of these philosophers, or merely give his impression of what he thinks they think?
More fundamentally, what gives these philosophers of religion the right to gatekeep religious thought? Yes, they have thought about religion, discussed the topic with each other, cited each other’s papers, and so on. Perhaps Dawkins’ arguments are all stupid. But so what? On questions of the divine and the ineffable, do philosophers really have more access to truth than, say, clergy? And on questions of the social impact of religion, are philosophers’ logical arguments superior to the quantitative research methods of sociologists, anthropologists, and so on? Why are the philosophers the owners of this piece of epistemic real estate, rather than trespassers themselves? (Update: In an ironic twist, several commenters inform me that many religious studies scholars view philosophers of religion as epistemic trespassers!)
And who cares if Plato thinks that craftsmen and poets need to shut up and stay in their lane? Who the hell is Plato?
Contrary to what Ballantyne says, it’s not always clear who’s an expert and who’s a trespasser. In practice, expertise is defined by consensus among communities of people who all or mostly accept and promote each other as experts. These communities can be formal, like the American Economic Association, or informal, like “DSGE macroeconomists”. But the basic idea here is that these communities are self-judged — they deny outsiders the right to adjudicate whether they possess actual expertise.
This strategy often produces good results, as demonstrated by the remarkable practical success of a bunch of scientific fields. But it’s not foolproof. Thanks to the quirks of human sociology, it’s possible for exclusive communities of self-described experts to arise who don’t actually have the expertise they claim — and even for these communities to be recognizes as experts by the broader public, at least for a time. For example, the world has now woken up to the fact that string theory is not empirically testable. But for a long time, the field was held in near-universal acclaim, with string theorists issuing repeated false promises of testability. For decades, mathematician Peter Woit — himself an epistemic trespasser! — has documented these false promises on his blog, along with various efforts by the string theorists to defend the exclusivity of their intellectual enterprise.
Another example is the so-called “freshwater” school of academic macroeconomics, who for decades created models that bore little resemblance to reality, but who all cited each other’s papers and boosted each other’s work, and even won Nobel prizes. Unlike string theory, the faux-expertise of freshwater macro ended up mattering a lot for the real world; freshwater “experts” gave very bad advice during and after the financial crisis, which probably influenced policymakers to withhold needed fiscal stimulus.
Now, in both of these cases, the piercing of the bubble of faux-expertise was done mostly (though not entirely) by scholars in adjacent fields — for example, Lee Smolin’s critiques of string theory and Paul Krugman’s critiques of the freshwater school. But in both of these cases, the critiques were made publicly, to non-experts, and that’s what made all the difference. Economists had been accurately debunking the “freshwater” gang’s models for years, and yet through the power of social dynamics and collective self-promotion they managed to hold onto their intellectual dominance until the case against them was taken to the public.
And in some fields, it’s not even clear what the relevant criterion of expertise should be! In chemistry it’s fairly easy to verify who’s a crank and who’s a real expert — the experts can predictably make cool and useful stuff happen with chemicals, and they can tell you how to replicate their feats at your factory or with your home chemistry set — and the cranks cannot do this. But in fields like literature, religion, or even philosophy itself, it’s harder to point to any sort of objective test of validity. This probably makes fields like these even more subject to squatting by groups of people who have little claim to expertise beyond a mutual agreement to represent each other as experts. But as we’ve seen with string theory, even the “hardest” sciences are not immune to this sort of epistemic squatting.
In other words, the adjudication of expertise — and thus of epistemic trespassing versus epistemic squatting — is a fundamentally human endeavor, subject to all of the pitfalls and foibles of other collective human endeavors. As Juvenal (surely an expert) once asked, “Who will guard the guardians?” Who assigns the property rights to various pieces of epistemic territory to specific groups of human beings? Ballantyne’s essay implicitly raises this crucial question, but fails to really address it.
Costs of epistemic enclosure
Ballantyne’s essay is flawed in another major way — it fails to wrestle with the question of whether epistemic trespassing might have clear, systematic benefits, in addition to the potential costs he outlines. He discusses some “defenses” that trespassers might use in order to justify why their extra-disciplinary efforts aren’t so bad, but doesn’t even seem to consider reasons they might be actively good.
Yes, it’s easy to point to examples of experts straying outside their lane and acting like cranks — in fact, when it comes to physicists doing this, it’s a popular joke. But Ballantyne could just as easily have cited examples of people making big breakthroughs by straying outside their core expertise — mathematician John von Neumann’s dabbling in economics, brewer W.H. Gossett’s contributions to statistics, Richard Feynman helping solve the Challenger disaster, and so on.
In fact, it’s not necessary to rely on anecdote to show that an outside pair of eyes can improve a group’s ability to find novel solutions to a problem. Business professors Marion Poetz, Nikolaus Franke and Martin Schreier write:
Over the course of years of studying innovation, we’ve found that there’s great power in bringing together people who work in fields that are different from one another yet that are analogous on a deep structural level. Such as makeup and surgical infections, surprisingly. Or inventory management and robot games. Or malls and mines.
Bringing in ideas from analogous fields turns out to be a potential source of radical innovation. When you’re working on a problem and you pool insights from analogous areas, you’re likely to get significantly greater novelty in the proposed solutions, for two reasons: People versed in analogous fields can draw on different pools of knowledge, and they’re not mentally constrained by existing, “known” solutions to the problem in the target field. The greater the distance between the problem and the analogous field, the greater the novelty of the solutions…
To get a sense of the value of accessing and implementing knowledge from analogous fields, consider our recent study in which we recruited hundreds of roofers, carpenters, and inline skaters to contribute their insights to the problem of workers’ reluctance to use safety gear because of discomfort…Each group was significantly better at thinking of novel solutions for the other fields than for its own.
In other words, a fresh pair of eyes doesn’t always help solve a problem, but it can think outside the box. This novelty of perspective is a potential benefit of trespassing that Ballantyne doesn’t even address.
A second potential upside to trespassing is that as in industrial competition, a bit of outside entry might make incumbents work harder and improve their game. For example, when the sabremetrician Nate Silver decided to venture into political forecasting — using original and bespoke models — his solid and well-publicized forecasting record motivated a wave of innovation and improvement in the space. He may or may not be the best in the business, but his aggressive competition against the entrenched community of political-forecasting experts has undoubtedly raised the standard of quality. Nowadays, of course, Silver is known primarily for his political forecasting (as Dawkins is now known mostly as a professional atheist), but when he started out he was unquestionably what Ballantyne would label a trespasser.
Another example would be economists’ forays into sociology; whatever deeper truths these forays uncovered, they undoubtedly motivated sociologists to upgrade their statistical methodology! And a third example is Google’s AlphaFold algorithm, which is leapfrogging existing research efforts and forcing academics to incorporate more A.I. into their methods.
A third potential benefit is the puncturing of socially driven consensus. Outsiders are usually not part of the social group that defines a community of expertise, which of course deprives them of lots of knowledge, but also frees them from much of the social need to conform to existing opinions. This conformity pressure is probably documented in some empirical study somewhere, but just to see how it works, check out Richard Feynman’s description of how experimental physicists fudged their value for the charge of the electron so as not to stray too far from existing results.
A modern example — with potentially significant health consequences — was the journalist Zeynep Tufecki’s successful effort to counter CDC and WHO guidance on the usefulness of masks in early days of the COVID-19 pandemic. To make a long story short, Tufekci was right and the epidemiological experts were wrong, as they later (to their great credit) admitted. Many people’s lives were probably saved by the efforts of Tufekci and other epistemic trespassers who strayed outside of their lanes on this issue.
This must be weighed, of course, against the panoply of cranks who downplayed the virus and spread antivax messages. No one disputes the harm those trespassers have done. But inveighing against epistemic trespassing in a blanket manner would seem to throw out the good along with the bad.
Now, Ballantyne’s essay does briefly grapple with a few ideas for criteria that could maybe be used to distinguish bad trespassing from good trespassing. For example, he talks about the possibility of general skills that can transfer from one domain to another (one of these could be “logical thinking”, which Ballantyne should probably hope is a general-use skill given that it’s philosophers’ main tool!). But apart from briefly describing one experiment on underwater shooting from the early 20th century, Ballantyne provides little evidence with which to assess whether, when, and how much different skills transfer between different pairs of domains. Presumably, assessing skill transferability is a complex task requiring a great deal of empirical research, and the results will differ greatly by field pairs and even by topic pairs. So I’ll stay in my lane on this one, except to note that this is a nontrivial question requiring nuanced thinking and thorough empirical work.
None of this is to say that Ballantyne’s postulated harms of trespassing are wrong (though most could stand to draw on a bit more supporting evidence from the fields of sociology, organizational behavior, and so on). But if you’re criticizing a behavior and demanding large-scale remedies, you should at least be evaluating the potential benefits in parallel with the potential costs!
To sum up, I’m highly suspicious of the practice of epistemic enclosure — of using social consensus to delineate the boundaries of fields of inquiry, and to assign each bounded region to a single club of human beings. This practice, to me, raises the specter of a gang of epistemic squatters claiming a piece of epistemic territory for themselves and using social censure to defend it against anyone who might point out the gang’s errors or force them to raise their game. As I see it, boundaries between fields of thought should be porous — the interdisciplinary research Ballantyne suggests is one way to cross those boundaries, but it shouldn’t be the only way.
Of course, I’m no expert.
Update: London School of Economics philosopher Liam Bright agrees, and has thoughts about the fundamental unity of science:
As a scientist I've spent my entire adult life campaigning for the importance of expertise. The importance of climate change, for one, motivated me, as well as the goofy letters from cranks we would get positing who new kids of physics.
Covid has knocked me back a solid 10% on this. The fields of public health and epidemiology seem to have put forward really poor efforts, from giving bad advice, to being unable to make even the very roughest projections or paint plausible pictures of how this would play out (remember test, trace, and extinguish the virus without vaccines? Then models that showed every human on the planet infected inside a month? Then flatten the curve? Then ethics experts telling us challenge trials, which would have saved hundreds of thousands of lives, were unethical? Then approving vaccines for 12 year olds but not for 11 year olds six months later? Then not allowing tests unless they were perfect, but slow? Then constantly deflating vaccines? Then fighting perpetual war on case loads post-vaccine?).
Over and over reasonably smart generalists with backgrounds in statistics routed them.
It's been a change in perspective for me for sure.
Epistemic Trespassing implies Epistemic Property implies Epistemic Rent-Seeking. Thank you for coming to my TED talk.