A conspiracy theorist dies and goes to heaven. After entering the Pearly Gates, he comes face to face with God, who embraces him and says,
“Welcome! As a reward for your good deeds on earth, I offer you this gift: the true answer to any question of your choice.”
The conspiracy theorist thinks for a moment and asks, “Who shot JFK?”
“Ah, a fine question,” God replies. “It was Lee Harvey Oswald, acting alone.”
“Incredible,” says the theorist. “This goes higher than I thought…”
You may have heard of cognitively closed systems—various isms that trap their adherents in an erroneous web of belief, systematically obscuring the path back to truth. Think of Plato’s Cave, Wittgenstein’s fly-bottle, or Hobbes’s bird stuck indoors, “fluttering at false light” (i.e. banging itself against the window rather than flying back out the chimney).
The philosopher/logician/mystic Raymond Smullyan gives a number of more modern (merely potential) examples, which I repeat here without endorsement: the Freudian who views every objection to psychoanalysis as the product of repressed desire; the Marxist who spurns all counterevidence as bourgeois ideology; the feminist who dismisses all critics as agents of the patriarchy; the solipsist who insists that his detractors are figments of his own imagination; and so on. To this, we could add the centrist whose critics are all “tribalists,” the capitalist whose critics are just “envious,” and the male incel whose critics are all “NPCs.”
In each case, you can see how the closed system (supposing it really is closed) works to insulate itself against legitimate critiques.
In “Problems in the Theory of Ideology”—a prescient paper from 2001—Joseph Heath introduces a fascinating spin on the closed system: the self-radicalizing theory. The basic idea is that some theories tend to degenerate into more extreme versions of themselves over time, because they nudge their proponents to respond to valid counterevidence not by abandoning their theories, but by doubling, tripling, and eventually n-tupling down. The self-radicalizer ends up like the conspiracy theorist who, after being refuted by almighty God, concludes that even He must be in on it.
Heath’s main example of the self-radicalizer is the “ideology critic,” a kind of social theorist who thinks of oppression as the result of a widespread cognitive error, typically blamed on the oppressors. The ideology critic looks around at the social world and sees that certain individuals—the working class, women, etc.—are collectively failing to promote their own interests; rather than locking arms in resistance, they’re mostly submitting to or even reinforcing an oppressive social system. What’s going on? “Why aren’t these people acting rationally?” wonders the critic. “The workers of the world should be uniting to overthrow the bourgeoisie, just as women should be coming together to overthrow the patriarchy!” The fact that this isn’t already happening cries out for explanation. And so the ideology critic explains the persistence of oppression as the result of a systematic belief in a pernicious “ideology”—some theory that deludes or tricks the oppressed class into reproducing the conditions of their own oppression. Think: “a woman’s status should be determined by her appearance,” or “rich people earn their riches through hard work and thriftiness.”
The problem with ideology critique, says Heath, is that it assumes cooperation is easy. The critic supposes that rational people generally would work together to promote their shared interests; so, when the proletariat (or whoever) fails to do so, we infer that they must be suffering from some kind of cognitive illusion—hence, ideology.
But cooperation isn’t easy! It’s perfectly normal for people to fail to promote their shared interests, even if everyone involved is acting rationally. This is called a “collective action problem,” and it can arise for a variety of well-known reasons, such as:
Misaligned incentives — often the socially beneficial action, like fighting for reform or refraining from pollution, is prohibitively costly for the individual who does it (the benefits, rather than being “internalized,” flow to “free riders”).
Lack of trust — some actions, like joining a revolution or switching industry standards, can be disastrous unless most people do it together; if no one has reason to trust that others will join, no one will risk going it alone.
Adapted preferences — poor/oppressed people may have preferences that are “adapted” to their low expectations, which can cause frustration when trying to “break out” of social roles, and which can be a disadvantage when bargaining with the rich/powerful.
It bears emphasizing that none of these explanations imply that the individuals involved are dumb or delusional. If anything, they’re victims of their own rationality! Cooperation would be easier if we all superstitiously assumed that we’d each be rewarded for pro-social actions, as if by moral magic.1
No wonder, then, that “ideology critique” has often led to disappointment. The workers still haven’t flocked to the ramparts even after decades of critical theory conferences; the gender pay gap still persists even after concepts like “patriarchy” have become so popular as to be the stuff of Hollywood cliché. The main prediction of ideology critique appears to have been refuted by experience. We have arrived at the moment when God pins the blame on Oswald.
But like the stubborn conspiracy theorist, the ideology critics would rather stick to their explanatory guns. Instead of looking for some other explanation beyond pure ideology, the critic instead concludes that they must have underestimated ideology, which turns out to be more pervasive and well-hidden than anyone expected. This goes higher than I thought.
Thus begins the “vicious cycle of theoretical self-radicalization.” Here’s how Heath describes the process:
The more serious problem for critical theory arises as follows: after having presented the criticism, and having it widely accepted, the critic expects to see some kind of social change. When none is forthcoming, the critic begins to suspect, not that there is a practical problem preventing implementation of the desired improvements, but that the criticism itself was too superficial, that it didn’t get to the root of the problem. The ideology must be more pervasive than originally suspected. Perhaps the original criticism was insufficiently radical, because it used concepts that were in general circulation, and hence complicit in the ideological system. The solution may be to deconstruct these concepts, and form an entirely new set.
Once this line of thinking has been engaged, the critical theory becomes increasingly baroque, increasingly obscure, and of course, increasingly unlikely to change anything. This can generate a vicious cycle of theoretical self-radicalization, in which critics respond to the increasing irrelevance of their theories by further radicalizing them, making the entire apparatus more and more remote from the concerns and the vocabulary of everyday life. (p. 188)
If I may add own objection: as self-radicalizing theories grow more elaborate, they also tend to become more ambiguous, with the result that critics will increasingly be unable to agree about the content of the ideology they’re critiquing. Does patriarchy demand that women dress modestly (thus controlling their personal choices), or does it encourage women to wear revealing clothes (thus catering to the male gaze)? If there’s no clear answer, then citing patriarchal ideology isn’t going to do much to explain real-world outcomes.
So, whether you’re worried about women’s rights or workers’ wellbeing, Heath thinks you should at least eventually try to analyze the problem as a normal failure of collective action rather than an overcomplicated triumph of ideology. Think of Claudia Goldin’s Nobel-winning work on the gender pay gap, or Daron Acemoglu and James Robinson’s Nobel-winning critique of “extractive institutions.” It’s the 21st century: we’re living in a golden age of empirical methods and sitting on a gold mine of wisdom about collective action problems (courtesy of 20th-century social science). There’s really no need to theoretically radicalize yourself into irrelevance.
But enough prelude! I’m not here to defend Heath’s (admittedly interesting) view of ideology. I’m more interested in his concept of self-radicalization, which is fascinating in its own right, and which applies across a wide range of intellectual contexts. The key is to notice that Heath’s giving an informal example of what scientists call a dynamical explanation. If you wanted to make it a bit more formal, you might put it like this.
We assume that people choose from a variety of different belief systems, but they change their beliefs only gradually, as they try to remove tensions and contradictions; people don’t just spontaneously convert to a completely different worldview. The space of all possible belief systems is called a “configuration space,” and the amount of tension in a system is its level of “energy.” The key methodological assumption is that people will change their beliefs iff there’s a sufficiently close belief system with lower energy. Simplifying (because I suck at thinking in 4+ dimensions), we can draw the configuration space as a 2D grid, which allows us to plot the energy levels as a curvy shape floating overhead, like so:
Now we have a beautiful, simple model of how a person’s beliefs develop over time. Like a marble rolling to the bottom of a bowl, a person’s belief system continuously develops towards the local minimum of tension and dissonance. The problem, as you can see in the above diagram, is that the local minimum isn’t always the global minimum. A believer might end up trapped—there’s a better location on the other side of the configuration space, but there’s no way to get there without “rolling uphill.”
This is how I think of the self-radicalizer: they’re stuck in a bad local minimum. Each time they resolve a tension in their worldview by making themselves more extreme, they are tragically moving away from the best possible viewpoint and towards a far more mediocre equilibrium that just happens to be closer to where they started. The only way out is to somehow shake things up—think of Brian Eno’s Oblique Strategies, cards designed to jolt you out of a cognitive rut.
Where else do we see self-radicalization? Some examples:
Echo chambers — on some difficult topics, we have to rely on experts to figure out what’s going on. But as Thi Nguyen warns, “our inexpertise may lead us to pick out bad experts, which will simply reinforce our mistaken beliefs and sensibilities.” We thus end up in a “runaway echo chamber,” where we distrust the very sources most able to correct us.
Cults — yeah, I’m linking to Thi Nguyen twice in a row, but it’s not my fault he’s interesting. As Nguyen puts it, an echo chamber can be like a cult, which “isolates its members by actively alienating them from any outside sources. Those outside are actively labelled as malign and untrustworthy. A cult member’s trust is narrowed, aimed with laser-like focus on certain insider voices.” Distrust of outsiders then creates “a social buffer against any attempts to extract the indoctrinated person from the cult.” Those who dip their toes into the cult quickly end up completely submerged.
Conspiracy theories — not to repeat myself, but often a conspiracy theory will include some suspiciously convenient tenet designed to discredit nonbelievers. (Say it with me: This goes higher than I thought.) See also: “The Devil put the fossils in the ground to deceive you,” which is a thing I was taught at my religious afterschool program in 4th grade. (Really.)
Bigotry — a bigot believes that their in-group is superior to some out-group, which naturally inclines the bigot towards in-group socializing and out-group distrust. This tends to insulate them from disconfirming evidence and sweep them up in group-polarized deliberations. (The group may even evolve metanorms, so that members are punished if they don’t enforce in-group solidarity—Axelrod gives the example of white people who were violently attacked by other whites for refusing to participate in a lynch mob.) Thus someone who starts out with mild in-group preferences can end up self-radicalizing into a hardcore racist, sexist, anti-semite, etc., utterly detached from reality.
Something that all of these worldviews have in common is that they’re Manichean. Once you buy into them, you begin to distrust those who lack the view (since they’re bad, evil outsiders). Then you buy into the Manichean worldview even more, which means you distrust the outsiders even more, and so on.
The possibility of self-radicalization seems to me a structural danger in any Manichean view. I hesitate to call it a defect, however, since it’s possible that some Manichean view might actually be true, in which case it’s good if that particular view is self-radicalizing.
Consider our own (good) beliefs about the law of induction, according to which the future will tend to be like the past. We assume that the laws of physics that worked up until today will continue working tomorrow; we don’t expect things to randomly change in never-before-seen patterns. Given that we believe in induction, the continued success of inductive reasoning seems like ever more evidence in its favor; our inductive worldview is self-radicalizing. But imagine we’d instead started out believing in counter-induction, according to which the future tends to be unlike the past. Then we would’ve ended up in a quite different equilibrium, as in the old fable.
A philosopher travels to a distant village where everyone reasons by counter-induction. Every night, their gurus predict that the sun will not rise tomorrow, and every morning, they suffer a humiliating refutation.
One day the philosopher can’t help but ask the elder guru, “Why do you keep using counter-induction? Haven’t you noticed it’s failed you every day?”
“Of course!” replies the guru. “That’s why it’s bound to work tomorrow.”
One last potential upside. In general, self-radicalizing does seem pretty bad. But in certain contexts, even if it’s individually costly, I wonder if it might be socially beneficial.
This seems to be what Ian Shaprio thinks of Jeremy Bentham’s version of utilitarianism, the view that we should always do whatever maximizes global happiness. Bentham’s view is infamously extreme—no special treatment for loved ones; no natural rights; etc.—but it’s also beautifully pure. Because Bentham pursues his core idea to its logical conclusions, rather than adulterating it in hopes of arriving at more respectable conclusions, Bentham teaches us something deep about what that core idea really contains, even if the contents aren’t ultimately correct.
Here’s Shapiro:
There are some thinkers in the Western tradition—I guess in any tradition—who have a particular characteristic that Bentham certainly has, and I think…Karl Marx had and Robert Nozick had. And the thing I’m thinking of here is they are the kind of person who takes one idea to the most extreme possible formulation. They ask themselves a question, “How would the world be if this idea that I have is the only important idea?” and they take it to its logical extreme, to an excessive kind of formulation.
And they will go places with their idea that nobody else will go, and so that makes them a little bit crazy. They’re monomaniacal, obsessively consumed with their idea. …
But what’s always interesting about people like this is that they play out an idea to its logical extreme, and that exhibits both its strengths and its limitations, just because they’re willing to go when others will not go—think the unthinkable, think politically incorrect things for their time—in pursuit of really pushing this idea to the absolute hilt.
And so Bentham is the kind of thinker who I suspect, at the end of the day, nobody will be fully convinced by, but he’s very useful. He’s a very useful diagnostician of what it is about utilitarianism that’s going to be appealing to you, and where eventually you’re going to want to put some limits on it, just because he goes beyond the limits.
The point isn’t that Bentham’s pure view is more plausible than a more moderate version of utilitarianism—such as a “consequentialized” hybrid. While I do like that point, Shapiro’s saying something else: even granting that Bentham’s extreme view is less plausible, it might be more revealing. Bentham pays the price of implausibility so that we can enjoy the benefit of understanding what True Utilitarianism looks like. In terms of our dynamical model, a monomaniac like Bentham can be seen as an intrepid explorer of the configuration space.
I may not always follow in your footsteps, intrepid explorers. But I do salute you.
There are also plenty of respectable worldviews that grease the wheels of cooperation through a kind of cosmic, pro-social ideology. According to some religions, for example, the afterlife will sort out the misalignment of mundane incentives: you are rewarded (or punished) after death in proportion to the good (or evil) you did beforehand.
Compare this to the much sillier Aristotelian view that, whenever we act virtuously, that is in our best earthly interests. (The soldier who dies for his country, says Aristotle, is better off dying in a blaze of glory than retreating in shame.) You can see hints of this view in the “prosperity gospel,” according to which God rewards believers with blessings during their worldly lives. This is a particularly interesting doctrine because it makes empirical predictions that should be amenable to econometric confirmation. Get on it, economists!
(Related fact: Robert Barro and Rachel McCleary have a classic paper about religious belief and economic growth. Their findings suggest that belief in heaven and hell causes more economic growth, but attending church has the opposite effect!)
On your dynamical system metaphor - the way I was thinking of it is not that the self-radicalizing theory is in a local minimum, but rather that they found an infinite well in the energy space! The truth isn’t necessarily lowest energy (particularly given imperfect information). The energy being minimized is at best an imperfect guide to truth.
I think the thinkers you’re describing at the end, like Bentham, are actually just minimizing a different function than everyone else, and that gets them over the energy barriers we find, whether or not it gets them to the truth. (And this is why, as my first post said, experts are usually wrong - but interesting!)
But consider people at the local minimum that coincides with the best global minimum. Don’t they typically get and stay there through the same process being described as “self-radicalization”? And don’t they typically explain why other people locate themselves at other minima by hypothesizing cognitive errors based on group dynamics and self-interest—in other words, “ideology”?