Turning Over Explanatory Inversions

On Finding Our Epistemic Bearings

Photo by Quino Al on Unsplash

This article from

Dan Williams

elicited some feverish buzz and endorsements in my corner of Substack recently. Williams’ thesis is that widespread pessimistic attitudes towards social epistemology (like all the chatter about misinformation) deserve to be turned upside-down in an “explanatory inversion.” He uses the issue of poverty to illustrate the idea: poverty is an enduring problem that folks often seek to assuage by locating treatable underlying causes, and the more conspiratorial species of leftist is liable to attribute poverty to the machinations of evildoing capitalists or whatever. But that approach is backward in a crucial way: it too readily adopts a mindset that poverty is an elusive mystery, and the economically literate intelligentsia enjoys pointing out that wealth—rather than poverty—is the true historical aberration and non-obvious phenomenon. Poverty is easily achievable; wealth isn’t. In fact, here’s

Noah Smith

doing this exact spiel just recently:

To ask why some societies in the world are still poor is the wrong question. Poverty is the default condition, not just of humanity but of the entire Universe. If humanity simply doesn’t build anything — farms, granaries, houses, water treatment systems, electric power stations — we will exist at the level of wild animals. This is simply physics.

Williams recommends a similar inversion for social epistemology, arguing that a scarcity of reliable information is a similar civilizational default—not a novel feature of exotic modern trends in culture/technology.

He relies on a couple of criteria for justifying an explanatory inversion, like (1) how commonplace the explanandum is and (2) whether the explanandum occurs naturally. For instance, Williams argues that poverty is both historically common and reflective of the state of nature, whereas wealth isn’t. Similarly, he characterizes social cooperation as a surprising triumph over natural inclinations. Analogously, accurate belief formation is the real unexpected deviation from normality, not the adoption of fallacious beliefs or vulnerability to stuff like biases, disinformation, and superstition—that stuff saturates everyday life. Accruing inaccurate views is straightforward and rampant; cultivating accurate views is challenging and atypical, so just as economic wealth is more puzzling than economic poverty, epistemic wealth is more puzzling than epistemic poverty.

What are we really doing by highlighting the enigma of wealth, though? I cosign the thinking that there’s something eye-opening and radical about realizing that wealth is more puzzling than penury, but some of the proposed rationales/features of these “explanatory inversions” merit closer scrutiny. Firstly, the appeal to rarity as the benchmark for what deserves explanation seems mistaken. Gravity is ubiquitous, but that doesn’t make Newtonian physics less scientifically prize-worthy—if anything, the opposite is more true: something being commonplace makes its want of explanation more pressing. Moreover, Einstein was able to explain special deviations from Newtonian mechanics, but does that make Newton’s efforts poorly targeted?

Williams also concedes that poverty might seem unusual from an extremely provincial viewpoint inside a narrow region of history, but zooming out geographically and temporally reveals its overwhelming prevalence, and he uses a famous graph of historical global GDP to emphasize the novelty of wealth. But if society is wealthy for long enough into the future, would that transform this explanatory inversion into a mistake and warrant re-inversion? Or if a time traveler reported longstanding future societal affluence, would the entire project of studying wealth generation suddenly be wrongheaded? The criterion of naturalness is equally suspect. Nothing is more natural (and common) than evolution by natural selection, but few insights have been more profound than Darwin’s theory, and we don’t accuse him of getting everything backwards.¹

Williams is surely correct that the path toward reliable knowledge is fraught with pitfalls, and without fastidious knowledge acquisition procedures and assiduous refinement of evidence/reasoning, truth-seeking is liable to be derailed. But does a plenitude of epistemic hazards really unveil that ignorance is some kind of absolute default? Consider Newtonian physics again: friction incessantly disrupts the tendency of objects in motion to remain in motion, but accepting friction as the default of physics was a scientific blunder. Newton understood that mechanics had an underlying preference for inertia, and if you want to coax more inertial behavior from some object, you’d want to eliminate sources of that friction.

Even this fashionable opinion about wealth being a spurious default risks venturing into the sophomoric, and these explanatory inversions should be handled cautiously. These ideas can easily be overextended to justify unending governmental tinkering and overbearing regulatory conditions. Even the top-down command-and-control strategies of communism are tempting if you operate within a framework that presumes wealth creation has to be actively fostered to elude the tractor beam of poverty. While it’s true that Adam Smith framed his economic studies as an effort to demystify how nations managed to become wealthier, it’s also true that he described market economics as sorting resources efficiently in an unguided (read: default) manner, as if by an invisible hand.

So, the proposed criteria for judging something’s eligibility for an explanatory inversion are misidentified, and we should be more careful about extrapolating from these inversions even when properly diagnosed, but Williams still seemingly has a point—what’s going on here? And what is the upshot of a successful explanatory inversion? I propose slightly amending the concept. Maybe being commonplace or natural or whatever supplies a partial, commonsense explanation for X, or partly explains X in terms of ~X, rather than proves that X doesn’t require an explanation at all. Maybe an explanatory inversion doesn’t entail swapping out explananda so much as it readjusts prevailing assumptions about the difficulty of explaining competing features of the world.

Unfortunately, further problems arise from Williams’ unsteady application of explanatory inversions to social epistemology. In short, he can’t seem to decide if we’re blessed or screwed and appears completely unreceptive to the probable answer of “both.” First, he rightly insists that our modern epistemic bounty is the outgrowth of “the scientific revolution, the Enlightenment, and centuries of cultural and institutional development designed to overcome the many sources of ignorance and misperception in human judgement.” Moreover, he argues, whatever epistemic prowess we enjoy is contingent on inculcating delicate truth-seeking norms:

Many educated people in affluent, liberal societies take this norm for granted today. Even though motivated reasoning is widespread, people are typically embarrassed by it. They think letting self-interest or tribalism bias judgement is shameful and strive to present themselves as disinterested, rational truth-seekers.

This attitude and its associated social norms are neither universal nor the default way people treat their deepest convictions. For most people in most places, there is little embarrassment in the fact that their worldview, religion, or ideology are designed not for truth but for things like identity formation, celebrating the glory of their tribe, achieving cooperation, and demonising enemies.

When Enlightenment philosophers celebrated reason and the rational pursuit of knowledge (“Sapere aude!”), they were calling for a profound—a radical—norm change.

Later on, however, Williams outlines his takeaways from all this explanation-inverting, and they’re total non sequitors, concluding that he dislikes popular usage of terms like “misinformation” and “post-truth” because, among other reasons, “the modern panic about these things is historically illiterate. There never was a ‘truth’ era.”

But if we enjoy comparative epistemic abundance thanks to peculiar norms, why does Williams insist that the notion of a truth era is nonsensical? I’m baffled by his claiming that both (1) our epistemic well-being is contingent on fragile norms and institutions, and (2) worrying about deviation from those norms/institutions is historically illiterate. Shouldn’t we rigidly police any suspension of those precious norms? If our epistemic progress is so precarious, why is he steadfastly discouraging any worries about backsliding? And doesn’t the moral panic about deteriorating epistemic norms cut directly against the claim that we take our epistemic circumstances for granted? This outlook is utterly confused.

Ultimately, the notion of an explanatory inversion, even if applied without error, is outstretched by Williams’ various conclusions. Again, I’m not dismissing his point in its entirety. I agree that epistemic virtue is a scarce and precious resource, but I’m puzzled how that deflates either the fears that we’re losing ground or the urgency with which we ought to try correcting any slippage. The explanatory inversion of poverty doesn’t undercut the worthiness of trying to alleviate poverty, nor does it render any concern about an uptick in poverty absurd. You wouldn’t dismiss worries about declining household wealth by reassuring everyone that poverty is historically commonplace. A similar explanatory inversion for social epistemology likewise fails to achieve this. We should cherish any positive epistemic norms we’ve cultivated, and we should fight tooth and nail to protect them, even if it’s unsurprising that we often fall short.

1. N.B., Daniel Dennett, alluding to a critique of Darwin from 1868, used the phrase “strange inversion of reasoning,” to characterize the insights of Darwin and Turing.