Statistics as Storytelling

Believe the science. The phrase seems innocuous on its face, the logic simple enough: Because science is a reliable means of acquiring empirical knowledge about the world, it is trustworthy. Skepticism, even more so outright disbelief, is therefore a sign of confusion, misinformation, lack of education, or malice. Bad actors sow doubt, and conspiracy theories flourish when science is viewed with this sort of suspicion. Hence the redoubled commitment to Believe the science! The imperative becomes a kind of societal duty, forming not only behavior, but identity.

Science is indeed worthy of trust. But as philosophers of science have long observed, and as the rest of us have experienced during the Covid pandemic, matters are not so simple as that. Science is a highly imperfect affair — not because it is perfidious, but because it is human. “Science” comprises a diverse set of inquiries and practices, which are premised on varied commitments, ordered to diffuse and perennially disputed ends. “Science,” in other words, is a verb, not a noun; it names the ongoing work that scientists do — a noble and brilliant bunch, certainly, but also a fractious and fallible one. Their purview, moreover, is restricted: They study what is detectable by our senses and by instruments built to extend them. Of what is not thus detectable, their expertise grants them not one whit of authority.

The account I have just offered understands science as a restrained enterprise, circumscribed in both methods and scope. But some see science as comprehensive. It can and does concern anything and everything; its subject matter is reality as such. Accordingly, if a topic or question or entity cannot be studied by science, this second view draws the natural conclusion: It must not exist. It must be fiction, deception, error, or gibberish. With certain significant ways of knowing and experiencing the world — art, drama, ritual, poetry, prayer, revelation — now removed from the equation, science, or rather Science, assumes magisterial authority. Thereby it is our only way to deliver true and certain knowledge of all we can ask or imagine. It is, in sum, a global way of knowing because it is the only way of knowing.

This second, far-reaching understanding of science is “scientism” — what Pepperdine professor of political science Jason Blakely, in his new book We Built Reality, defines as “an inflated, unwarranted confidence in the power of science to explain all of human life.” Ordinarily, when we hear “scientism” we think of the natural sciences, for example the belief that all reality at rock bottom is the domain of physics, or that human desires are nothing but a product of evolutionary biology. Blakely spies similar beliefs at work in the social sciences. More indictment than manifesto, the book lays out in persuasive detail how the social sciences have “infiltrated” our common life of politics and ethics.

The problem with scientism, as Blakely sees it, is not primarily the inflation of scientists’ egos or even of science’s influence. It is, rather, twofold: on the one hand, the colonization of our collective moral, social, and political imagination with an artificially narrow way of making sense of the world and of our lives therein; on the other hand, the unwitting power of social scientists not merely to describe but to remake the world in the image of their theories, goals, and desires.

We may picture scientism as the drift of the natural sciences’ methods and prestige to the social sciences. This drift makes non-scientific disciplines such as philosophy, ethics, literature, and history appear relativistic. And it creates incentives for intrinsically value-laden enterprises like psychology to “scientize” themselves, lest they too be left by the wayside. Consider Jesse Singal’s recent book The Quick Fix, which documents numerous examples of “fad psychology” in action: Despite questionable methodologies and non-replicable experiments, we constantly hear how “studies show” that science can, among other things, solve crime, make you happy, and cure racial bias.

“There’s only one problem with all this,” Blakely writes: “there are very good reasons to believe that unifying all our understanding of the world under the banner of the natural sciences is impossible. Like medieval alchemists trying to turn base metals into gold, such a task may be impossible, not due to a lack of ingenuity or intelligence on the part of those working on the goal, but because the structure of reality itself does not permit it.” In this case, the structure of reality is the nature of human beings.

Blakely opposes to scientism “the hermeneutic or interpretive outlook”:

Interpretive philosophy holds that achieving the unity of science is an impossible task because humans create and embody meanings in a way that requires the art of interpretation and not simply scientific explanation. In this view, human beings are fundamentally different within the order of things. As creative agents they continually spin new webs of meaning that form into practices, institutions, and the entire weave of social reality itself. Human social and political behavior does not fit under the conceptual logic of the natural sciences because it is not law-abiding or mechanistic in nature. The human sciences therefore have their own unique set of descriptive and explanatory concepts, and they are above all interpretive, humanistic disciplines, not formal, mechanistic ones.

The upshot: “no predictive science of human behavior is possible,” since human actions and beliefs are not fully determined by any set of conditions preceding them. Even Nate Silver agrees. His political and sports forecasts trade in probabilities in the aggregate, which means the equation bakes in the assumption that events sometimes turn out otherwise than as predicted. They were never judged impossible, only more or less unlikely.

How then can we understand human beings in their fullness, including their social reality? Blakely’s answer is simple: narrative. He doesn’t mean this in the pejorative sense, as in “don’t let the facts get in the way of a good narrative.” Rather, because human beings continually create new meanings and seek to make sense of things, “explaining human actions requires telling particular stories or narratives about what has happened.” Storytelling might sound subjective — and it is, inasmuch as it depends on perspective: a tale has a teller. But perspective does not rule out objective truth; it does not entail relativism, in the sense that a story is true only for the one telling it — “my” truth versus “yours.” Worries about the loss of neutrality or rationality in such an approach are therefore misplaced. If I recount to you what I saw when the truck ran the light and hit the minivan, my testimony isn’t called into question simply because it’s mine, much less because it lacks a God’s-eye view of the accident. Nor is it considered worthy of dismissal if I say, “The driver was wrong to do that,” though that is a value judgment, or “The driver is my uncle,” though that goes beyond a bare description of the event. Narrative interpretation and rational inquiry are not mutually exclusive; we can make rational sense of things only if there is a story holding them together.

In short, “stories are simply the only rational way to explain human actions. The central task of the social sciences is therefore to tell the most objective, nuanced, complex, and true story about a given slice of social reality.” To accomplish that in its fullness calls for hermeneutics, which is the art, not the science, of interpreting meaningful human action.

Blakely draws attention to three essential distinctions. The first is between scientism and the art of interpretation.

The second is between popular and elite discourse. Many of Blakely’s targets represent varieties of popularized social science — think Nate Silver, Ezra Klein, Malcolm Gladwell, Steven Pinker, and Tyler Cowen. Actually, some academics may well find themselves in agreement with Blakely’s criticisms and claim exemption from his accusation of scientism, believing that only opinion pages and trade books, not monographs and peer-reviewed journals, perpetrate the errors and exaggerations he lays out.

Not so fast, Blakely counters. The thesis of the book is that all social science — whether popular or elite — serves the same function. The distinction between popular and elite discourse is not much of a distinction at all. For however technical and purely descriptive social science imagines itself to be, it should be understood “as a way of exercising authority in society at large.” Moreover, the popularizers are often also members of the elite journalistic and academic establishment. All too often, the technical scholarship one is translating for a popular audience is one’s own. Such popularization, then — whether or not it ought to be seen as bastardization — is a feature, not a bug, of social science’s public performance.

That feature leads to Blakely’s third basic distinction. Marx wrote that the philosophers sought to understand the world, when the point was to change it. Social scientists, by comparison, claim only to describe the world, but they actually transform it. That they change the world is not a problem per se, particularly when they are explicit in their aims and modest in their claims. The problem is when they allege to be merely analyzing but their analysis is itself an agent of change. It is a self-deception that only compounds the force of the procedure.

Since the cloak of that deception envelops us all, Blakely suggests that the ability to see the dynamic at work requires a certain reading strategy, heterodox or subversive in character. This strategy interprets social-science literature “not as dry, technical tracts describing the world but as vividly cultural and ideological meanings that endeavor to transform the world.” Behind such an interpretation lies “a radical paradigm shift in how we read the genre of social science, not simply as an empirical exercise but as a form of cultural and ideological production.” For “social science is a veritable factory of meanings,” writes Blakely, wryly echoing John Calvin’s aphorism that the human mind is a perpetual factory of idols. In this way social science is “every bit as dynamic, poetical, and fecund as a creative writing department.”

Blakely employs a term of art scholars have used for this unsuspecting feedback loop, the way that a description in social science ends up changing the thing it is describing. This is the “double hermeneutic effect,” or “double-H effect” for short. Consider by contrast a mathematical theory about black holes. The theory does not interfere with the object of one’s study, for black holes do not respond to stimuli such as mathematical theories. But the objects of social science are human beings, who read and react to stimuli such as social science.

Take for example the so-called “clash of civilizations” idea. Proposed by Samuel Huntington in the 1990s, it suggests an intractable division and conflict between two entities: the West and Islam. These entities, according to Huntington, have objective status; however complex, they are coherent cultural realities whose inner logics stand opposed to each other. Blakely’s observation is that this vision is less a reporting of facts than a framework for interpretation and action. What Huntington proposes as description is an invitation to enact the classifications. And to enact them may prove to give them more life than they had prior to their description.

Social science, again, “is not simply descriptive but also performative.” Its white papers and working papers, its NBER abstracts and statistical regressions are nothing less than “artifacts of culture that participate in enacting and inaugurating certain political realities.” All told, it adds up to a vast laundering operation. It is not laundering money; behind the façade of scientific neutrality, it is laundering value.

On this understanding, the public performance of social science is what speech-act theorists call illocutionary: it is an utterance that “does what it says.” Consider phrases like “I promise,” or “You are now husband and wife,” or “Beware!” They do not just convey information, they perform an action. When they are spoken, the promise is thereby made, the marriage thereby begun, the hearer thereby warned. The word is itself a deed; it performs a task as much as it communicates content.

Blakely wants us to see the role of social science in public discourse in a similar way. Social science is a public performance not through rational persuasive power, much less by design, but by virtue of its undisputed social status and its infiltration into every corner of the culture.

This produces two ironies. The first is that the more we fall under social science’s sway, the less agency we have over against it. We may suppose that social science informs and supplements democratic deliberation, when it actually turns out to circumvent it, since the answers to disputed social and political questions are decided in advance.

From this follows the second irony. Wittgenstein warned against our bewitchment by language gone on holiday: When we use language that means nothing intelligible, we fall under a spell. For example, we speak of the human body and mind as hardware and software, then reverse the metaphor (forgetting it is a metaphor) and describe a computer as having, or being near to acquiring, artificial intelligence. Philosophy, then, is a kind of therapy — or counter-magic — to dispel the confusions wrought by our inadvertent misuse of language.

Blakely speaks in similar terms when describing our common life under the social scientific regime. The irony is that the very means of our growth in knowledge — the rationalization of social and political life — is a conjuring act. We are bewitched. And We Built Reality is an attempt to break the spell.

Blakely’s book canvasses a number of major topics in a short span: laissez-faire economics and the 2008 crash; the idea that neuroscience and genetics can explain people’s political preferences; the “rational violence” of the war on terror — to name only a few. He is not after exhaustive documentation. What he wants is to set forth enough examples, from sufficiently different slices of life, to convince the reader that the phenomenon he is describing is real, widespread, and detrimental to our common life.

The simplest case is the most extreme: the neo-phrenology hawked by theorists of race, intelligence, and criminality. What claims to be patient and painstaking descriptive analysis of a preexisting social reality — in this case, statistical correlations linking certain racial groups to specific behaviors, capacities, and life outcomes — functions instead as a plausible social framework for individuals, institutions, and the culture as a whole to adopt or inhabit. It proposes, in other words, a way of seeing the world and interacting with one’s neighbors, decked out in the authority of stats and grids, graphs and charts. Although the purveyors of this “analysis” retain plausible deniability through various distancing mechanisms, no such fine-grained distinctions remain once the theory touches terra firma.

Blakely wonders: “What if the central problem of superstitious forms of pop social science was not that they were insufficiently scientific, but rather that they were too literal in their reading of the original imaginative suggestions and metaphors?” What if, when black men are profiled on the street when minding their own business or receive harsher sentences than white men for committing the same crimes, this is not a misapplication of high theory, but its logical extension?

Borrowing from sociologist Aaron Panofsky, Blakely terms this practice “astrological genetics.” Instead of reading our destinies in the stars, racial “science” reads them in our bodies. The shape of our skulls, the color of our skin, the land of our ancestors tells us all we need to know — or at least a good deal — about our behaviors, our beliefs, even our fates. This biologization of culture codes the superficial as the significant, and societal catechesis trains us to decipher these codes through complex legends, spoken and unspoken. The result: “Race constitutes a massive case of a double-H effect, in which social scientific theories and a culture of scientism penetrate into social reality, suggesting that we read certain traits as races.” Race is a story that gives us meaning, providing grammar and syntax for a particular way of interpreting the world. The social science telling this story may thus “be read as a kind of repressed imaginative literature in which new identities are continually dreamt up.” Its authors, then, “did not unearth the world but helped to contrive it.”

Or consider a seemingly more benign example. The 2008 book Nudge, by Richard Thaler and Cass Sunstein, describes the “choice architecture” in which we live and move and have our being, and the various incentives for behaviors built, intentionally or not, into that architecture. Think of having to “opt in” instead of “opt out” of being an organ donor. You are less likely to opt out if you’re already a donor than to opt in if you’re not a donor. Default settings set an unchosen status quo. Since most of the time we leave the settings at default, the status quo remains. Thus can we be “nudged” to do, or at least to accept, what we never consciously chose.

This state of affairs opens up a whole world of power and influence. If our choice architecture is so influential on our daily behaviors, then even the tiniest of variables can generate enormously different outcomes. Moreover: If an architecture, then an architect.

Naturally, the ideal choice architect for the masses, mired in their biases and unreflective habits as they regrettably are, is the duo of Thaler and Sunstein, together with people like them. Blakely is unsparing in his condemnation of this parody of self-appointed oligarchy. “Thaler and Sunstein’s almost unflagging sunniness” works to mask the book’s central proposal, namely “that social scientific elites benevolently modulate and engineer the behavior of others.” Whereas ordinary people are unable to recognize or overcome the nudges and reflexes that dominate their lives, “a technocratic elite is able to momentarily overcome these limitations for the sake of constructing rational, scientific public policy that promotes social harmony.” In other words, this “privileged ability to exit the automatic mind” contains, or trades on, “an implied form of class hierarchy.”

This hierarchy is one in which well-meaning liberals are on top. Democracy, in this scheme, amounts to little more than occasional “automatic” voting — all while inequality grows, working people try to make ends meet, and choice architects draw up blueprints, luxuriating in their freedom from bias. Blakely argues that Nudge “could be read as a modern version of Plato’s city–soul analogy: just as the human mind has a hierarchical cognitive architecture, so too is society ruled by a cognitive elite.” Given that Sunstein went on to join the Obama administration after the book’s publication, it might also “be read as a looping double-H effect, justifying and enacting rule by Ivy League technocrats.”

What galls in this example is not Thaler and Sunstein’s concepts (nudge, choice architecture) or even their supposition that they are uniquely qualified to nudge us into utopia. Perhaps they are. No, what galls is that they do not know what they are doing; or, if they do, that they do not admit it. They project themselves as enlightened democrats eager to help as best they can, when in fact they are elites atop the class hierarchy, writing for fellow elites — hence that imperious, impermeable “we” — deigning to stoop low in benevolence for the unwashed masses. And this benevolence reinscribes the very hierarchy on which their status depends. This is pure ideology, dressed up in the trappings of scientific authority.

In “After Jericho,” the Welsh poet R.S. Thomas writes:

There is an aggression of fact
to be resisted successfully
only in verse, that fights language
with its own tools. Smile, poet,

among the ruins of a vocabulary
you blew your trumpet against.
It was a conscript army; your words,
every one of them, are volunteers.

The walls of Jericho, for Blakely, are the premises and public performance of scientism. He has nothing to stand against their “aggression of fact” except words. But words are powerful. Though he is not a poet, he counts poets among his ranks; or rather, he seeks to represent them along with all those who belong to the classical liberal arts. We Built Reality thus concludes with a call for the recovery of the humanities as the only bulwark against, and alternative to, scientism run amok. Only in a culture in which all ways of knowing, both the poetic and the scientific, are reintegrated into a single whole will we be able to escape “the current delusions and disappointments of our reigning scientism.”

Unfortunately, “the fear of scientism and the abuse of scientific authority can sometimes lead to an even stronger attachment to the notion that scientific authority is a kind of omniscient savior.” Rightly troubled by groundless suspicion of legitimate science, we react by overestimating rather than properly qualifying the scope of scientific reason. What we need is not More Science! but something else entirely: the virtue of prudence.

Prudence names the sort of practical wisdom that informs discretion in judgment; it incorporates the facts, as well as the whole narrative context that gives those facts meaning, into a single perspective that aims to understand and to serve the persons, goods, and ends in view. Neither the hard nor the social sciences can offer us prudence, although prudence is key in their practice. Where to find it, then?

Blakely rightly points us to the humanities. Yet while a return to humanistic inquiry and understanding may be necessary, the journey there seems an unlikely one. To acquire prudence we must be schooled in it, and our institutions of humanistic education are barely keeping the doors open, much less filled with confidence about what they are for. Perhaps amid the wreckage of a scientized culture they will find it again.