The boy rode to school through Mexico City traffic while his father talked. Charles Stephens (b. 1937) taught history at the University of the Americas and used the drive to rehearse his lectures. Munich. Yalta. The show trials. The gulag. The passenger was his son Bret (b. November 21, 1973), who later said he received his education in twentieth-century politics in that car, years before any school assigned it. The father spoke of the century as a war between civilization and totalitarianism. The son absorbed the frame and never put it down.
Almost everything in Stephens's career runs back to that inheritance. He became a Pulitzer Prize-winning columnist, editor-in-chief of The Jerusalem Post at twenty-eight, deputy editor of The Wall Street Journal editorial page, the house conservative of The New York Times, and editor of Sapir, a quarterly on American Jewish life. He also became a man without a party. The Republican Party he joined as a young writer treated American power as the guarantor of a liberal world order. The Republican Party of Donald Trump (b. 1946) treats that order as a racket run at America's expense. Stephens spent the decade after 2016 arguing, in the pages of a newspaper whose readers mostly despise his economics, that the tradition which formed him deserved to survive.
The family war
The family history reads like a syllabus for his politics. His paternal grandfather was born Louis Ehrlich in Kishinev in 1901. The family left the Russian Empire after the pogroms and settled in New York, where Louis eventually took the surname Stephens, by family account after the Irish writer James Stephens. He moved to Mexico City, founded a chemical company called General Products, and prospered. He started as a New Deal Democrat. After World War II he became a fierce anti-Communist.
His wife, Annette Margolis, was a New York-born painter who had studied at Hunter College and Columbia. In post-revolutionary Mexico she moved in the circle of Diego Rivera (1886-1957) and Frida Kahlo (1907-1954). The marriage ended, and Annette married the American composer Conlon Nancarrow (1912-1997), a veteran of the Abraham Lincoln Brigade in the Spanish Civil War and, in his stepgrandson's description, a devout Stalinist.
So Charles Stephens grew up inside the Cold War at his own dinner table: a father who hated Communism, a stepfather who served it. He chose his father's side and kept choosing it. He worked as a foreign-policy speechwriter for Senator George Smathers (1913-2007), the Florida Democrat, and briefly dated an office secretary named Mary Jo Kopechne (1940-1969). As the Democratic Party moved left, he moved right. In 1974, the year of Watergate, he ran for Congress as a Republican on continued military aid to South Vietnam. He lost, returned to Mexico, and helped run the family business.
Bret's mother, Xenia, was born in Milan in 1940 to a Jewish family with roots in Vilna. Her family reached the United States in 1950 after a decade of European upheaval. The household Bret grew up in was secular, Spanish-speaking, prosperous, and marked on both sides by flight. Jewish vulnerability was a family memory before it was a political position.
Charles Stephens found intellectual companionship in Commentary, the monthly Norman Podhoretz (1930-2025) had turned into the central organ of neoconservatism. In his liberal social world, Charles held minority opinions, and Podhoretz's essays told him those opinions belonged to a tradition. Decades later, going through his late father's effects, Bret found an issue of Commentary from the 1980s with a Podhoretz cover story titled "The Hate That Dare Not Speak Its Name." Stephens had used the phrase in a foreword of his own, believing it his. The line had planted itself in his memory in childhood and waited there. He apologized to Podhoretz in print and called it an honest mistake. It was also a measure of how far down the magazine went in him.
Chicago
At about thirteen, Stephens decided to finish his education in the United States. He attended Middlesex School in Concord, Massachusetts, graduating in 1991, and later returned as a trustee and as a parent of Middlesex students. He then entered the University of Chicago through the Fundamentals program, a great-books course of study created by Leon Kass (b. 1939), Amy Kass (1940-2015), Allan Bloom (1930-1992), and James Redfield, and took his degree in 1995.
Fundamentals organized education around permanent questions rather than a department. Stephens read Aristotle, Locke, Smith, Tocqueville, Arendt, Hayek, and Miłosz, and argued about them in seminar rooms where argument was the point. He has credited the Kasses above all. From Chicago he took a definition of intellectual humility that he repeated for the rest of his career: intelligent people of good faith can study the same question and reach opposite conclusions, and a serious community is built through that disagreement. In his sophomore year he read Sick Societies, the anthropologist Robert Edgerton's attack on cultural relativism, and on a whim wrote about it for Commentary. The magazine that had steadied his father now published the son. After Chicago came a master's in comparative politics at the London School of Economics, then a job at Commentary under the editors who had shaped his household's politics from three thousand miles away.
Jerusalem at twenty-eight
Stephens joined The Wall Street Journal in 1998 as an op-ed editor in New York, then moved to Brussels to write editorials on the European Union. The Journal editorial page of that era stood for free trade, low taxes, immigration, and American power, and it treated those commitments as a package. Stephens absorbed the package. Then came September 11, and in January 2002 he left to become editor-in-chief of The Jerusalem Post, the youngest in the paper's history.
Picture the newsroom he walked into. The Second Intifada was at its height. Buses and cafés were exploding in the city where the paper's reporters lived. The staff included veterans who had covered Israeli politics for decades, and their new boss was a twenty-eight-year-old American whose management experience consisted of editing opinion pages. Some of them regarded him the way career diplomats regard a donor ambassador. He ran the paper anyway: news, editorial, foreign editions, the works. He redesigned it, pushed its editorial line toward the conservative and internationalist wing of Israeli opinion, and treated its English-language readership as a global audience rather than a local one.
The claim, sometimes repeated, that he rescued the paper financially has no public documentation. What Jerusalem gave him was harder to audit and more durable. From the editor's chair, terrorism did not look like a symptom of poverty awaiting a development program. It looked like a strategy, chosen by movements whose doctrines rejected compromise and whose targets were civilians drinking coffee. Stephens generalized from this, and the generalization carried both insight and risk. He understood earlier than many that Hamas, Hezbollah, and al-Qaeda meant what they said. He also acquired a habit of sliding from the pathologies of movements to sweeping judgments about the cultures around them, a habit that would cost him.
Global View
He returned to the Journal in late 2004, became deputy editorial-page editor with responsibility for international opinion, joined the editorial board, and wrote the weekly "Global View" column. He was no freelance contrarian; he helped write the official voice of the most influential editorial page in American conservatism. He supported the 2003 invasion of Iraq and accepted its premises: that Saddam Hussein (1937-2006) was a gathering danger and that his removal could open the region to something better. He later conceded he had been too optimistic about America's power to democratize another society. He never accepted that the invasion had been wrong in principle. The revised position, worked out over years, held that the United States must act as the world's policeman while abandoning the ambition to be its priest: enforce order, punish aggression, and give up on redeeming foreign societies through occupation.
That argument became America in Retreat: The New Isolationism and the Coming Global Disorder (2014), aimed at Barack Obama (b. 1961). Stephens argued that American withdrawal is also a policy, with victims, and that vacuums get filled by Russia, Iran, and jihadists. Events in Crimea and Syria gave the book a good first decade. Its weakness ran the other way: Stephens audited the costs of restraint in detail while treating the political aftermath of intervention as a secondary problem, and Iraq stood as the permanent exhibit against him.
The Pulitzer board gave him the 2013 prize for commentary, citing the force and contrarian edge of his columns. The submitted work attacked Republicans as well as Democrats, defended Fareed Zakaria during a plagiarism controversy, and argued for protecting vulgar speech. He later chaired Pulitzer juries and judged the Livingston Awards. By his early forties he sat inside every institution of establishment journalism while writing, increasingly, that those institutions were failing.
The move
In April 2017, Opinion editor James Bennet (b. 1966) brought Stephens to The New York Times. The logic was simple: the paper's readership had just been blindsided by an election, and Bennet wanted a conservative who rejected Trump and accepted the institution. The execution detonated on arrival. Stephens's first column, "Climate of Complete Certainty" (April 28, 2017), argued that climate advocates hurt their cause by claiming more certainty than the models support. Climate scientists dismantled it line by line. Readers posted screenshots of their canceled subscriptions. Colleagues denounced the hire in public.
Stephens had walked into a structural problem no column could solve. The Times wanted ideological range; a large part of its staff and audience believed some positions sat outside legitimate debate, and Stephens's climate record sat there. He, for his part, wanted to unsettle liberal readers from inside an institution they trusted, and he sometimes seemed to prefer the unsettling to the strongest available case.
Five years later he went to Greenland, stood on the melting ice, talked to the scientists, and wrote that his doubts had dissolved. He accepted anthropogenic warming as real and serious while continuing to argue for nuclear power, innovation, and adaptation over restrictions on growth. Critics asked, fairly, why a man who prized evidence needed a glacier when the published record had been available for decades. The episode displayed both halves of his temperament. Suspicion of consensus delayed him. Absence of loyalty to his own prior positions let him move, in public, with his name on it, which remains rare in his trade.
The bedbug
On August 26, 2019, word spread that bedbugs had been found in the Times newsroom. David Karpf, a George Washington University professor, tweeted a joke to his modest following: "The bedbugs are a metaphor. The bedbugs are Bret Stephens." The tweet drew nine retweets and no likes. It was on its way to oblivion.
That evening Karpf opened his email and found a message from Stephens, with the university's provost copied. Stephens invited the professor to visit his home, meet his wife and children, and "call me a 'bedbug' to my face." Karpf posted the email. The internet did the rest. Stephens deleted his Twitter account within a day and then wrote a column on the history of totalitarian regimes describing their enemies as insects, which readers took as a comparison of his own situation to that history, and the mockery started over.
The episode deserves its place in the biography because it tested his stated creed. Stephens's best-known lecture, "The Dying Art of Disagreement" (2017), argues that democratic life requires thick skin, curiosity about opponents, and the discipline to argue without seeking to destroy. The lecture is good. The email copied a man's employer over a joke. Stephens argued that vermin metaphors carry a dark lineage, which is true, and beside the point; a Pulitzer-winning columnist at the world's most powerful newspaper held every advantage over an associate professor with a small Twitter account, and he reached for the man's boss. The gap between the creed and the conduct is the most honest datum we have about how hard the creed is to live, including for its author.
Four months later came "The Secrets of Jewish Genius" (December 27, 2019), an essay attributing disproportionate Ashkenazi achievement to a culture of literacy, argument, and dissent. The original version cited a 2005 paper proposing a genetic explanation, co-authored by Henry Harpending (1944-2016), an anthropologist with ties to racialist thought. The Times removed the citation and appended an editor's note. Stephens maintained his thesis was cultural. The defense was plausible and the citation was careless, on a subject where the history of race science leaves no room for carelessness. In 2021 he wrote a column defending reporter Donald McNeil, pushed out after using a racial slur while discussing the slur with students; the Times declined to run it, and it leaked. Each controversy repeated the pattern: a writer who prizes force and surprise, handling evidence and power less carefully than his own standards require.
Homeless
Stephens opposed Trump from 2016 forward, and the objection went deeper than policy. His conservatism existed to defend the liberal constitutional order: alliances, courts, truthful speech, the peaceful transfer of power. Trump treated each as an obstacle. When the Republican Party submitted, Stephens concluded the party had left him, a phrase he used as a title for public appearances in 2026. At the University of Iowa in October 2025 he described his position as membership in "the conservative wing of the liberal church."
He refused the easier exits. He never became a liberal; he kept his Zionism, his market economics, his hostility to identity politics. He also refused to soften on Trump when softening would have restored his standing on the right, though he grew more willing to distinguish Trump from Trump's voters, arguing that a governing class which failed at war, finance, and borders had manufactured its own populist revolt. When JD Vance (b. 1984) lectured Europeans on democracy at Munich in 2025, Stephens answered that a man who would not say Trump lost in 2020 had no standing to give the lecture. When the second Trump administration did things he had advocated for decades, he said so, which brings the story to Iran.
On February 26, 2026, with American strikes visibly imminent, Stephens published "The Case for Hitting Iran." He faulted Trump for refusing to explain the war to the country, called that refusal a serious moral and political failure, and then made the case Trump would not make: Iran's nuclear program, its alliance with Moscow and Beijing, its proxies, its attacks on shipping. Two days later the United States and Israel struck, and the war began. Two weeks after that, with Iranian missiles falling on Israel and the Gulf states, his column opened with the question David Petraeus (b. 1952) put to a reporter during the drive on Baghdad in 2003: "Tell me how this ends." Critics of the war filed Stephens among the pundits who had learned nothing since Iraq. His own framing was that he had learned exactly one thing: support the use of force, demand an accounting of its purpose, and stop promising that force will remake the society it strikes. Whether that distinction survives contact with a long war is the open question of his late career. He is writing a book about the future of the free world while the free world's leading power is led by a man he considers unfit, which is either the best or the worst possible condition for the project.
The Jewish turn
Israel has been a constant since Jerusalem, and after October 7, 2023, it moved to the center. Stephens read the massacre, and the celebration of it on Western campuses, as the failure of an entire educational and media class. His positions since then resist the usual sorting. He rejected the genocide charge against Israel in July 2025, arguing the term's application there worked to twin the Jewish state with Nazi Germany. The next month he wrote that a permanent Israeli reoccupation of Gaza would force thoughtful supporters of Israel into opposition to Benjamin Netanyahu (b. 1949), and he demanded an immediate, large increase in food and medical aid. He held both that Hamas bore responsibility for starting and prolonging the war and that this responsibility released Israel from nothing.
Since 2021 he has edited Sapir, a quarterly funded by the Maimonides Fund, and the job marks a change in his ambitions. The columnist addresses a national audience; the editor of Sapir helps set the agenda of American Jewish institutional life, issue by issue: power, continuity, education, antisemitism, chosenness. His argument to Jewish philanthropy has been to spend less on monitoring haters and more on building Jews: schools, texts, confidence, attachment. His essay "My Liberal Faith" gave the project its widest frame, describing liberalism as an inheritance of conscience, freedom, and restraint that Jews have both served and depended on.
Podhoretz died on December 16, 2025, at ninety-five. Stephens's tribute in Commentary, "The Exemplar," opened where this essay opened: with Charles Stephens, born into a family of ideological feuds, finding in Podhoretz the assurance that his convictions had a lineage. The son who wrote the tribute now edits a Jewish magazine of his own, commissions the essays, writes the editor's notes, and supplies to younger writers what Podhoretz supplied to his father. The circuit closed in one generation.
The ledger
Stephens married the journalist Pamela Paul (b. 1971) in the late 1990s; the marriage was brief, and both later wrote for the Times, she as Book Review editor and then columnist. He married Corinna da Fonseca-Wollheim, a classical-music critic, and they have three children and a family life divided between New York and Hamburg. Russia banned him for life in 2022, a distinction he can be presumed to value. In spring 2026 Cornell appointed him a Zubrow Distinguished Visiting Journalist, and he appeared in Klarman Hall on March 6 to discuss democracy and journalism with the investor Seth Klarman, before an audience of students, many of whom regard his politics as archaeology. His eight-year weekly dialogue with the liberal columnist Gail Collins (b. 1945) ended on May 5, 2025, when both turned to books. In the final installment Collins told him that "hanging out with you like this for eight years was such a pleasure." Detractors called the feature a performance of civility between two comfortable people. Its defenders could answer that a decade in which two Americans of opposed politics argued weekly without contempt required no defense, and that the country produced few other examples.
The ledger, then. On the credit side: a writer who saw Trump clearly in 2016 and paid the full price on the right without collecting the reward on the left; who changed his mind on climate in public; who has spent thirty years arguing that ideological movements mean their doctrines, a claim that October 7 and February 28 keep re-proving; and whose prose, at its best, strips euphemism off institutional failure faster than any columnist of his generation. On the debit side: Iraq, and the asymmetry it exposed. Stephens examines the risks of restraint under a microscope and the risks of action through a window. He knows that weakness invites aggression. He is slower to see that strength invites hubris, that allies farm American commitments, and that wars end politically or do not end. The bedbug email sits on the same side of the ledger, a small entry with large print: the apostle of disagreement, disagreed with cheaply, reached for institutional power.
His deepest belief, inherited in a car in Mexico City traffic, is that liberal civilization is a single order, that markets, free speech, alliances, Jewish sovereignty, and American power stand or fall together, and that the West's mortal danger is shame at its own history. His deepest blind spot follows from the belief: the assumption that a West confident enough to use its power will know how to use it well. His father rehearsed the lectures. The son is still delivering them, to a country that has changed the course, in a war whose ending no one can yet describe.
The Bedbug and the Ritual Order: Erving Goffman Reads the Bret Stephens Email
On Monday, August 26, 2019, someone at The New York Times sent the staff a memo about bedbugs in the newsroom. The memo leaked to Twitter within the hour, because everything leaks to Twitter within the hour, and the joke wrote itself several thousand times that afternoon. One of the jokers was David Karpf, an associate professor of media and public affairs at George Washington University, a man who studies political communication on the internet for a living. He tweeted: “The bedbugs are a metaphor. The bedbugs are Bret Stephens.” The tweet earned nine retweets and zero likes. By the standards of the platform it was a failure, a pebble dropped down a well. Karpf went about his evening.
That night an email arrived. It came from Stephens’s New York Times account, and it copied the provost of George Washington University. Stephens wrote that he was amazed at what supposedly decent people say about strangers online. He invited Karpf to visit his home, meet his wife and children, and “call me a ‘bedbug’ to my face.” He promised to be courteous no matter what Karpf had to say.
Within twenty-four hours the columnist had deleted his Twitter account, the hashtag #BretBug had trended nationally, and a joke nine people had shared became the most famous insult of the year. The episode is usually filed under internet pile-ons, a genre with a large and undistinguished literature. It deserves better company. Erving Goffman (1922-1982) built a sociology for exactly this collision, and reading the bedbug affair through his work explains what every hot take missed: why the email, and not the tweet, destroyed face; why Stephens’s diagnosis of Twitter was half right; and why a man who commanded every institutional advantage lost a contest whose rules he did not know he had entered.
Goffman published “On Face-Work: An Analysis of Ritual Elements in Social Interaction” in the journal Psychiatry in 1955, and reprinted it in Interaction Ritual (1967). Face, in his definition, is “the positive social value a person effectively claims for himself” during a particular contact. The claim is social. Others must honor it, and the honoring runs on a tacit compact: I will work to protect your face, you will work to protect mine, and both of us will steer the encounter away from moments that would discredit either claim. Goffman called the whole apparatus the ritual order. His larger point, developed across The Presentation of Self in Everyday Life (1959), “The Nature of Deference and Demeanor” (1956), and Relations in Public (1971), was that this order is self-enforcing and universal. No rank exempts a person from it. A king can violate the ritual order and the violation costs him face on the spot, in front of the court, with no appeal.
Start with the tweet, because Goffman would. The joke was what he called aggressive face-work: scoring a point at another’s expense before an audience, the everyday blood sport of wit. Goffman treated point-scoring as a normal, regulated activity with a crucial condition: the target is usually present, or the remark stays in a region where the target will never hear it. Every social world runs on this segregation. The office lunch where the boss is mocked, the seminar hallway where the famous scholar is deflated, the family car after the dinner party. Goffman’s term was backstage talk, and his rule was that regions are kept separate by walls, doors, and discretion.
Twitter demolishes the walls. Joshua Meyrowitz extended Goffman to electronic media in No Sense of Place (1985), arguing that broadcast technologies merge situations that physical architecture kept apart, and the social-media scholars who followed him, Alice Marwick and danah boyd among them, gave the merger its current name: context collapse. Karpf’s tweet was backstage talk delivered in a technically public region. Its intended audience was a few hundred followers who shared his politics and his mordancy. Its possible audience was everyone, including the target. Karpf, whose scholarly field is this exact phenomenon, understood the risk in the abstract better than almost anyone alive, and rated it low, reasonably, because nine retweets is a whisper. Stephens was, in Goffman’s vocabulary, an unratified hearer, a man with his ear to a door that no longer exists.
Now the email, the document on which the whole episode turns. Stephens clearly experienced the tweet as what Goffman called a challenge to face, and he initiated what Goffman called the corrective process. Relations in Public lays out the remedial interchange in sequence: a virtual offense occurs, the offended party signals the breach, the offender supplies remedy (an account, an apology), the offended party grants relief, the offender expresses appreciation, and the ritual equilibrium is restored. The process is delicate, and Goffman’s rule for it is proportion. The signaled breach must be calibrated to the offense. An overweight challenge does not restore the ritual order. It commits a new offense, which then requires its own remedy, with the roles reversed.
Measure the calibration. The offense: a pun, nine retweets, zero likes, from a stranger. The challenge: a personal email from a Pulitzer Prize winner at the most powerful newspaper on earth, sent to the joker’s employer, with the sanctioning officer of his university copied on the correspondence. Goffman spent his career cataloguing the small change of interaction, and the cc line is small in exactly the way he taught readers to notice. Two lines of header. It converts the genre of the document. Without the cc, the email is a challenge between persons, eccentric, aggressive, but ritual. With the cc, it is a complaint to an institution, and it imports the possibility of professional consequence into what pretends to be an affair of honor.
That doubleness is the analytical core. The body of the email speaks the grammar of the duel. Come to my home. Meet my family. Say it to my face. Goffman wrote about this grammar in “Where the Action Is,” his essay on character contests, the run-ins where each party stakes his claim to courage, composure, and integrity before witnesses, and where the stake is the self. The duel is the character contest in its pure historical form, and the duel has one absolute requirement: the parties meet as equals, stripped of external advantage, or the contest proves nothing. The cc line speaks a different grammar, the grammar of the personnel file. It summons a superior. It converts the challenger from a gentleman demanding satisfaction into a customer demanding a manager. No document can speak both grammars at once. A man cannot invite you to a test of nerve and notify your boss in the same breath, because the notification concedes what the invitation denies, that he is unwilling to meet you inside the ritual order and has gone outside it for leverage. Every reader of the email understood this instantly and pre-theoretically. Goffman supplies the theory: the email failed as a challenge because it broke the equality condition of the character contest, and it failed as a complaint because the offense was beneath institutional notice. It fell between two orders and discredited its author in both.
The strange thing, and the thing a Goffman reading recovers that the pile-on could not, is that Stephens’s underlying intuition was sound. His email reached for copresence. Say it to my face is folk Goffman. It expresses the correct observation that face-to-face interaction carries ritual constraints, mutual monitoring, embarrassment, the sheer difficulty of being cruel to a body in the room, and that mediated channels dissolve those constraints. Goffman’s entire body of work rests on the special properties of copresence; Meyrowitz’s rests on what happens when media subtract them. Stephens diagnosed the disease accurately. His remedy administered a stronger dose of it. He answered a context collapse with a larger context collapse, dragging Karpf’s employer into a joke, and he enforced the norms of copresence through a channel that carried, however he later denied it, the shadow of a threat. The doctor prescribed the pathogen.
Watch Karpf’s counter-moves, because they are a clinic. He did not reply to Stephens. He tweeted that he had received the email and asked his followers whether he should post it. The move looks like hesitation and functions as ratification: it assembles an audience and obtains its consent, converting a private receipt into a public exhibit with the audience implicated in the exhibiting. Then he posted the screenshot with Stephens’s contact information cropped out, and said in interviews that he wanted no one to lose a job over this. Goffman’s essay on deference and demeanor distinguishes the two currencies of ritual: deference, the regard shown to others, and demeanor, the composure through which a person shows he deserves regard. Karpf paid both. He displayed poise, withheld escalation, protected his antagonist’s privacy while destroying his position, and performed, before a national audience, the courtesy that Stephens’s email had promised and failed to deliver. The demeanor was the argument.
It rested on a material base, and Karpf said so plainly afterward. He had tenure. The provost could read the email and do nothing, and did nothing. An adjunct, a lecturer, a graduate student receives the same cc and feels the floor move. Goffman is sometimes accused of ignoring power, of describing the ritual order as if all players held equal chips. The bedbug episode shows the relation between the orders with unusual sharpness. Institutional power determines who can afford to play the ritual game with composure; the ritual game then redistributes face without regard to the chips. Stephens held the institutional advantage and it purchased nothing, because the contest occurred in the ritual order, where the only currency is conduct. Karpf’s tenure did not win him the exchange. It bought him the calm with which he won it.
By Tuesday the corrective process had reversed polarity. The offense on the table was no longer the tweet; it was the email, and the internet, acting as what Goffman would recognize as an enormous, undisciplined sanctioning audience, demanded remedy from Stephens. He supplied two, and both misfired in ways the literature predicts. Marvin Scott and Stanford Lyman, building on Goffman in their 1968 paper “Accounts,” distinguished excuses, which admit the act was wrong while denying full responsibility, from justifications, which accept responsibility while denying the act was wrong. Apology, Goffman’s own preferred remedy, splits the self in two, a blameworthy past self and a reformed present one that joins the audience in condemning it. Stephens, on MSNBC the next morning, chose justification, the one form that restores no face. He said that likening people to insects was wrong in all cases, that the world’s worst regimes had talked of men as vermin before killing them, and that his email had threatened nothing, an assertion the cc line quietly contradicted on every screen in America. He condemned the offense he had suffered instead of the offense he had committed. The remedial slot stood open, and he filled it with a second challenge.
His other move was avoidance, and here he followed Goffman to the letter. The avoidance process, the 1955 paper explains, is the face-saving option of last resort: when an encounter can only cost face, stay out of the encounter; when a region has become impossible, leave the region. Stephens deactivated his Twitter account within a day, pronouncing the platform “a sewer” on his way out. The judgment was defensible and the timing was legible. Everyone could see that the sewer had been survivable for eight years and became intolerable in the specific week it turned on him. Exit read as flight, and flight, in a character contest, is a verdict.
Then came the strangest artifact of the week. On August 30, Stephens published his Times column, pegged to the eightieth anniversary of the German invasion of Poland, on the rhetoric that prepares slaughter, including the language that renders human beings as insects. He never mentioned Karpf, the email, or the week he had just lived, and no reader failed to supply them. In Frame Analysis (1974) Goffman described keying, the transposition of an activity into another register, the fight keyed as play, the tragedy keyed as rehearsal. The column attempted the largest upward keying available in Western letters: a personal embarrassment re-keyed as the prologue to genocide. Audiences are expert frame-detectors, Goffman insisted, and this audience detected the key in one reading. A man mocked as a bedbug had responded, in the paper of record, by associating his mockers with the architects of the Holocaust. The laughter that followed was the sound of a frame collapsing under its cargo. And the column produced a final ritual injury the columnist seems not to have anticipated: embarrassment, Goffman showed in his 1956 paper on the subject, radiates. It attaches to everyone identified with the flustered performer. The Times had been handed no role in the affair until the column gave it one.
Settle the ledger, then, in Goffman’s currency. The tweet cost Stephens almost nothing; nine retweets is below the threshold at which face is engaged, and unchallenged, the joke dies in the well Karpf dropped it into. The email cost him nearly everything, because it was performed incompetently in an order that forgives incompetence in no one: an overweight challenge, a duel with a cc line, a demand for deference issued through a failure of demeanor. Each subsequent move, the justification, the exit, the keyed column, was a remedy that renewed the offense. Karpf ended the week with more face than he began it, having done little except display composure while his antagonist spent his own. The affair confirmed the oldest and hardest claim in Goffman, the one that sounds like moralism and is actually mechanics. The self a man defends in these contests is a ritual object, constructed in the encounter, sustained by conduct within it, and it can be lost only by its owner. Nobody called Bret Stephens a bedbug and made it stick. Nine people saw a pun. Millions watched a remedial interchange performed backward by its injured party, and the performance, as Goffman knew, is the man.
The Cartographer’s Complaint: Bret Stephens and Hallin’s Spheres
On April 28, 2017, The New York Times published the first column of its newest hire, and by nightfall readers were posting screenshots of their canceled subscriptions. The column, “Climate of Complete Certainty,” made a claim so hedged it bordered on the trivial: that warming is real, that models carry uncertainty, and that advocates who claim total certainty overreach. The reaction treated it as heresy. Scientists rebutted it line by line. Colleagues denounced it in public. A petition circulated. The gap between the modesty of the claim and the violence of the response is the puzzle, and the puzzle has a solution, worked out three decades earlier by a scholar studying an entirely different war.
Daniel Hallin published The Uncensored War: The Media and Vietnam in 1986, and buried in its account of how American journalism covered that conflict is the most useful map ever drawn of how the press decides what counts as an opinion. Hallin proposed that journalists sort all political statements into three concentric regions. At the core sits the sphere of consensus, the realm of motherhood and the flag, where journalists feel no duty to present another side and act as celebrants of shared values. Around it lies the sphere of legitimate controversy, the region of elections, budgets, and wars-in-dispute, where the rituals of objectivity and balance apply. Outside lies the sphere of deviance, the region of actors and views that journalists and the political mainstream “reject as unworthy of being heard.” A position in the outer sphere gets no rebuttal and no balance. It gets exclusion, and the exclusion is performed as a professional duty rather than a political act.
Hallin built the model to demolish a myth. The press did not, he showed, oppose the Vietnam War from conviction; it moved when the boundaries moved, and the boundaries moved when elites split. Coverage turned critical after establishment figures turned critical, a pattern W. Lance Bennett later formalized as the indexing hypothesis: the range of debate in the news tracks the range of debate among officials. The deepest implication of the model is the one Jay Rosen (b. 1956) drew out when he revived it for the internet age in 2009: the sorting is invisible. No editor announces that a view has been assigned to deviance. The assignment happens in a thousand small decisions about who gets quoted, which claims need balance, and which need only condemnation, and journalists experience these decisions as news judgment rather than as the most consequential political act they perform.
Read through Hallin, Bret Stephens’s career stops looking like a sequence of controversies and becomes one continuous activity. He is a boundary litigant. For thirty years he has fought over the placement of the line between legitimate controversy and deviance, sometimes as advocate, sometimes as defendant, sometimes as mapmaker, and the fights he has won and lost track the largest redistricting of American public speech since Hallin’s own subject.
Begin with the hire, because the hire was cartography before it was employment. James Bennet brought Stephens to the Times in April 2017 as an explicit response to November 2016. The election had falsified the paper’s map. Sixty-three million voters had endorsed a candidate whom the Times’s opinion pages had placed, with almost no dissent, in the sphere of deviance, and the paper concluded that its sphere of legitimate controversy had been drawn too small to contain the country. Stephens was hired as an annexation: a conservative, anti-Trump, pro-market, hawkish, who would extend the map rightward without extending it to Trump. The plan embodied a theory, that the spheres should be drawn to match the electorate. The staff revolt that greeted his first column embodied the rival theory, that some boundaries are drawn by evidence and should not move for any electorate.
The climate fight is the cleanest specimen of a sphere dispute in the modern record, because the two sides were not arguing about the answer. They were arguing about which sphere holds the question. Stephens’s column, read carefully, conceded the science and contested the certainty of downstream projections, a position whose natural home is legitimate controversy. His critics answered that the anchoring claims, warming, human causation, serious risk, sit in the sphere of consensus, placed there by evidence rather than convention, and that a Times column treating them as debatable performs the one act Hallin says the middle sphere performs: it legitimates. This is the sharpest edge of the model. Admission to the sphere of legitimate controversy is a prize independent of winning any argument inside it, because admission announces that reasonable people disagree. The tobacco industry understood this for forty years; its memos asked for doubt, never for victory. The scientists who fought Stephens’s column had read that history and were policing the border, and the indexing hypothesis explains why the border was contestable at all: American political elites were split on climate while scientific elites were not, and Hallin’s model has no rule for which elite a newspaper must index. The Times, hiring Stephens, indexed to the Senate. His critics demanded it index to the National Academy. The column was the ground on which two indexing regimes collided.
Stephens settled his own case in 2022, standing on the Greenland ice, writing that his doubts had dissolved. In Hallin’s terms he did not lose the argument; he emigrated, moving his residence from the contested border zone into the consensus sphere while keeping his policy preferences, nuclear power, adaptation, growth, inside legitimate controversy where they belong. The emigration was tacit acknowledgment of what his critics had claimed in 2017, that the empirical core had never been controversy at all.
His earlier career shows the machinery running the other way, and it shows him profiting from it. In 2002 and 2003, the years he edited The Jerusalem Post and returned to The Wall Street Journal, hawkish internationalism occupied the sphere of consensus in American media. Hallin’s Vietnam finding replayed almost without revision: with both parties’ leaderships supporting the Iraq invasion, opposition was indexed to the margins, and the mainstream press treated skeptics of the war roughly as it had treated skeptics of Vietnam in 1965, as unserious, as deviant, as unworthy of the balance rituals. Stephens flourished inside that map. He did not draw it, but he never protested it, and the columnists who spent 2003 explaining that war opponents stood outside respectable opinion were performing the same boundary enforcement he would later suffer. Then the map inverted under his feet. Trump attacked the Iraq war from inside a Republican debate stage in 2016 and paid no price, which announced that elite consensus had cracked on the right, and within a few years the neoconservative position had been pushed toward the margin of the party that invented it. Stephens’s positions barely moved across these twenty years. The line moved across him, twice, in opposite directions, which is the experience that separates him from colleagues who have only ever policed a border and never been deported over one.
Inside the Times, the boundary wars of 2020 and 2021 ran through his corridor. The paper published Senator Tom Cotton (b. 1977) arguing for military deployment against riots in June 2020; the newsroom rebelled; Bennet resigned within days. Bari Weiss (b. 1984), hired in the same annexation as Stephens, resigned the next year with a letter arguing that the paper’s sphere of legitimate controversy had contracted to the tastes of its angriest online readers, that “Twitter has become its ultimate editor.” Stephens defended the Cotton publication and, in 2021, wrote a column defending the reporter Donald McNeil against dismissal over quoted use of a slur; the paper declined to run the column, and it leaked. Set aside the merits of each episode and observe the form. Every one is a jurisdictional fight over who draws the map: editors, the newsroom, the subscriber base, the platform. Hallin wrote about an era when a small elite drew one national map in private. The Times of 2020 was an institution in which the drafting room had been stormed, and Stephens spent those years as the house litigant for the widest possible middle sphere, losing most of the cases.
He has also crossed the line in the other direction, and the sanctions he drew are worth reading as procedure. The 2016 column describing antisemitism as a disease of the Arab mind, and the 2019 essay on Jewish genius with its citation of a racialist-adjacent intelligence paper, were judged by his own paper’s readers, and in the second case by his own editors, to have entered the sphere of deviance, where race science lives by settled assignment. The response to the second case followed the pattern the sociology of journalism calls paradigm repair: the offending citation was removed, an editor’s note was appended, and the note functioned as a boundary marker, informing readers that the paper knows where the line runs and regrets the trespass. Thomas Gieryn named the general activity boundary-work in 1983, writing about scientists; Matt Carlson and a generation of journalism scholars have documented the press performing it on its own members. Stephens’s editor’s note belongs in that literature as a specimen: the deviant crossing acknowledged, the border restored, the columnist retained. Exclusion has degrees, and a note is exile’s cheapest form.
Then came October 7, 2023, and the redistricting that now organizes his work. On the campuses, in the NGO world, and across large territories of online discourse, the position Stephens has held since boyhood, that Jewish political sovereignty is legitimate and necessary, began moving toward the sphere of deviance: Zionist speakers shouted down as beyond debate rather than debated, Zionism bracketed with racism in the resolutions of student governments, the word deployed as a disqualification. The fight over the genocide charge, which Stephens joined in July 2025, is at bottom a classification war conducted in Hallin’s terms, since the side that wins the label decides which position becomes undiscussable: if the charge holds, defense of the war is deviant; if it fails as he argued, the charge’s casual use is. And here the model requires the amendment that Stephens’s career has been quietly demonstrating for a decade, the amendment that constitutes an actual contribution to the literature rather than an application of it.
Hallin assumed one map. He could. His subject was a media system with three broadcast networks and a handful of national newspapers, institutions that drew a single national geography of opinion, which is why exile from it was total, and why his Vietnam-era dissidents simply vanished from the screen. That condition is gone. The spheres have gone jurisdictional. A position now holds a different status in each of several partially sealed discourse systems: Stephens’s Zionism sits in the sphere of consensus at Commentary and Sapir, in the sphere of legitimate controversy at the Times, and in the sphere of deviance at a growing number of elite universities, simultaneously, with no authority able to harmonize the maps. His anti-Trumpism sits in legitimate controversy at the Times and in deviance across the MAGA media system, which runs its own outer sphere and has assigned him to it by name. Under one map, a career was a location. Under many maps, a career becomes arbitrage, and Stephens is the type case: a writer who holds citizenship in several jurisdictions and full standing in none, and whose weekly output can be read as a customs declaration, arguing cargo across borders that face different directions.
His institutional response fits the amended model precisely. When a writer loses a boundary fight inside someone else’s jurisdiction, the historically available moves were surrender or exile. Fragmentation created a third: found a jurisdiction. Sapir, which Stephens has edited since 2021, is a machine for drawing a map, an American Jewish discourse in which the questions he considers legitimate, power, chosenness, the costs of the diaspora reflex, receive the balance rituals of the middle sphere, and the positions he considers deviant, the apologetics for Hamas above all, receive exclusion. This is what founding a journal has always meant; Podhoretz’s Commentary did the same for an earlier map. What is new is only the multiplicity, the fact that the founded jurisdiction does not replace the lost one and does not need to.
Which returns the story to his creed. “The Dying Art of Disagreement,” the 2017 lecture that stands as Stephens’s manifesto, is best read as normative cartography: an argument that the sphere of legitimate controversy should be drawn as wide as the society can bear, and the sphere of deviance as narrow. It is a serious position with a distinguished ancestry, and it contains the concession its admirers skip. Stephens keeps a deviance sphere too. He would not platform Holocaust denial for balance, or debate the massacre of October 7 as a question with two sides; he has said as much, and he is right to. No one in this fight proposes abolishing the outer sphere. Every combatant, the climate scientists, the campus activists, the newsroom staffers, the columnist, accepts Hallin’s geography and fights over the surveying, and the fights are unwinnable in the way boundary disputes are always unwinnable, by argument alone, because the boundary is upstream of argument; it decides what argument is for. Stephens has spent a career on that terrain, mostly without the vocabulary for it. Hallin supplies the vocabulary. The man is a cartographer’s complaint in human form: drawn onto three maps at once, legitimate here, deviant there, consensual in the journal he edits, and still filing columns each week that assume, against the visible evidence of his own biography, that somewhere a single map of the sayable exists and can be argued into shape.
Exit, Voice, and Bret Stephens
Albert O. Hirschman (1915-2012) knew about exit before he theorized it. He left Berlin in 1933, weeks after the Reichstag fire, a Jewish teenager who read the map faster than the adults around him. He fought in Spain, and in 1940 he worked in Marseille with Varian Fry (1907-1967), forging documents and scouting foot routes over the Pyrenees, running what amounted to an exit operation for several thousand people whose voice had failed absolutely. Thirty years later, as an economist at Harvard, he compressed the experience into the most portable framework the social sciences produced in his generation. Exit, Voice, and Loyalty (1970) begins with a small question, what customers do when a firm’s quality slips, and ends as a general theory of decline: members of any organization, firm, party, church, or state, have two responses to deterioration, leaving and complaining, and the fate of the organization depends on the mix. Exit is the economist’s response, silent and efficient. Voice is the political response, articulate, costly, and carrying what Hirschman called “the messiness and heartbreak of voice.” Loyalty is the third term, the attachment that delays exit and thereby gives voice time to work.
Bret Stephens’s career is a controlled experiment in the framework, run across every institution he has touched, and the family history that produced him is a chapter of its prehistory.
Start there, because the deep structure starts there. The Stephens family archive is a ledger of exits taken when voice became impossible. Kishinev after the pogroms: no petition available to a Jewish family in the Russian Empire of 1903, so Louis Ehrlich’s people left for New York. Milan, 1950: Xenia’s family out of a Europe that had spent the previous decade demonstrating what it thought of Jewish voice. Hirschman would have recognized both departures professionally; they were the kind he facilitated in Marseille. And the political tradition Stephens inherited, Zionism, is the exit option built into permanent form. The founding Zionist argument, made explicit from Pinsker and Herzl forward, holds that Jewish voice within other peoples’ polities has a ceiling, that emancipation-era Jews were the most loyal members European states possessed and the loyalty purchased nothing, and that a people in that position requires what no diaspora can supply, a guaranteed exit and a sovereign voice. Stephens’s attachment to Israel, usually parsed as ideology, parses more cleanly as Hirschman: he descends from people who needed exits and got them barely, and he supports the one institution on earth whose purpose is that no Jew should ever again depend on the mercy of someone else’s voice.
Now run the experiment forward, to 2016, the year the framework caught up with him personally.
Through 2015 he held a position Hirschman would classify as ideal membership: a loyalist-critic inside the Republican coalition, wielding voice from the deputy editorship of The Wall Street Journal editorial page, an organ that functioned as the party’s quality-control department. When Trump entered the primaries, Stephens deployed voice at maximum volume, column after column, and discovered what the volume was worth. The primary electorate was not reading the quality-control reports. Voice failed completely, and it failed in the presence of an audience, which is the most instructive kind of failure. In November 2016 he announced he would vote for Hillary Clinton (b. 1947), and the announcement completed the sequence: voice exhausted, exit executed.
But watch what he exited, because the precision is the whole game. Hirschman’s customer leaves a firm because the product deteriorated; the exit is an act of fidelity to the product, performed against the firm that abandoned it. Stephens left the party while refusing to leave the tradition, and his stated warrant, condensed in the title of his 2026 public appearances, “The Republican Party Left Me,” is a product-quality claim in its pure form: the firm quit making conservatism, so the customer quit the firm. The line carries an irony he certainly hears. It reverses Ronald Reagan (1911-2004), whose founding conversion story, that he never left the Democratic Party, the party left him, supplied the modern GOP with its origin myth. The myth now runs backward, recited by the people departing the church it built.
Hirschman has a name for what the party did next: the lazy monopoly. Most firms fear the exit of their most quality-sensitive customers, since those customers are the early-warning system. A lazy monopoly welcomes their departure, because their voice was the only discipline it faced, and their exit leaves the remaining membership quieter and the deterioration unopposed. The post-2016 Republican Party is the concept’s textbook case. It did not fight to retain its Never Trump intellectuals; it celebrated their departure, primaried their remaining allies, and converted the word “conservative,” in its internal usage, into a synonym for loyalty to one man. Hirschman’s model predicts the consequence with some exactness: an organization that expels its complainers loses the capacity to perceive its own decline. The two-party structure then closes the loop. Hirschman assumed exiting customers defect to a competitor, disciplining the original firm through the market. Stephens had no competitor to defect to; the duopoly offers a conservative internationalist no second firm. So he did the thing the model treats as the exile’s remaining option. He relocated his voice outside the coalition entirely, to the op-ed page of The New York Times.
The relocation solved one problem stated in the framework and created another. Voice after exit is the weakest voice there is; Hirschman notes that organizations discount the departed critic as an apostate, and the post-2016 right performs the discount on Stephens daily, his criticism dismissed as the grievance of a man who left. What the Times post restored was cost, and cost is the currency of credibility. A conservative columnist writing for conservatives pays nothing for his conservatism; the same sentences published before an audience of hostile progressive subscribers are billed at full price, in cancellations, in colleagues’ denunciations, in the standing invitation to be made an example. Every column Stephens files at the Times is costly voice in the strict sense, expensive to utter and therefore informative, and the arrangement runs both directions: his anti-Trump columns cost him on the right, his Zionism and market economics cost him on the left, and a writer taxed from both sides is a writer whose statements cannot be explained as audience service. Hirschman’s framework, built for firms, here prices sincerity.
The subscriber side of that arrangement deserves its own paragraph, because the customers were running Hirschman too. The canceled-subscription screenshot, the signature gesture of the 2017 backlash to his hiring, is a hybrid the book anticipates in its discussion of boycotts: exit performed as voice, the departure staged before an audience so that the firm hears it. A silent cancellation is a market datum; a photographed one is a speech act. The Times absorbed thousands of them and kept the columnist, which was the institution’s answer in the same grammar, a signal that this quality complaint would not set the product line.
Inside the building, the channels kept illustrating the framework’s fine print. In 2021 Stephens wrote a column defending Donald McNeil (b. 1954), the reporter pushed out over quoted use of a slur; the paper declined to publish it. Blocked voice does not disappear, Hirschman teaches; it converts or reroutes. The column leaked within days, voice rerouted through the unofficial channel, and Stephens stayed, which is the datum worth pausing on. He did not resign over the spiking, or over the Cotton affair, or over any of the boundary fights he lost. Set his conduct beside the alternative course, resignation-with-public-letter, the genre that combines exit and voice in one flourish and spends the member’s entire remaining influence in a single transaction. Stephens has never used the genre against an employer. His pattern is the opposite investment: loyalty, in Hirschman’s technical sense, the attachment that holds the quality-conscious member inside through deterioration he has publicly named, preserving his voice for the long argument. The pattern extends past journalism. He serves as a trustee of Middlesex, the boarding school that formed him, while writing that elite education has curdled; he chaired Pulitzer juries and judges the Livingston Awards while arguing that his profession’s political culture is failing; he remains a contributing editor of Commentary six decades into its arguments. The portrait Hirschman drew of the member most valuable to a declining institution, the one whose options make exit easy and whose attachment makes it unthinkable, is a portrait of this specific career.
There is one clean case of the other kind, and its exception status is the point. In August 2019, after the bedbug affair, Stephens deleted his Twitter account within a day, pronouncing the platform a sewer on the way out. Hirschman’s exit is normally silent and his voice normally stays; the flounce combines them, exit announced as a verdict. Stephens exited Twitter because voice there had catastrophically failed him, and the departure was rational in the framework’s terms: voice presupposes that the organization can hear, and a platform is an organization with no ears. What the case demonstrates by contrast is how anomalous it is in his record. One exit-with-verdict in a career of loyalist persistence, and it was directed at the single institution in his life that runs on exit alone, where attention is the only vote and leaving is the only sentence anyone reads.
The Conversation, his eight-year weekly dialogue with Gail Collins that ended May 5, 2025, sits at the framework’s far pole. If the bedbug exit was voice’s failure, the Collins feature was voice domesticated, ritualized, and sold: two members of opposed coalitions performing disagreement on schedule, the messiness and heartbreak converted into a subscription product. Detractors called it civility theater, and inside Hirschman’s terms the objection has content, since voice that risks nothing purchases nothing. Its defense also has content: the feature modeled, weekly, before millions, the premise on which every use of voice depends, that the other side is worth addressing rather than exiting from, and a decade of the model may have been the feature’s actual product.
Scale the framework up one last level and Stephens’s foreign policy comes into focus as the same argument conducted among states. America in Retreat (2014) is a book-length brief against exit. Its thesis, that American withdrawal from the world order is a choice with victims and that the vacuum fills with predators, is Hirschman’s logic applied to the superpower: the United States built the postwar order, holds the largest stake in it, and is the member whose exit destroys the institution outright. Alliances, in this reading, are loyalty structures, devices that make defection expensive on purpose, and the Trump-era treatment of NATO as a protection racket is the lazy monopolist’s view of membership, all dues and no product. The apparent contradiction, a man who exited his party condemning his country’s exit from its order, resolves on the book’s own terms: exit is warranted when the firm abandons the product, and an America abandoning the order abandons a product of its own manufacture. He left the GOP for quitting conservatism; he opposes America quitting the world for the same reason, fidelity to the thing over the vehicle. Whether the resolution fully survives the Iran war he endorsed in February 2026 is a question his next book, on the future of the free world, will have to answer, since a war is voice’s final form and Hirschman’s framework goes silent at the water’s edge of force.
The ledger, then. Voice failed him once, inside his party, completely and in public. He took the exit, kept the loyalty one level up, and rebuilt the voice on hostile ground, where its cost is its credential. The institutions he stayed inside, the Times, Middlesex, the Pulitzer apparatus, Commentary, hold him as the member Hirschman prescribed for declining organizations, the complainer who will not leave. The institution he left without a backward glance was the one that could not hear complaint at all. And the deepest layer never changed: a man descended from exits, loyal to the state that exists so his people will never need another one, spending his working life on the proposition that the heartbreak of voice beats the neatness of leaving, in every case where anyone is listening.
