The Firehose

An abuser who makes one person doubt their own perception and a state that makes an entire population doubt whether truth itself is knowable are, this book argues, running the same operation at different magnitudes. That equivalence is the argument; what follows is the record. It traces that operation at the scale of nations through named, budgeted, publicly acknowledged programs — because unlike almost every mechanism examined so far, this one is official doctrine, written down in the government’s own manuals.

United States military doctrine defines psychological operations as planned activities conveying selected information to audiences to influence their emotions, reasoning, and ultimately their behavior — the government’s own language, not an outside critic’s. The capability was renamed MISO in 2011, and the Defense Secretary’s stated reason is itself an instance of the mechanism it describes: the term “psychological operations,” he said, “tends to connote propaganda, brainwashing, manipulation, and deceit” — so the rename was itself an information operation aimed at perception. It was renamed back in 2025.¹

The Soviet Union ran its own parallel doctrine under the name active measures: front organizations, controlled newspapers, and deliberate disinformation to weaken adversaries and shape the global environment. Its clearest documented campaign, popularly known as Operation INFEKTION, planted and sustained the false claim that the United States had invented HIV as a bioweapon at Fort Detrick — traceable to a 1985 KGB telegram to Bulgarian state security.² It is the same narrative-control mechanism seen at interpersonal scale, here injected directly into an entire species’ understanding of a disease.

A RAND Corporation analysis of the contemporary Russian model gave this mechanism its most memorable name: the firehose of falsehood, a strategy that is high-volume, multi-channel, rapid, repetitive, and entirely indifferent to truth or even internal consistency.³ Its goal is not to win an argument but to entertain, confuse, overwhelm, and exhaust an audience’s appetite for fact-checking at all — and it works precisely because it violates the ordinary rule that persuasion requires truthfulness; falsehoods land harder, not softer, when they already fit a narrative the audience wanted to believe. This is Seligman’s learned helplessness from the first chapter, at the scale of a public rather than a single mind: when everything might be false, people stop trying to know anything.

America’s own Cold War cultural operations ran on a related but distinct logic. Radio Free Europe and Radio Liberty were secretly CIA-funded until 1972, broadcasting as psychological warfare behind the Iron Curtain under the appearance of independent journalism. The Congress for Cultural Freedom operated as a CIA front across thirty-five countries, using intermediaries — including, famously, the Museum of Modern Art — to fund artists and intellectuals who advanced American interests without ever knowing whose interests they served.⁴ This is influence laundered through plausible-deniability fronts, the secrecy-as-infrastructure pattern of the previous chapter applied not to organizational loyalty but to the shaping of ideas themselves.

The military has a plainer name for the same craft. “Perception management” is a defined entry in the Defense Department’s own dictionary — written down, on paper, to describe the shaping of foreign audiences. In 2008 the New York Times documented the same apparatus turned inward, on Americans. Dozens of retired officers who appeared on television as independent military analysts had been privately cultivated by the Pentagon as “message force multipliers,” briefed and steered to carry the department’s line on the Iraq war while presenting it to viewers as their own detached expertise; the investigation won a Pulitzer Prize.⁵ The design is the CIA cultural front just described, only faster and at broadcast scale: the audience is not so much lied to about the facts as deceived about who is speaking, because a claim lands as your own conclusion precisely when its source is invisible. The same government would later contract for “persona management” software to operate fictitious online identities in bulk — the troll farm’s method, in American hands.⁶

The theoretical lineage runs back to Edward Bernays, Freud’s nephew and the founder of modern public relations, who argued in his 1928 book Propaganda that manufacturing public consent was not manipulation to be hidden but a necessary function of mass democracy. He demonstrated the principle at real scale twice: his 1929 “Torches of Freedom” campaign rebranded women’s smoking as a feminist act specifically to open a new market for a cigarette brand, and his consulting work for United Fruit Company in the early 1950s helped frame Guatemala’s president as a Soviet-aligned threat ahead of the CIA-backed coup that deposed him.⁷ Neither campaign required outright falsehood — Bernays’ method was reframing a true fact into a different, equally true-sounding one, enough on its own to manufacture public support for outcomes serving his clients rather than the public. Jacques Ellul later distinguished agitation propaganda, short-term and emotional, from integration propaganda: long-term, designed to make people conform durably to a social order precisely because it never looks like propaganda at all.⁸ That is why the most effective operations here don’t resemble manipulation. They resemble the ordinary background texture of culture, which is the entire design.

The most consequential recent convergence is where a state’s psychological-operations doctrine and a private company’s advertising business merged into a single tool. Cambridge Analytica harvested up to eighty-seven million Facebook profiles and built psychographic personality profiles on more than a hundred million American voters, delivering individually tailored political advertising meant to move how a specific person would vote. Whether that psychographic targeting actually changed behavior is genuinely disputed — independent researchers have found little evidence it worked, and the firm’s own claims for it were largely self-promotional — but what matters here is the apparatus, not its proven effect. The Federal Trade Commission fined Facebook five billion dollars in 2019 over the underlying data practices.⁹ This is population-scale influence doctrine narrowed to one mind, delivered through the same device that carries the consumer stalkerware examined earlier — state-scale technique and individual-scale targeting in one tool. The same capability was run by a state in the Internet Research Agency’s St. Petersburg troll farm, formally named in a 2018 U.S. Special Counsel indictment of thirteen Russian nationals and three entities. Per the indictment (and per the bipartisan Senate Intelligence Committee’s own later findings, which reached the same conclusions independently of the untested criminal charges), thousands of fake personas posed as American activists across the entire political spectrum at once — not favoring one side but running genuinely divisive content in both directions. They purchased political ads and organized real-world rallies attended by unwitting Americans, recruited online by accounts they never knew were operated from Russia — the first time this mechanism produced a named criminal indictment rather than mere analyst attribution.¹⁰

A parallel British capability surfaced through the Snowden disclosures: GCHQ’s Joint Threat Research Intelligence Group operated under an internal doctrine its own documents describe as “the four D’s” — Deny, Disrupt, Degrade, Deceive — explicitly targeting individuals and groups, including hacktivist collectives, with fake personas, reputation- destruction campaigns, and honeypot-style deception, described internally as “online HUMINT” and “effects operations.”¹¹ It is the identical method stack documented in digital coercion: isolation, narrative control, reputational destruction, run by a state intelligence unit against named targets.

Research published in the American Political Science Review, extrapolating nationally from a cache of leaked internal emails from one county’s propaganda office, estimated the Chinese government fabricates roughly four hundred forty-eight million social-media posts every year — and found, crucially, that these posts are overwhelmingly cheerleading and distraction rather than direct argument or rebuttal.¹² The goal was never to win the argument but to change the subject, at scale.

The firehose is only half of the doctrine it belongs to, and its opposite half is worth naming, because the two are complementary rather than alike. The firehose floods to exhaust, until an audience gives up on knowing anything; its counterpart floods nothing, and instead feeds a target the precise, curated inputs that will lead them to reason their own way to a chosen conclusion. The Soviet military theorist Vladimir Lefebvre gave it a name in the 1960s — reflexive control — and defined it with unnerving precision as conveying to an opponent the very grounds on which they will decide, so that they arrive, apparently freely, at the decision the controller wanted all along.¹³ Where the firehose produces paralysis, reflexive control produces a choice — and the target defends that choice as their own, because in the only sense visible from inside, it was: they reasoned to it themselves, on premises someone else supplied.

This is the state-scale name for a move the opening chapters of this book already described at the intimate one. The controller who arranges which information reaches a partner, which options appear available and which are quietly foreclosed, which friend is made to seem untrustworthy just before the partner “decides” on their own to pull away, is running reflexive control in a single household — engineering a decision and then pointing to its voluntariness as proof that no coercion occurred. It is the most deniable influence there is, precisely because its product is a free choice, and the freer the choice feels, the better the operation worked. The counter is the discipline Chapter Eleven set out, turned on one’s own decisions rather than on a claim: when a choice matters, ask who supplied the premises it rests on, and whether the party who benefits from the outcome is the same one who arranged the reasons — because a decision reached freely on premises another party selected is not, in the end, a free decision at all.

A closing discipline note, because “psyop” is the term here most prone to being stretched past what it can honestly support. The Tavistock Institute of Human Relations is a real, small British nonprofit research body, founded in 1947, producing conventional academic work on group dynamics and social systems. A persistent conspiracy narrative casts it instead as a secret coordination hub for global mind control, linking it to claims as far-flung as the Beatles’ popularity being an engineered phenomenon. No documented evidence supports any of this; the Institute’s published output is exactly what its founding charter describes.¹⁴ Named, budgeted, sourced programs earn full weight; “it’s all a psyop” earns none — and the reflexive version of that claim is itself a symptom of the firehose effect it describes, manufacturing the same can’t-know-anything paralysis it warns against.