What Is Propaganda? Persuasion With the Sources Hidden

Jacques Ellul and, later, Garth Jowett and Victoria O'Donnell converged on a definition worth stating: the deliberate, systematic attempt to shape perceptions and direct behaviour to achieve a response that furthers the propagandist's intent. Three features follow. It is purposive, aimed at a specific outcome for the sender rather than at informing the receiver. It is systematic rather than incidental, run as a programme with repetition and coordination. And it conceals or obscures its own source, purpose or method, which is the feature that distinguishes it from open advocacy: a clearly labelled political argument from an identified party is persuasion, and the same argument presented as spontaneous grassroots opinion is not. Truth is orthogonal to all of this, and the most durable propaganda is generally accurate in its individual claims and highly selective about which claims it makes, which is why the fact-checking response to it is necessary and insufficient.

The catalogue was largely assembled by the Institute for Propaganda Analysis in the late 1930s and has not needed much revision:

• •Name-calling and glittering generalities, attaching a loaded label to a target or a vague virtuous one to a cause, so that evaluation is replaced by association
• •Transfer, borrowing the authority of a respected symbol, flag, uniform or institution by placing it next to the message
• •Testimonial, using an admired figure whose expertise is irrelevant to the claim
• •Plain folks, presenting an elite communicator as an ordinary member of the audience
• •Card stacking, presenting only the evidence that supports one side, which is the technique that works with entirely true statements
• •Bandwagon, asserting that everyone already agrees, which exploits the documented human tendency to align with a perceived majority
• •Repetition, which is the most reliable of all, since mere exposure increases both liking and perceived truth, an effect measured repeatedly and known as the illusory truth effect

The professionalisation happened in the First World War. Britain's Wellington House and later the Ministry of Information, and the American Committee on Public Information under George Creel, ran coordinated campaigns of extraordinary reach, and several of the atrocity stories circulated were later shown to be fabricated, which produced a public cynicism that made mobilising opinion harder in 1939. Edward Bernays, a nephew of Freud who had worked for Creel, took the methods into commercial public relations, wrote a book called Propaganda in 1928 arguing that the conscious manipulation of the opinions of the masses is an essential part of democratic society, and pioneered techniques including the manufactured news event and the recruitment of third-party experts. The totalitarian states then took it further than any democracy, with a dedicated German ministry from 1933 controlling press, radio, film and art, and Soviet practice integrating propaganda with education, employment and youth organisation so completely that the categories blur. Radio was the decisive technology, since it reached the illiterate, arrived in the home and permitted a single voice to address a nation simultaneously.

The modern form differs in mechanism rather than in principle. Computational propaganda uses automated accounts and coordinated inauthentic networks to simulate a volume of opinion that does not exist, which exploits the bandwagon and social proof effects directly. Microtargeting allows different and mutually inconsistent messages to be delivered to different audiences without either seeing the other, which removes the accountability that public messaging carried. Astroturfing manufactures apparent grassroots campaigns funded by interested parties. Firehosing, a term applied to Russian state communication, floods the space with high volumes of contradictory claims, with the aim not of being believed but of producing exhaustion and the conclusion that nothing is knowable, which is a shift in objective from persuasion to demobilisation. And the economics of attention reward emotionally activating content regardless of source, so that a system optimised for engagement amplifies the same material a propagandist would choose, without any propagandist being involved.

The countermeasures with evidence behind them are mostly about process rather than content. Inoculation, explaining the technique in advance so that a person recognises it when used, has been tested in randomised trials including large studies using short videos, and produces measurable and durable improvements in the ability to identify manipulation, and it works better than correcting individual claims after the fact. Lateral reading, leaving a source to check who is behind it before evaluating the content, is the single habit that most distinguishes professional fact-checkers from ordinary readers in comparative studies. Transparency requirements on political advertising, including archives of who paid for what and who was targeted, address the concealment that the definition turns on. Media plurality reduces the effect of any single channel. What does not work well is direct rebuttal, which risks repeating the claim and can entrench the belief, and the general finding across this literature is that the only reliable defence is against the method rather than against each message.

Propaganda is defined by purpose and concealment rather than by falsehood: a systematic attempt to shape behaviour for the sender's benefit while obscuring the source, purpose or method, and the most durable examples are accurate in detail and highly selective in scope. The standard techniques were catalogued in the 1930s and still apply, repetition being the most reliable because exposure alone raises perceived truth. Modern forms simulate consensus with fake accounts, target inconsistent messages privately, and in one variant aim at exhaustion rather than belief.