Gatekeeping is control over who can access or circulate information. Whoever sits at a chokepoint — an editor, a platform’s policy engine, a records custodian, a community moderator — decides what passes and what is held back. The classic studies are old and durable: Kurt Lewin named the “gate” concept in the 1940s, and David Manning White’s 1950 case study of a wire-service editor showed how much of the public record turns on individual selection choices at a single desk.
Gatekeeping has entirely legitimate forms. Editorial judgment, peer review, security classification, and moderation of illegal content all filter information for defensible reasons. It becomes manipulation when the filter is applied not by a stated, consistent standard but to protect a preferred narrative — when inconvenient facts meet delay, redaction, or denial while convenient ones flow freely. Because the mechanism operates through authority and scarcity of access, it rarely looks like deception; it looks like process.
The defensive problem is that a well-run gate hides its own selectivity. The audience sees only what passed through, not what was stopped, so the picture feels complete. Recognition depends on comparison: noticing that access rules bend with the content, that a single channel is being installed as the sole authority, or that gaps appear at suspiciously useful moments. The counter is structural — refuse to let one gate define the record, and gather independently from sources the gatekeeper does not control (adjacent to information isolation, Cat 22, and censorship, T13.3).