Hate speech rules defy clean policy because of diverse legitimate contexts
Evidence
As one policy representative from LiveJournal put it, “Obviously a site does not want to wake up one morning and discover it has become the watering hole for Neo-Nazis and the KKK. It is, however, nearly impossible to write a policy for, since you also get (a) people using slurs in a reclaimed fashion for an ‘in-group’ they are part of; (b) activists documenting groups that hold those positions to ‘name and shame’; (c) people who are entirely innocent of the alternate usage/history of a particular term; (d) someone saying something thoughtless or ignorant not motivated by animus; (e) people who do hold those opinions fully aware that they are unpopular and devising elaborate vocabularies to talk about them in code.” — Tarleton Gillespie, Custodians of the Internet, 190
Categorization
- SEED: The five enumerated cases of ambiguous hate speech (reclaimed slurs, name-and-shame activists, innocent usage, thoughtless ignorance, coded evasion) could anchor an essay on why algorithmic moderation fails: every category needs human contextual judgment.
- XREF: connects to: Vague moderation terms proliferate rather than settle disputes — Tarleton Gillespie, Custodians of the Internet: Platforms,…, 235; Constitutional free speech protections create an intractable dilemma over hate speech regulation