Autonomous vs Dependent Errors
also Self-sustaining error · Context-dependent error · Autonomous vs non-autonomous errors
A distinction between errors that stand on their own as causal actors and errors that survive only because supporting ideas keep propping them up.
CF sorts mistakes along many axes, and one of them concerns how an error stays alive. A dependent (non-autonomous) error works only in context: it relies on other ideas, and would collapse if those supports were withdrawn. An autonomous error stands on its own, acting as an independent causal agent that can persist, spread, and effectively take on a life of its own even after the original mistake that spawned it is gone.
The practical payoff is diagnostic. Knowing which kind you face tells you where an error-correcting intervention must land. A dependent error is best removed indirectly: refute the idea it leans on and it falls with no support. An autonomous error must be confronted head-on, because attacking its neighbours leaves it untouched. CF connects autonomous errors to mechanisms that self-perpetuate, such as habits, intuitions, an “error-factory” autopilot, and static memes that replicate regardless of their host’s reasoning.
This cuts across CF’s other error categories. It is orthogonal to whether an error is major or minor and to whether it is systemic (systemic errors obstruct the correction of others). An autonomous error tends to have higher reach and resists the casual “oops, noted” response people give before repeating it, because the underlying generator was never addressed. CF’s broader aim is not heroic one-off correction but making the right idea natural and habitual, so the wrong autonomous pattern stops regenerating itself, which is also why CF treats taking error seriously as a real goal rather than an excuse-laden afterthought.