Systemic vs Non-Systemic Errors

In CF’s taxonomy of errors, a systemic error is one that makes it harder to fix other errors, while a non-systemic error is comparatively self-contained. The defining property is not size but effect on your error-correction machinery: a systemic error degrades the very process you rely on to catch and repair mistakes, so it has leverage over a whole class of downstream problems.

This is why CF treats systemic errors as priorities. If your method for detecting and correcting mistakes is itself flawed, then patching individual non-systemic errors is like bailing a leaking boat without plugging the hole. Fixing the systemic error multiplies your ability to fix everything else, so it gives outsized return per unit of effort.

The idea parallels CF’s knowledge-skyscraper point: a flaw low in your structure of ideas threatens everything built on top of it, which is also why such errors drive error propagation. It connects to major vs minor errors but is a distinct axis — an error can be locally minor yet systemic because of what it blocks. It is also related to, but separate from, the autonomous vs dependent distinction, which classifies whether an error can act on its own rather than how it affects correction.

CF frames this within a broader, deliberately mechanistic effort to categorize error types so you can target the high-leverage ones first instead of treating all mistakes as interchangeable.