IGC (Idea, Goal, Context)

CF holds that an idea cannot be evaluated alone. To say an idea “works” or “fails” is meaningless until you ask: works for what purpose, under what circumstances? So CF’s actual unit of evaluation is the IGC — an (idea, goal, context) triple. The idea is the proposal; the goal is the objective it must accomplish; the context is the relevant background circumstances (including your existing library of criticisms). Each triple is judged as a whole.

Every IGC is either refuted or non-refuted — there is no middle ground and no degree of strength (see yes-or-no philosophy). An IGC is refuted when you know a decisive error: a reason the idea fails at that goal. Non-decisive defects do not count, because they are compatible with success. This makes binary judgment tractable, because the goal supplies a clear pass/fail bar.

The key consequence: one idea generates many IGCs and so many evaluations. The same idea can pass for one goal and fail for another — voicing a complaint may fail at problem-solving yet succeed at starting a fight. “Mostly true” then translates to: most relevant IGCs containing the idea are non-refuted.

CF deliberately splits idea, goal, and context apart rather than folding goal and context into the “idea,” and treats a modified idea as a new variant rather than a mutated original — so a refuted IGC can never silently become non-refuted. Decisions, like judgments, always take the form of accepting one IGC while rejecting competitors, unifying evaluation and choice.