Constraint Applied to Epistemology
also Bottleneck of Ideas · Constraint-Focused Thinking
CF's transfer of Goldratt's constraint concept into reasoning: spend detailed attention only on the few factors that actually bind an idea's success, and grade the rest pass/fail.
In Theory of Constraints, a factory’s output is set by its bottleneck; improving any non-bottleneck step has excess capacity and adds nothing to throughput. CF transfers this directly to thinking. When you evaluate an idea or option, most of the factors involved already have a comfortable margin and are not what limits success. Detailed attention belongs only on the few decisive considerations that actually bind the result; everything else gets a quick pass/fail grade.
This grounds CF’s sharpest opposition: the credence-weighting and Bayesian programs that try to score and continuously update 50+ factors. Temple argues that updating on a factor with excess capacity is not high-effort precision but an outright error, since the global picture did not change. Most evidence falls within a working system’s margin of error, so the right default is usually not to update. Treating every factor as worth optimizing optimizes local optima that do not move the goal.
The mechanism is breakpoints: quantities matter only when they near or cross the threshold between good-enough and failure, converting a spectrum into qualitative yes/no judgments. This is why CF rejects weighted-factor schemes and argument strength gradations, replacing them with binary multiplication. As Goldratt put it, optimization away from the constraint is wasted; CF makes the same move the core discipline of rational evaluation.