[thirdtruck]

Non-scientific explanations do have to account for the neurological research that can explain the causes behind those very idea themselves, though.

We're dealing with faulty analysis software (our brains), one with documented bugs and all kinds of premature optimizations. We have to account for that.

A developer can convince his teammates, QA, management, and even the customer that a given bug is a feature. When that "feature" causes the plane guided by that software to go crashing into a mountain, however, we can't declare the mountain a "moral hazard" or "something beyond the understanding of requirements gathering". Instead, we fix the bug to account for the mountain, tell the developer to stop making that mistake, and try again.