Good explanations must be hard to vary

Deutsch's central diagnostic tool for separating real knowledge from myth is deceptively simple: a good explanation is one that can't be easily altered while still accounting for the same observations. He contrasts the ancient Greek myth attributing seasons to Demeter's grief with the actual explanation involving Earth's axial tilt. The myth can be endlessly rewritten — a different god, a different emotion — and still "explain" the same seasonal pattern, which reveals it's explaining nothing at all. The tilt explanation, by contrast, makes specific, checkable claims that would break if altered.

This criterion, he argues, is what separates science from superstition, and it applies far beyond physics — to economic theories, political justifications, and personal excuses. An explanation that could rationalize any outcome is providing the illusion of understanding, not the substance of it.

Takeaway: before trusting an explanation, ask whether it could be swapped for a different one without changing what it predicts — if so, it's explaining nothing.