Whole: Rethinking the Science of Nutrition

by T. Colin Campbell

But the key point is, you accept in advance that some configuration of observable facts could partially or completely invalidate your theory.