Mathematicians must be more prepared to test their truths

Mathematical proofs embody a Platonic ideal of eternal truth. Two major contested proofs this week show the need for more controversy, not less