A theory of objective mathematical correctness is developed. The theory is
consistent with both mathematical realism and mathematical anti-realism, an
d versions of realism and anti-realism are developed that dovetail with the
theory of correctness. It is argued that these are the best versions of re
alism and anti-realism and that the theory of correctness behind them is tr
ue. Along the way, it is shown that, contrary to the traditional wisdom, th
e question of whether undecidable sentences like the continuum hypothesis h
ave objectively determinate truth values is independent of the question of
whether mathematical realism is true.