Classification and Prediction: A Reply to Kitts