We argue that it is a mistake to approach goodman's new riddle of induction by demarcating projectible from non-Projectible predicates and hypotheses, And put forward an alternative way of looking at the whole question.