A theory of language must guarantee how the rules of its grammars apply to give the desired outputs. It has generally been assumed that the order in which rules apply is determined partly by universal principles (e.g., the Cyclic Principle) and partly by language specific ordering restrictions. Recent studies,' however, provide evidence for the hypothesis that the order in which rules apply in derivations is completely determined by universal principles. Then a natural hypothesis about rule application in general would be the following
抄録全体を表示