The source of syntactic ambiguity and facts concerning the resolution of such ambiguity are discussed in this paper. The attitude of qenerative linguists towards ambiguity is examined, and a working distinction is drawn between vaqueness and ambiguity. The consequences of this distinction are then examined for syntactic ambiguity and an ordering hypothesis for syntactic rules is proposed. The main issue is the ordering that must hold between optional and obliqatory rules. Three observations about syntactic ambiquity are presented: (1) it often escapes attention; (2) even when pointed out to a hearer, the hearer may find it difficult to resolve; and (3) disambiguation by paraphrase is often possible. It is shown that if a linguistic theory wants to take these three factors into account, it must be concluded that syntactic ambiguity is the result of the application of optional rules. The consequence of this conclusion, given the approach to rule orderino, is that for every "n" ways that a sentence is syntactically ambiquous, there must be at least n-1 unambiguous paraphrases. This consequence is tested on three well-known types of ambiguity. In conclusion, observations are made about possibly controversial points, namely, the definition of synonymy and disjunctive rules. (Author/AMH)