民謠在傳統研究上有兩個缺點：一、素材僅限於少數韻行；二、只著重語言個別派生規律，而缺乏普遍性。本文建立一個720行的彰化閩南語民謠語料庫，並提出一個非派生分析。從這個語料庫顯示，大部分的韻行屬陽性節律，占90.67%，而拍板分享則是避免陰性節律的手段。本文在Prince and Smolensky(1993)的優選理論框架下，提出一組節律制約。這些制約屬於普遍語法的一部分，而在彰化民謠中呈現語言個別性的層級排序。 Traditional treatments of folk verse have two disadvantages. First, they focus only on a small number of lines. Second, they pursue derivational rules on a language-specific basis, but lack universal validity. This paper establishes a corpus of 720 lines of Southern Min folk verse in Changhua and provides a non-derivational analysis of the data. The corpus shows a preference for masculine rhythm, which is found in 90.67% of the data. Metrical beat sharing, then, allows feminine lines to be avoided. This paper puts forth a set of metrical constraints under the framework of Prince and Smolensky's (1993) Optimality Theory. The constraints are part of Universal Grammar, but are ranked language-specifically in Changhua folk verse.