東勢客語的連讀變調現象

Sequential and Tonal Markedness in Dongshi Hakka Tone Sandhi

本文以優選理論 (Optimality Theory, Prince & Smolensky 1993[2004], McCarthy & Prince 1993) 分析東勢客語的連讀變調現象。本文指出，東勢客語中有兩股促使聲調產生變化的力量。這兩股力量一是希望相鄰聲調在調素層 (tonemic tier) 一致，屬於同化作用，並反映在NOJUMP-t 這個音韻制約上；另一是希望相鄰聲調在聲調層 (tonal tier) 及曲拱層 (contour tier) 不要相同，屬於異化作用，並反映在OCP-T(11)、OCP-C( ) 以及OCP-C( ) 這三個音韻制約上。然而，並非所有違反這些要求的聲調組合均會產生變調。一般而言，只有在領頭 (head) 位置出現有標聲調時，變調才易產生。這樣的現象可以運用de Lacy (2002) 領頭標記聲調的概念，透過聯合領頭標記制約 *HD/ 以及OCP-C( ) 和 NOJUMP-t 制約來補捉。

This paper investigates Dongshi Hakka tone sandhi within the output-oriented framework of Optimality Theory (OT, Prince & Smolensky 1993[2004], McCarthy & Prince 1993). Two different forces are shown to motivate the tonal alternation in Dongshi Hakka. The first force is assimilatory in nature and forces intersyllabic tone features to agree. Completely contradictory to this force is a dissimilatory effect that requires elements at the tonal level and the contour level to be different. These facts are captured by NOJUMP-t, OCP-T(11), OCP-C( ), and OCP-C( ), which regulate the well-formedness of tonal combination. In addition to tonal sequential markedness, the markedness status of a tone itself also plays a role. A low register tone occurring in a head position is shown to be marked and indirectly decides whether a tonal combination that violates a certain sequential markedness constraint will undergo tone sandhi. This can be explicitly captured by the conjunction of tonal and sequential markedness constraints.