本文根據諸多方言材料，訴諸舌根塞音塞擦化、塞擦音簡化、口腔發音部位解除三條規律，解釋宜蘭漳州腔「身軀」今讀hun55[33] su55，並著眼方言差異及詞彙擴散，根據規律演變的邏輯順序推測可能的方言形式，最終一一獲得證實。本文還主張以諧聲為軸，適度跳脫中古分類，再依循方言橫向比較的內在邏輯，合理演繹語音的縱向發展。兩個理論意義是：（一）有別於文獻將部分端透定母字讀作塞擦音視為後起的音變，本文論證不排除存古。（二）方言偶發的端組塞音讀如見組、見組塞音讀如端組，並非單純的塞音移轉，而分別體現舌根塞音塞擦化及塞擦音簡化的未然和已然。

Based on a wide range of dialect materials, this paper appeals to velar stop affrication, affricate simplification, and debuccalization to explain the pronunciation of hun33 su55 ‘body’ in the Yilan Zhangzhou dialect. Possible onset predictions following logical orders of rule development from the viewpoints of dialect variation and lexical diffusion are all empirically borne out. With Xiesheng 諧 聲 as a pivot, it is suggested that more attention be paid to synchronic phonologies rather than ancient rhyme books. Two theoretical implications lie in: (a) sporadic affricate pronunciations of Middle-Chinese Duan 端, Tou 透, and Ding 定 initials reflect ancient relics, as opposed to a later sound change advocated in previous literature; (b) sporadic pronunciations of velar stops for Middle-Chinese Duan, Tou and Ding initials and dental stops for Middle-Chinese Jian 見, Xi 溪, and Qun 群 initials do not simply indicate stop conversion, but rather manifest incomplete and complete velar stop affrication and affricate simplification, respectively.