Abstract
In this paper we look at two seemingly unrelated historical processes: affrication of the Old English (OE) palatalized velars [kj] > [tʃ], e.g., OE cild > PDE child, OE cīosan > PDE choose, and the Middle English (ME) vowel unrounding [y] > [i] and [ø] > [e]. More specifically, it is argued that the front rounded vowels [y] and [ø], as well as the palatalized velars [kj] and [j], are complex melodic expressions containing two antagonistic resonance elements |I| and |U|. Furthermore, it is proposed here that the phonological system of ME witnessed a drastic change as a consequence of the introduction of the ban on the |I| and |U| merger. This *|I U| constraint is responsible for the loss of the resonance element |U| from the internal structure of both segments, which leads to the unrounding of the i-umlauted vowels and the affrication of the palatalized velars. This paper provides a detailed analysis of velar palatalization and its subsequent affrication, while additionally we address the questions of the lack of affrication before both i-umlauted and unrounded vowels, the palatalization and vocalization of the voiced velar fricative /ɣ/ and the chronology of affrication in the history of English.