Replacive overwriting by morphological tone has been used as major evidence for cyclic construction approaches to phonology (Hyman 2013, Inkelas 2018, Rolle 2018). In this article, I show that this argument is unwarranted: tonal overwriting can be derived by simple concatenation of tonal morphemes, general phonological constraints, and minimal access to morphological information—in Autosegmental Colored Containment Theory. In detailed analyses, I demonstrate that neither morpheme-specific phonology nor cyclicity is necessary to capture tonal overwriting, whereas cyclic analyses actually make wrong predictions on possible overwriting patterns.

Supplementary data

You do not currently have access to this content.