Key points are not available for this paper at this time.
Geminate consonants and long vowels frequently resist the application of rules that would a-priori be expected to apply to them; i.e., they are frequently ‘inalterable’. This article argues that, by invoking the theory of CV Phonology, it is often possible to predict which phonological rules are unable to affect long segments. The prediction follows from rather minimal assumptions about how rules apply to forms.
Bruce Hayes (Sun,) studied this question.