ABSTRACT
This chapter discusses the general consequences of the diacritic and domain treatments of invisibility. It shows that the treatment of invisibility available within Prosodic Lexical Phonology theory has greater descriptive and explanatory power than the diacritic feature approach. Virtually from its first treatment in the literature, invisibility has obtained formal representation as an entity located within the phonological string. Discontinuous constituency and medial invisibility obtain exactly the same representation in the prosodic domains theory. The prohibition against exhaustive invisibility follows directly from the representation accorded to invisibility in the domains approach. An area where the domains account derives important conditions which must be stipulated on the diacritic account involves the scope of invisibility. The chapter presents several cases of lexicalized invisibility in which what is invisible is a non-constituent. In the case of rule-governed invisibility, the single- constituent condition follows as a natural consequence of an independently motivated constraint on phonological rules.
