This paper offers a novel account of a familiar typological observation, namely the tendency of phonological inventories to consist of segments that are dispersed through the available auditory space. In contrast to previous approaches, which have treated dispersion as a goal explicitly encoded in the grammar, this paper shows that the cross-linguistic pattern follows automatically from the interaction of two independently motivated factors: phonological representations in which only contrastive features are specified, and the enhancement of these features in phonetic implementation. The merits of this approach are illustrated by examples involving both vocalic and consonantal inventories.

This paper is concerned with syllable universals, especially the claim that all languages have syllables. Expanding beyond my earlier work, I take a new look at Gokana, the major counterexample to the universal syllable, and present overlooked (but ambiguous) evidence for a weight-insensitive bisyllabic trochee. After demonstrating the theory-dependent nature of absolute universals, and distinguishing between analytic vs. descriptive claims, I focus on the latter as a means of ‘normalising’ the discussion of what constitutes evidence for the syllable, both in Gokana and in general. A typological approach is argued for in which languages differ in the nature and extent of the ‘activation’ of phonological properties, with Gokana representing a language which only marginally activates the syllable, if at all. The paper ends by situating the issue within the context of recent discussions of universals and diversity (Evans & Levinson 2009), which have not dealt primarily with phonology.

In Optimality Theory with Candidate Chains (OT-CC; McCarthy 2007), candidates are multi-step derivations, and precedence constraints, which regulate the order of derivational steps, can inspect entire candidate derivations. This means that OT-CC opens the door to certain kinds of ‘global rules’ – that is, effects in which the application or non-application of a process is decided with crucial reference to derivational history. This paper investigates what limits may exist on OT-CC's global rule powers, focusing on two forms of opacity which are possible under a theory where all rules apply simultaneously, but not under sequential rule application: mutual counterfeeding and mutual counterbleeding. It is shown that the original version of OT-CC allows none, but that each of them could be made possible with relatively simple revisions to the original theory. Possible examples of these forms of opacity are discussed.