ABSTRACT
In 2003, Jean-Roger Vergnaud published his last phonology paper (chapter 3 of this volume). The paper starts by stating the “metrical hypothesis” in which a stress contour of a linguistic expression is a rhythmic organization of constituents of that expression. In particular, Vergnaud questions the formal correspondence between metrical grids and stress-bearing units, an issue which was discussed in varying degrees of depth since Liberman (1975), Liberman and Prince (1977), Halle and Vergnaud (1987a and 1987b), and Hayes (1995), among others. The whole inquiry leads Vergnaud to delve into the formalization of metrical structure. He stresses that observations as diverse as metrical structure, clocks, and other analogous rhythmic systems can be conceptually unifi ed given a domain-general notion of occurrence. One goal of Vergnaud (2003) is that occurrences and the notion of “chains” as a set of occurrences can be fully instantiated across cognitive and computational domains. To Vergnaud, any cognitive domain can be described by listing the inventory of basic formatives along with their operations, and in this regard, it is the occurrence which defi nes or restricts how operations generally apply. The entire vision that any grammatical system can be adequately defi ned by the interaction between elements and contexts is termed by Vergnaud as “Items and Contexts Architecture” (ICA) ( chapter 5 of this volume). The main purpose of this chapter is to revisit and expound the notion of occurrence as one major primitive of the ICA, and moreover to suggest possible empirical extension for further verifi cation.
