This course will start with a brief overview of linear phonology. The basic assumption of linear phonology was that sound can be linearly sliced up into discrete matrices of segments. The way in which phonological adjustments were accounted for in linear phonology will be briefly dealt with. After that, we will discuss the main arguments that led to the development of nonlinear phonology. Nonlinear phonology does not represent one single theory in a definite form, but rather consists of a number of subtheories (stress, feature representation, syllable structure, underspecification etc.) each of which is characterized by a number of competing descriptive models. We will discuss, among other things, syllable structure, metrical structure and feature geometry. We will focus on metrical structure and discuss a number of competing descriptive models.