In recent phonological theories the idea that segments are composed of unstructured lists of features has been replaced by a model in which features are grouped together in bundles or gestures. There are a number of competing feature geometries in the literature, which differ in detail, but which are similar in overall structure. The structure has a ROOT node dominating a branching tree structure. This node itself can be marked with features, usually [consonantal] and [sonorant] (or the equivalent elements) and dominates a number of other nodes, including PLACE and LARYNGEAL. The motivations for this more complex segment structure are various, but include rule simplicity. For example, place of articulation assimilation can be viewed as the copying of the PLACE node from one segment to another without having to reference the individual features that the node dominates.