We consider duplicate address detection in wireless ad hoc networks under the assumption that addresses are unique in two hops neighborhood. Our approaches are based on the concepts of physical neighborhood views, the information of physically connected nodes, and logical neighborhood views, which are built on neighborhood information that is propagated in networks. Since neighborhood information is identified by addresses, inconsistency of these two views might be caused due to duplicate addresses. It is obvious that consistency of physical and logical views on each node's neighborhood is necessary for a network to have unique addresses, while the sufficiency depends on the types of information contained in views of neighborhood. We investigate different definitions of neighborhood views. Our results show that the traditional neighborhood information, neighboring addresses, is not sufficient for duplication detetion, while the wireless nature of ad hoc networks provides powerful neighborhood information in detecting duplication.