We propose three methods for obtaining solutions in fuzzy constraint networks and study their application to the problem of ordering fuzzy numbers. The techniques proposed may be classified as defizziification functions which are applicable to any set of mutually dependent fuzzy numbers in which the dependence relationships are represented by means of metric constraints. We suggest the use of these techniques for ordering linked variables in an efficient manner, and discuss their behavior regarding several quality criteria. The first application realm of these techniques is temporal reasoning.