When all records in the navigation state are assigned a given dimension value, that dimension value is an implicit selection.
In addition to being selected explicitly by the application, dimension values can be selected implicitly. For example, if all Champagnes are from France, then the explicit selection of Wine Type: Champagne causes the implicit selection of Region: France. Implicit selection is a function of the set of records in the navigation state, regardless of what combination of search, navigation, and record filters was used to obtain them.
Implicitly-selected dimension values trigger precedence rules in exactly the same way as explicitly-selected dimension values. This behavior helps ensure a consistent user experience, by providing the same dimensions for refinement of a given result set, regardless of whether that result set was obtained through search, navigation, or a combination of the two.
For this reason, two navigation paths leading to the same set of records will always have exactly the same set of navigation selections (differing only in whether the selections are implicit or explicit). Because of this equivalence, the set of precedence rules fired in both states will be identical.
If this behavior is undesirable, it can be suppressed or modified at the application layer. In some cases, implicitly-selected dimension values reflect redundancy in the data modeling, such as the presence of duplicate dimensions.