Most approaches to social positions define them relationally. For Marx, capitalists can only exist if there are workers, and vice versa. The two "roles" are defined by the relation between them (i.e. capitalists expropriate surplus value from the labor power of workers). Husbands and wives; men and women; minorities and majorities; lower caste and higher caste; and most other roles are defined relationally.
The regular equivalence approach is important because it provides a method for identifying "roles" from the patterns of ties present in a network. Rather than relying on attributes of actors to define social roles and to understand how social roles give rise to patterns of interaction, regular equivalence analysis seeks to identify social roles by identifying regularities in the patterns of network ties - whether or not the occupants of the roles have names for their positions.
Regular equivalence analysis of a network then can be used to locate and define the nature of roles by their patterns of ties. The relationship between the roles that are apparent from regular equivalence analysis and the actor's perceptions or naming of their roles can be problematic. What actors label others with role names, and the expectations that they have toward them as a result (i.e. the expectations or norms that go with roles) may pattern - but not wholly determine actual patterns of interaction. Actual patterns of interaction, in turn, are the regularities out of which roles and norms emerge.
These ideas: interaction giving rise to culture and norms, and norms and roles constraining interaction, are at the core of the micro-sociological perspective. The identification and definition of "roles" by the regular equivalence analysis of network data is possibly the most important intellectual development of social network analysis.