I think it's more useful, in the MRI tradition, to think of interaction as the area of interest, rather than the family per se. While the family is a very rich source of interaction - and the interaction of the recursive variables of meaning and behavior, ideas and actions - one can all too easily slip into old lineal models of causality if one looks at the family as either the cause of problems or the solution. I like working with families because the relational networks are rich and meaningful for the members (though exceptions always disprove the rule). Glad I found this forum for discussion. It looks to be a great one!