This comment addresses Kelley's target article that provides a model aimed at providing a broad theoretical model for all of our ongoing interpersonal life. We underline Kelley's assertion to use abstract minimal representations for assessment purposes, but observe that similar principles should hold for theoretical models. Kelley's theoretical model defies falsification because there are many possible problems for confounding. In addition the model is heavily influenced by the dominant game theoretical protocol in that it focuses on strategic behaviour to be found in control-type relationships, thereby ignoring the affiliative type of interpersonal behaviour. Finally it is observed that although sequential±temporal aspects of behaviour are integrated in the model, the model neglects the embeddedness of the dyad in other groups and organizations, thereby ignoring the long-term effect of the externalities on the collective level.
`The next major challenge is, of course, to develop and test models that evaluate the roles of task structure, individual value, other's behavior, and time as they interact to influence social decisions in more complicated ongoing relationships of interdependence' (McClintock & Liebrand, 1988).
The purpose of Kelley's target article is to identify a hierarchy of choice points and their associated types of decision rules, that people encounter in their interdependent life. By identifying and understanding both the structure and the decision rules that people apply within that structure, we then would be able to predict patterns of interaction. In Kelley's approach the concept `social orientation' is extended because it now refers to regularity in decision rules for all the decision problems that interdependent people face. Thus far the research on social orientations was focused mainly on socio-economical relations where choices directly affected own and others' outcomes. The task Kelley has put forward is an ambitious one.