Societies are clearly made up of individuals and individual minds-but how? How do many interacting minds generate and constrain the group-level phenomena that are collectively called "society"? Or is it better to think of generation and constraint from the other direction, that the kind of minds people have depends on the societies they belong to? These questions are basic to the original and putative goals of cognitive science as a scholarly enterprise, touching on many familiar and persistent dualities: reductionism and holism; micro and macro; nature and society. Yet, the study of mind is currently pursued rather independently from the study of society. For the most part, the cognitive sciences study individual and interactive phenomena, such as communication, decision making, political attitudes, and moral reasoning. They also default toward the universal (Levinson, 2012). The social sciences, meanwhile, study collective phenomena, such as languages, economies, democracies, technologies, institutions, and laws; and they are chary of assuming generalizability across history or contexts. The differences between the cognitive and social sciences in methodological procedures and theoretical vocabularies are often profound, and this may be partly justified by their different levels of enquiry. However, these differences make it hard to discern the important implications each body of knowledge must have for the other. Too often, we seem to be on one side or another of a large divide.How to build more substantive bridges is a puzzle. Academic rewards work mostly in the direction of narrow specialization. The deep expertise acquired in this way is essential and rightly valued very highly, but it also generates silos. Some scholars pioneer broad frameworks by starting from their own domain of expert knowledge and building outwards. This can be highly productive, but it also runs the risk of being cavalier, or even imperialistic, in that it may not link deeply with what is already known and discussed on the other side.In our view, if integration and consilience are the main goal, then the first, most pressing question should not be (for instance), "How can moral cognition explain legal systems?," but