The last two chapters defended the prospects of a real social science. Yet the idea that social science can be good science is only half the classical tradition I want to defend here. That tradition argued not just for a social science, but a science of society. A social science was to be a science of the large scale – of patterns and processes explained in large part at the macrosociological level. I investigate and defend this tradition in what follows, arguing that good holist social science is both possible and essential.

We thus shift topics in this chapter. Nonetheless, the results of previous chapters will play a key role. The work of Paige and that of Hannan and Freeman, I shall argue, are instances of successful and irreducible holist research. Furthermore, defending holism will also be further defending naturalism. Individualists claim that the only adequate social science is one that eschews reference to macrosociological entities. Yet most social science ignores this demand. So if the individualists are right, there are serious doubts about naturalism. Rejecting individualism will at the same time remove one further obstacle to science in the social sciences.