Philosophy of the Social Sciences

What it asks

The social sciences study things that act for reasons. A planet does not interpret its situation; a voter, a firm and a state all do.

One tradition holds that this changes nothing fundamental: behaviour has causes, patterns can be found, and the methods of natural science apply with adjustments for complexity. The other holds that explaining action requires grasping what it MEANT to the agent, which is a different kind of achievement from finding a regularity.

The dispute is old, it recurs in every social science under different names, and both sides have too much evidence for it to be settled by argument alone.

The recurring problems

Explanation or understanding. Whether social science should look for causal regularities or interpret meanings - and whether these are rivals or two jobs.

Individualism against holism. Are social facts fully explicable in terms of individuals and their relations, or do institutions, classes and norms have explanatory power of their own? This is the field's oldest structural question.

Looping kinds. People learn the categories applied to them and change accordingly. A classification of people can therefore alter its own extension, which no classification in physics does and which makes social kinds a distinct sort of thing.

Value freedom. Weber argued that social science can and should describe without endorsing. The choice of what to study, which categories to use and what counts as an outcome are all value-laden, and whether anything survives that is disputed.

How it developed

Comte and Mill proposed a science of society on the natural-science model, and that positivist ambition set the terms.

The German hermeneutic tradition objected at once, distinguishing explanation in the natural sciences from understanding in the human ones. Weber attempted a synthesis: causal explanation that runs through the meaning an action has for the agent.

The mid-twentieth century saw the positivist model at its most confident, in behaviourism and rational choice theory. Winch's argument in 1958 that social behaviour is rule-governed and therefore requires the kind of understanding one has of a language rather than of a mechanism reopened everything.

Hacking's work on looping effects gave the field its most portable idea: that classifying people is an intervention in what is being classified. The replication problems that have affected psychology and economics have since raised a further question the field is well placed to address - what these disciplines were assuming about evidence all along.

The positions within it

Naturalism and interpretivism; methodological individualism and holism; critical theory, which denies that description and critique can be separated; rational choice approaches; social constructionism. All are in the Schools section.

Where it stands now

The credibility revolution in economics and the replication crisis in psychology have made methodology the live area, and philosophers of social science have had unusually useful things to say about both.

The other growth is in the metaphysics of the social: what money, institutions, groups and social kinds actually ARE, given that they depend on collective attitudes and are not thereby unreal.