• Summary
  • Overview
  • Key Readings

Causality is a core problem in social science methodology, as the laws of causality found in physics – which state generalizations without exceptions – are not found in the social sciences. As a consequence, classical definitions of the causal relation, such as John Stuart Mill's definition in terms of invariant succession, need either to be modified and qualified, or replaced by a different concept of causality entirely. This has lead to a long and complex literature on the problems of causality.

This four volume major reference work covers the main issues, methods of analysis, and alternatives, of causality, including the classic texts applying these alternative concepts and methods to empirical cases. The volumes give a substantial historical and philosophical introduction relevant to the concerns of practitioners. ...

Editor's Introduction


Causality in Statistical Distributions

Causality has ordinarily been understood, in the physical sciences, or at least in the philosophy of the physical sciences, in terms of law, which is to say as an unconditional invariant succession that can be represented by an equation. This identification became so close that thinkers like August Comte and Bertrand Russell envisaged the disappearance of the notion of cause itself and its replacement by law. Yet cause hung on, especially in the social sciences and in places where there were no laws, but where intervention with predictable results was possible. Initially, many thinkers in the social sciences, including Comte and John Stuart Mill, thought that the physics model of law could be adapted to understand causality in the social ...

locked icon

Sign in to access this content