Thread: Quantum causality

In a 1940 article in the Physical Review Wolfgang Pauli provides a definition for quantum causality: it is 'implemented microscopically by the requirement that observables commute at spacelike separations'.

I find this confusing. Doesn't spacelike separation by definition exclude causal relationships between events? And how exactly does operator commutativity relate to causality? Might someone be able to offer a qualitative interpretation of this statement?