Catalan and Scott¹ propose an alternative interpretation for our findings² for CdCr₂S₄ in terms of Maxwell–Wagner effects. They also quote related isomorphs, such as HgCr₂S₄ (ref. 3), which has been discussed elsewhere^4,5. As we have shown, Maxwell–Wagner relaxations can indeed strongly affect the dielectric properties of transition-metal oxides^6,7 and semiconductors⁸. We do not find Catalan and Scott's arguments about CdCr₂S₄ convincing, however, and think that our experiments provide ample evidence for multiferroic relaxor behaviour.