The qualitative division between domain-general and domain-specific cognition is unsubstantiated. The distinction is instead better viewed as opposites on a gradual scale, which has more explanatory power and fits current empirical evidence better. We also argue that causal cognition may be more general than social learning, which it often involves.