In both the study of the mechanisms underpinning animal behaviour and its evolution, communication is said to occur when one animal’s behaviour (the signal) serves as a stimulus that physically causes a change in another animal’s behaviour, regardless of whether any of this behaviour is under voluntary control. However, signals are sometimes described as possessing a further, potentially mysterious property: that of carrying ‘information’. Some people don’t like this idea. They argue that explaining communicative behaviours in terms of senders influencing or manipulating receivers gets more of both the evolutionary and mechanistic facts right. I argue against this, and defend the idea that we should explain both the evolution of communicative behaviour and the mechanisms underpinning it using informational notions.