In this paper I argue that it is possible to attribute beliefs and other intentional states to mute animals. This kind of attribution is substantial, in that it does allow for some minimal form of co-referential failure.