The original mirror-neuron concept, involving observation in another of an intentional, meaningful, goal-directed action, is clearly motor, unlike synaesthesia for touch or for pain, and may play a key role in action-understanding and imitative learning, even perhaps the acquisition and evolution of language, skills, and tool manufacture and use. Thus, to understand the goal of another person’s behaviour, we must not only match it against our own motor system, but also covertly imitate the other’s action. Indeed, forty years ago, the Motor Theory of Speech Perception proposed that we understand another’s speech by covertly and subvocally recreating, in real time, the speaker’s likely mouth movements, rather than by merely following the speech sounds. This year dramatic support came from the finding that what you hear when listening to ambiguous samples like ‘head’ and ‘had’ is influenced by externally-induced experimental deformations of the skin around your lips; you almost literally hear with your mouth! The old motor theory of speech perception clearly anticipated features of the mirror neuron hypothesis. It is also compatible with the widely-held belief that verbal language evolved from gestural communication, rather than from earlier primate patterns of vocalisation.