The origins of music and musical emotions is still an enigma, here I propose a comprehensive hypothesis on the origins and evolution of music, dance and speech from a biological and sociological perspective. I suggest that every pitch interval between neighboring notes in music represents corresponding movement pattern through interpreting the Doppler effect of sound, which not only provides a possible explanation to the transposition invariance of music, but also integrates music and dance into a common form—rhythmic movements. Accordingly, investigating the origins of music poses the question: why do humans appreciate rhythmic movements? I suggest that human appreciation of rhythmic movements and rhythmic events developed from the natural selection of organisms adapting to the internal and external rhythmic environments. The perception and production of, as well as synchronization with external and internal rhythms are so vital for an organism’s survival and reproduction, that animals have a rhythm-related reward and emotion (RRRE) system. The RRRE system enables the appreciation of rhythmic movements and events, and is integral to the origination of music, dance and speech. The first type of rewards and emotions (rhythm-related rewards and emotions, RRREs) are evoked by music and dance, and have biological and social functions, which in turn, promote the evolution of music, dance and speech. These functions also evoke a second type of rewards and emotions, which I name society-related rewards and emotions (SRREs). The neural circuits of RRREs and SRREs develop in species formation and personal growth, with congenital and acquired characteristics, respectively, namely music is the combination of nature and culture. This hypothesis provides probable selection pressures and outlines the evolution of music, dance and speech. The links between the Doppler effect and the RRREs and SRREs can be empirically tested, making the current hypothesis scientifically concrete.