In this paper I sketch the outlines for a comprehensive theory of the psychogenesis of music. That is, a theory of how human beings may come to hear certain sounds and combinations of sounds as music. It is a theory that takes its empirical starting point in previous and well-known research findings on fundamental human interaction and communication. As such it incorporates at its core a developmentalpsychological theory about the human being's development of a sense of self in relation to others, from infancy on, and is further supported by findings from research on infants' behavior and reactions to music. It is argued that human interaction and communication is at the outset musical -or protomusicaland that which makes interaction and communication work is the emotive, or affective power of sound ("communicative musicality" is another term that has been used for mainly the same phenomenon). Although the empirical foundations are familiar, the comprehensive picture offered by the theory is new. The theory is structured according to a main thesis that states that music is a way of "making special" human self-development, our "sense of self ". However, for this affective-communicative theory to explain not only our reactions to protomusical sound, but also music at large (music understood in a broad universal sense), it must be extended to answer certain questions about human cognition. Therefore I start by referring to research in cognitive-psychology that explains cognition in part as a capacity to categorize phenomena according their level of detail or generality.