Emergence has many meanings, some of them trivial or of limited interest. There is the example of dynamical emergence as the time evolution of a dynamical system; structural emergence such as the construction of a self-supporting arch from pieces of stone, or the emergence of oscillation from a capacitor and the coil, which alone don’t oscillate. Perhaps the most interesting case of emergence is in open-ended biological evolution where new potentials arise together with the new structures that use them. How is this possible? Current interest in “niche construction” and ALife modeling puts an emphasis on the relational, or interactional properties of organisms. A relational property (like fragility or similarity) requires more than one agent to define. In a number of recent works we have studied the role of relational properties in the emergence of new evolutionary forces and the consecutive emergence of species and ecosystems. To understand social agents, we may generalize from emergent evolution. Similar to the relational phenotype in an evolutionary system, the agent in a social system finds itself in a changing dynamic environment which acts back to the very properties of the individual in a relational or “permeable” way. Individuals are “not always what they are”: they change, sometimes fundamentally, with the changing history of their contexts.