The purpose of this paper is to contribute to the theoretical discussion of scientific mobility in order to progress understanding of the role it plays in shaping and fostering knowledge transfers, which in turn could benefit the scientific and technological advancement of developing countries. We use Callon's socioeconomics of scientific research (1994, 2002) as the theoretical basis for posing the question: to what extent are the components of scientific knowledge embodied in scientific human capital (tacit knowledge, scientific skills, problem solving capabilities, etc.) actually appropriable and rival? Mainstream economics assumes that rivalry and appropriability are intrinsic properties of human capital. We argue, however, that 'external' factors, particularly the configuration of networks, play a role in determining the degree of rivalry and appropriability of knowledge embodied in scientific human capital, turning therefore these characteristics into extrinsic (and not intrinsic) ones. This has important consequences for the analysis of economic allocation and distribution of human capital and for the understanding of scientific mobility and the circulation and replication of knowledge. Callon's framework forces us to think about scientific mobility and its role in the diffusion of scientific knowledge in a different way, using the language of networks and emergent and consolidated configurations.