This chapter discusses the frames of reference used in goal-directed arm movement. To generate a hand movement toward a visual stimulus, various sensory messages must be collected and selected by the brain in order to build up the appropriate motor commands that will steer the limb toward the target. It is usually postulated that both sensory information and motor program are referred to an internal representation of space, and interfacing them implies at least two important stages. First, the target's image projected onto the retinal map has to be relocated onto a motor (or visuomotor) map of space in which target position is specified in a body-centred system of coordinates. Second, the position of the moving limb must be specified on a proprioceptive map using the same system of coordinates as the visuomotor map, so that limb position and target position can coincide. Reference frames can be formed in the physical world, in the sensors themselves, or in the effector system. They do not need to be “frames”; they can be centres of rotation, for instance. Reference frames could also be virtual in the sense of being constructed internally by the brain to perform computations in a topological space, for instance, in which relative positions or motions are the variables that are processed.
Vanden Abeele, S., Crommelinck, M., & Roucoux, A. (1993). Frames of reference used in goal-directed arm movement. In Alain Berthoz (ed.), Multisensory Control of Movement (p. p. 363-378). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780198547853.003.0181