Traditional analyses of language [1] emphasize an arbitrary correspondence between linguistic symbols and their extensions in the world, but recent behavioral and neurophysiological 2, 3 studies have demonstrated a processing link between a symbol and its extension: that is, comprehension of language about concrete events relies in part on a simulation process that calls on neural systems used in perceiving and acting on those extensions. It is an open question, however, whether this simulation process is necessary for abstract language understanding 4, 5. Here we report how, using a new technique based on use-induced neural plasticity [6], we have obtained evidence for a causal link between the motor system and the comprehension of both concrete and abstract language.
