ABSTRACT
Th e prediction of where and how people are going to move has obvious relevance for social interaction. As adults, we are extremely adept at predicting at least some of these behaviors automatically in real time. If, for example, we observe someone reaching in the direction of a half-fi lled glass on a table, we can predict with relative certainty that the reaching action is directed toward the glass. Oft en, we can also predict if the actor intends to drink from the glass or intends to remove the glass depending on the state of the actor as well as the context surrounding the action.
