Can people form coherent mental images for metaphorical actions that are physically impossible to perform such as grasping a concept, chewing on an idea, and coughing up a secret? This article reports the findings from two experiments that investigated the products of people's imagining metaphorical and nonmetaphorical actions given different forms of real and imagined bodily engagement. Our hypothesis was that people should be able to form coherent mental images for metaphorical actions because many abstract concepts are partly understood in terms of enduring embodied metaphors. For example, we conceive of “ideas” as physical entities that we can grasp, juggle, hold on to, chew, swallow, digest, and spit out given wide-spread conceptual metaphors like IDEAS ARE PHYSICAL ENTITIES. We argue that our data demonstrate the importance of embodied simulations of action in how people learn to make sense of why certain metaphors have the specific meanings they do, even in cases where these phrases refer to impossible human actions.