Image schemas have been a fundamental construct in cognitive linguistics, providing grounds for psychological, philosophical, as well as linguistic research. Given the focus in cognitive linguistics on embodied experience as a fundamental basis for language structure and meaning, the employment of image schemas in the analysis of gesture with speech is a logical extension. However, given their level of abstraction, to what degree do image schemas provide a useful explanatory tool for researching the concrete, physically embodied details of gestures? This article considers the answer to this question and then turns to a more recent theoretical development that complements the picture by encompassing a different realm of cognitive and linguistic phenomena. This research, on 'mimetic schemas', is shown to have great potential for thinking about some known phenomena of gesture in a new way. Schema research on these different levels thus provides a useful means to analyze behavior in another modality involved in spoken language use, namely the visual. © John Benjamins Publishing Company.