While the lexical items of human spoken languages are to a large part conventionalized and hence to a large extend non-iconic, there are pockets of iconicity in human communication. We will investigate three kinds of iconicity: speech-accompanying manual gestures, ideophones as e.g. helterskelter, and prosodic modulations as in boooring. The central research issue is the pragmatic status of such iconic meanings: Do they contribute to what is said (i.e., are they at-issue), and under which conditions can they become at-issue? If they are not at issue, are they a type of conventional implicature, or rather a type of presupposition? Existing proposals for iconic gestures that they are supplements like appositive relative clauses, or that they are cosuppositions will be tested experimentally for gestures and for the other two types of iconic meaning. We will investigate the precise alignment of iconic manual gestures and the oral expressions that they affect, the connections of gestures with ideophones and prosodic modulations, and the way how iconic meaning components contribute to the overall meaning, e.g., whether they project under semantic operators, or whether and how they can be addressed directly in communication. We will investigate the multimodal aspects of communication by means of existing corpora, production experiments with synchronized acoustic, video and motion capture data as well as audiovisual perception experiments and behavioral measure of cognitive processing (reaction times). The goal is to gain a better understanding of the often neglected iconic aspects of meaning in human communication, and of the interplay of various types of meanings.