Political speeches are an excellent example of how discourse often requires speakers to transfer multiple rivalling viewpoints, both their own and others’. Cognitive linguists have displayed how, in speech, speakers express viewpoint via individual choices at the lexical and grammatical level. Lately, cognitive linguists have also displayed that speakers express viewpoint employing speech-accompanying gestures. To date, the inquiry of viewpoint expression has centered on cases where speakers convey the same viewpoint across modalities. By exploring the persuasive uses of gesture in Obama’s A More Perfect Union speech, it is displayed how speakers can communicate multiple divergent viewpoints across gesture and speech, concurrently. There are moments when Obama expresses his opponents’ viewpoint in speech, while framing them in terms of his own viewpoint in gesture, and vice versa. This paper debates how the deviation of viewpoints across modalities supplies key insights into multimodal cognition, with respect to working memory, metaphor, and persuasion. Specifically, it is asserted that, as an implicit medium, gesture permits speakers to inject viewpoint into the uptake of speech, below the conscious radar of recipients, and it is debated how this rhetorical capacity is developing as a result of communication technologies.