Based on qualitative analyses of spontaneous interactions between native speakers of British English, this paper argues that speakers’ use of multimodal enactment during constructed dialogue can be motivated by stance-taking processes. Speakers use multimodal enactment (i.e. change in voice pitch, pantomime) when dis(s-)tancing themselves from a stance attributed to an absent subject. When endorsing an absent subject’s stance, they don’t use multimodal enactment, thereby iconically representing the outside stance as their own. Theoretically, this study re-evaluates Du Bois’s (2007) Stance Triangle as a Stance Tetrad: speakers simultaneously position themselves with respect to an object and both present and absent subjects.