This paper proposes a possible approach to present various temporal relations of virtual human actions. It combines precreated and dynamically generated (procedural) animation facilities into a unified mechanism, and focusses on controlling simultaneous animations by multiple animation channels. We present our work on language visualisation (animation) in our intelligent multimodal storytelling system, CONFUCIUS, and describe how the proposed approach is employed in CONFUCIUS’ animation engine. This approach allows the intelligent storytelling to take advantage of procedural animation effects in the same manner as regular animations, adding an additional level of flexibility and control when animating virtual characters.