Several computer based tools have been developed to support cooperative work. The majority of these tools rely on the traditional input devices available on standard computer systems, i.e. keyboard and mouse. This paper focuses on the use of gestural interaction for cooperative scenarios, discussing how it is more suited for some tasks, and hypothesizing on how users cooperatively decide on which tasks to perform based on the available input modalities and task characteristics. An experiment design is presented to validate the proposed hypothesis. The preliminary evaluation results, also presented, support this hypothesis.