A central debate in joint action research concerns what distinguishes joint action from individual action. Recent accounts propose that specific task representations support joint action. While initial experimental evidence supports this claim, it remains unclear how the representations that support joint and individual action are integrated with one another.

Here, we introduce an experimental paradigm that applies the logic of task-switching to examine whether joint and individual action rely on distinct or shared task representations. Across two experiments with independent samples (N = 24 each), participants were cued to press a key either individually or jointly with a partner. Crucially, the only difference between conditions was the instruction to act alone or together. By comparing response times in repetition and switch trials, we tested whether alternating between these types of actions incurred a switch cost.

One possibility is that joint and individual action rely on separate representations, such that switching between them incurs a switch cost. Alternatively, both may draw on a shared representation that can be fully or partially instantiated such that no switch cost should emerge.

Our results revealed significant switch costs, consistent with the proposal that joint and individual action depend on distinct task representations. Importantly, switch costs were larger when switching towards joint action than towards individual action, suggesting that the task representations supporting joint action may function as a dominant mode of representation that must be inhibited when acting alone.