Humans have to coordinate many different inputs to generate a goal-directed output. Although it seems trivial that we can execute most actions in our everyday life effortlessly - it is not. Several independent processes merge to produce seemingly trivial looking actions. In research on human action control, the processes of binding and retrieval have received increased interest in recent years. In this context, a unified account emerged that strives to specify binding and retrieval in action control (BRAC) over a range of related experimental phenomena and paradigms (Frings et al., 2020). In the second symposium, we take a broad look at research that contrasts the ubiquity and limitations of action control. The first talk looks at the role of binding and retrieval for action plans that are no longer needed. The following talk investigates the role of context and episode discriminability for retrieval processes and connects to the event segmentation literature. It is followed by an investigation on the influence of stimulus modality on the segmentation of action sequences. The last two talks specifically test prevalent assumptions in the action control literature and highlight important boundaries to action control mechanisms. The contributions presented in both symposia underline the diversity of the research areas investigating human action control and highlight the prominent role of binding and retrieval processes for moving forward in understanding goal-directed human action.