A deep-rooted presumption about the human mind and human agency is that mental states have significant causal influence on the way people act: When we perceive a person as performing a certain action, e.g. as reaching for a glass of water, we cannot but assume that her behaviour is driven by her thoughts, convictions, wishes or feelings, e.g. by the desire to have a drink of water. In particular, we presume that the contents of a person’s thoughts causally influence her outward physical behaviour: The person’s belief that the glass in front of her contains water, rather than machine oil, causes her reaching for the glass to quench her thirst. At the same time, it is commonly supposed that we have privileged access to our mental states, including beliefs and desires, without having to draw on empirical evidence: If a person believes that the glass in front of her contains water and desires to have a drink, she does not have to rely on empirical observation of her body and environment in order to know that she has this belief and desire. From a common-sense point of view, these assumptions appear highly plausible. From a philosophical perspective, however, they raise a number of controversially discussed issues. If one endorses, as most contemporary analytic philosophers do, a physicalist picture of the world, the first assumption—that there is mental causation—leads to the problem of causal exclusion famously pointed out by Kim: Given that each mental property is realized by a physical property and that each physical state has a sufficient physical cause, how is it possible that mental states can cause behaviour?