Thomas Huxley articulated the “steam whistle hypothesis” over a century ago (1874). It says conscious thought resembles the steam whistle on a train locomotive: It derives from and reveals something about activity inside the engine, but it has no causal impact on moving the train.
Skepticism about consciousness was particularly fueled by Libet’s (1985) research. In his studies, participants watched a highly precise clock and recorded when they made a conscious decision to initiate a finger movement. Brain wave activity showed a sharp increase prior to the conscious decision. Although the interpretations of these findings have been debated sharply (e.g., Mele 2009), many have taken them as further support for the steam whistle theory. Roediger et al. (2008), for example, said Libet’s findings contradict the “naive view” that “conscious intention causes action. Clearly conscious intention cannot cause an action if a neural event that precedes and correlates with the action comes before conscious intention” (2008, p. 208).
We are skeptical of uncaused causes. Hence arguments of the sort exemplified by the above quotation from Roediger et al. (2008)—that if a brain event precedes the conscious thought, then the conscious thought is not a cause of the subsequent behavior—are fallacious. The question is whether the conscious thought is a vital link in the causal chain as opposed to being merely a signal or side effect of the true causes. It is quite plausible, for example, that impulses to act generally originate in the unconscious, but the behavioral outcome depends crucially on what happens when they are contemplated consciously.
Libet (e.g., 2004) proposed that action begins outside of consciousness, but the conscious self can stop an action before it happens. Mele (2009) indicated the fallacy in the Roediger interpretation by making the analogy of a fuse: The existence of a previous and correlated cause (lighting the match) does not rule out a causal role for the fuse in setting off the bomb.