negative feedback. The fact that these findings appear to point to a deficit in performance monitoring in alcoholics is at first glance in opposition to our findings; however, it has been argued that the ERN and F-ERN do not represent the same component process (Gehring and Willoughby 2002). In fact, Holroyd and Coles (2002) found that while the ERN increased with learning, the F-ERN decreased with learning (Holroyd and Coles 2002). Furthermore, even when it has been argued that the ERN and the F-ERN are manifestations of the same underlying process (Nieuwenhuis et al. 2004), the relative or subjective relevance of an action moderates the amplitude of the ERN (Luu et al. 2000a, b; Hajcak et al. 2005; Hewig et al. 2007). In our task, the relevance of the responses was not manipulated. Therefore, it is possible that when the emotional or motivational state of alcoholics is affected, as it could potentially be affected in gambling tasks, a deficit may become evident; however, our data do not address the role of motivation or affect on behavioral or electrophysiological performance.
