Event-related potentials (ERPs) reflect sensory, cognitive, and motor processes or mechanisms. By using the classical Go/Nogo-task several ERP researchers claim to have found ERP correlates of a frontal inhibition mechanism. In the Go/Nogo task the subjects have to respond to one stimulus (or stimulus category, Go-stimuli) and to refrain from responding to the other stimuli (Nogo-stimuli). Most of the Go/Nogo ERP studies used visual stimuli. Pfefferbaum, Ford, Weller and Kopell (1985), using visual letter and symbol stimuli, found a P300 with maximum at Pz on Go trials, and equal P300 amplitudes at Cz and Pz for Nogo trials. Moreover, in Nogo trials a negative shift with Fz maximum was superposed on the positive-going flank of the P300. A similar shift was found when the Ss only had to count the Go stimuli. Kok (1986), using visual letter stimuli, found larger late positive components in Nogo than in Go trials at Cz. Moreover, at about 400 ms he found a negativity with maximum at Fz, which was larger on Nogo than on Go trials. Jodo and Kayama (1992), using visual LED stimuli, observed a short negative component with maximum at Fz in Nogo trials, which became larger with higher time pressure. Eimer (1993), using a visual cuing task, also found a negativity in Nogo trials with maximum at Fz, which was larger for attended (cued) than for unattended (noncued) stimuli. Nogo stimuli also elicited a larger P300 than Go stimuli at Fz and Cz. Roberts, Rau, Lutzenberger and Birbaumer (1994), using a warned Go/Nogo paradigm with visual letter stimuli, found a strong enhancement of the P3 at Fz and Cz, and (not explicitly reported) a more negative N2 at Fz for Nogo compared to Go trials. Thorpe, Fize and Marlot (1996), using picture stimuli, found a phasic negativity in Nogo trials which began already at about 150 ms after stimulus onset, though the reaction times were about 450 ms long. Kopp, Mattler, Goertz and Rist (1996b) modified an Eriksen-task by adding Nogo trials. These authors compared ERPs in Nogo-trials that were either specifically primed (by arrows) or unspecifically primed (by rectangles). By this technique, motor-related Go/Nogo differences could be excluded. Nevertheless, a more negative N2 was observed in specifically primed trials, which require specific inhibition. In contrast, Kopp et al. could not find a fronto-central enhancement of the P3 in specifically vs unspecifically primed Nogo trials.

In sum, the studies using visual Nogo-tasks report two effects in Nogo-vs Go-ERPs: (1) a negative shift with maximum at Fz with a latency of 150–400 ms, the “Nogo-N2”, and (2) (less consistently) a positive shift with maximum at Fz and Cz with a latency of 300–500 ms, the “Nogo-P3”.

The Nogo-N2 possibly reflects a frontal inhibition mechanism which is active on Nogo trials. This inhibition hypothesis is e.g. suggested by Jodo and Kayama (1992), who argued that higher time pressure should increase inhibition and hence the Nogo-N2, as well as by Kopp et al. (1996b), who argued that specific (but not unspecific) response priming requires inhibition and hence reveals a Nogo-N2. The inhibition hypothesis is further corroborated by animal studies. These authors trained monkeys to perform hand movements in a visual Go/Nogo task. At about 150 ms after Nogo stimuli, they found a sharp surface negative potential in the dorsal bank of the principal sulcus, a formation in the frontal cortex. The assumption that this potential is linked to inhibition was strongly supported by the second study of these authors. The source of the “Nogo”-potential, i.e., the caudal-dorsal principal sulcus, was stimulated by electrical bursts with different time lags after a Go stimulus. The stimulation caused a delay and even suppression of the motor activity normally occurring after the Go stimulus, and mainly so when the stimulation was given at about 150 ms after the Go stimulus, i.e., the time when the Nogo-potential usually appears after a Nogo stimulus. Hence it appears that the frontal cortex is the source of an inhibition potential, which may be reflected in the Nogo-N2 in humans. However, there is a fundamental problem with this hypothesis: after auditory stimuli a Nogo-N2 is hardly seen. Karlin, Martz and Mordkoff (1969), who recorded from Cz only, rather found a more negative ERP in Go than in Nogo trials in the N2 latency range. Hillyard, Courchesne, Krausz and Picton (1976) found no Go/Nogo difference for the auditory N2 at frontocentral and frontal leads. In an auditory cuing task, Schröger (1993) found a slightly more negative N2 after Nogo than after Go stimuli, if the stimuli were attended (cued). Falkenstein, Koshlykova, Kiroj, Hoormann and Hohnsbein (1995) found no significant Nogo-N2 after auditory stimuli in an audio-visual Go/Nogo task. In sum it appears, that the Nogo-N2 is either absent or very small after auditory stimuli, which is evidence against the inhibition hypothesis.