Detecting gender before you know it: How implementation intentions control early gender categorization

• Gender categorization was reflected in the N170 and P300 components.
• Goal intentions that aimed at controlling the gender bias altered the N170 effect.
• Implementation intentions eliminated automatic gender categorization on the N170.
• Intentions did not affect the P300 component.
• We found neural indicators of nonconscious goal striving.

Gender categorization is highly automatic. Studies measuring ERPs during the presentation of male and female faces in a categorization task showed that this categorization is extremely quick (around 130 ms, indicated by the N170). We tested whether this automatic process can be controlled by goal intentions and implementation intentions. First, we replicated the N170 modulation on gender-incongruent faces as reported in previous research. This effect was only observed in a task in which faces had to be categorized according to gender, but not in a task that required responding to a visual feature added to the face stimuli (the color of a dot) while gender was irrelevant. Second, it turned out that the N170 modulation on gender-incongruent faces was altered if a goal intention was set that aimed at controlling a gender bias. We interpret this finding as an indicator of nonconscious goal pursuit. The N170 modulation was completely absent when this goal intention was furnished with an implementation intention. In contrast, intentions did not alter brain activity at a later time window (P300), which is associated with more complex and rather conscious processes. In line with previous research, the P300 was modulated by gender incongruency even if individuals were strongly involved in another task, demonstrating the automaticity of gender detection. We interpret our findings as evidence that automatic gender categorization that occurs at a very early processing stage can be effectively controlled by intentions.