Abstract:  The current study reports three experiments to test whether and how unconscious activation of distractors with subliminal presentation, especially at stimulus level, affects response to targets in a letter flanker task. In each experiment, the flanker letters were made unconscious using visually backward masking letters. As classic Flanker tasks, the congruency between target letters and flankers were manipulated to form three conditions – conflicting at stimulus level, conflicting at response level, and non-conflicting. Stimulus conflict referred to trials in which the target and the flankers differed but linked to the same response key, indicating that the competition between the target and flankers occurs at stimulus level. In contrast, response conflict referred to trials where the target and flankers were not only different but also associated with distinct response keys, indicating that the competition between the target and flankers emerges at both the stimulus and response levels. Non-conflict referred to congruent target and flankers trials, used as a baseline condition. Accordingly, the stimulus conflict effect was the difference between stimulus conflict and non-conflict conditions, while the response conflict effect was the difference between response conflict and non-conflict condition. A total of fifty seven participants this study. Experiment 1A was a baseline experiment with supraliminal flankers, in which the classical effects of stimulus conflict and response conflict were observed. However, when the flankers were made unconscious in Experiment 1B using subliminal flankers, a reversed stimulus conflict effect emerged but the response conflict effect maintained. These results were replicated in Experiment 2 when symbolic stimuli were used to reduce the difficulty in forming a stimulus-response association. Likewise, the effects were observed in Experiment 3 with more stimuli and response types were introduced. Crucially, the effect of stimulus conflict dynamically varied along different time windows, while the effect of response conflict was stable across time windows. Altogether, the results provided systematically behavioral evidence for the subliminal activation of distractors that affects target performance at both the stimulus and response levels in a flanker task. The data indicated that the unconscious representation of distractors dynamically influences stimulus processing of targets over time but exerts a stable impact on responses. Our findings, especially the unconscious representation at stimulus level can clarify the mechanism and integrate previous contradicting conclusions of unconscious processing.

Key words: unconscious processing, stimulus representation, cognitive conflict, level of representation, flanker task