The article states the necessity of logical explanation of empirical phenomena. This approach is demonstrated on analysis of Stroop interference, because of it is one of the most reproducible and valid eff ects in psychology. It is usually said that interference happens as a result of some mythical constraints of cognitive activity. Authors suggest that this is not the case. Interference is a result of verifying the correctness of performing a paradoxical task: not to realize (ignore) the result of automatically performed actions. It is shown that this explanation is heuristic for describing results of multiple researches. Refs 39.