Four memory scanning experiments investigated the effect of the probability of occurrence of one case of a letter (e.g., "A") on response time to the other case of that same letter ("a"). There was no effect: Responses to one case of a letter did not depend on the probability of occurrence of the other case. This finding indicates that facilitation of visual encoding by high probability of occurrence is not caused by increased activation at the level of the name code. Additional results lead to the conclusion that facilitation is caused by activation along specific routes by which visual features activate letter names.