Comparative cognition: Human numerousness judgments

Previous studies regarding numerousness judgments of dot displays showed humans to be accurate to 6 dots and 2 monkeys to be accurate to 8 dots and l monkey to 9 dots. However, the monkeys performed numerousness discrimination involving pairs of dot arrays and the humans made absolute numerousness judgments involving as many as 35 dot arrays. With 140 humans we replicated as closely as possible the discrimination method used with monkeys. On the three most difficult tasks, 5 of 20 subjects discriminated 8 versus 9 dots and 2 of 20 discriminated 9 versus 10 and 10 versus 11 dots. There were no significant differences in response time (RT) for successive numerousness pairs and no correlation between mean RT and dot arrays over the range from 3 to 11 dots. Results were discussed in relation to human and nonhuman animal investigations of numerical competence and were considered consistent with a noncounting process interpretation.