PAST MIDNIGHT - SEMANTIC PROCESSING IN AN IMPLICIT LEARNING-TASK

Four experiments are reported which examine the nature of representati ons underlying an implicit learning task. When shown a series of clock faces, each bearing a time between 6 and 12 o'clock, subjects subsequ ently show a selection preference for novel clock faces between these times. Furthermore, they show no signs of being aware of the underlyin g rule governing this preference. This effect is also present when the representation of time of day is changed from analogue to digital bet ween learning and test. In the final experiment subjects show no prefe rence for seen over unseen clocks between these critical times. These data suggest that, for this particular task, implicit learning involve s abstract representations.