Sequence learning has been the subject of research in various paradigms but has not been investigated for learning deferred imitation of long spatial sequences. In this task no guiding stimuli support the sequence reproduction and all sequence information has to be recalled from memory. We investigate this kind of imitation learning with a task in which a long sequence of spatial positions has to be reproduced without guiding stimuli, either by manual pointing or by ocular fixations. Sequences consisting of 20 positions were acquired after only 25 training trials. The persistence of learned sequences over several days showed that the sequence was retained in long-term memory. A transfer test revealed that the learned sequence was independent of the effector. A detailed analysis of the error distributions of pointing and ocular fixations was performed to characterize the guiding control signal. The independence of the variable position errors from the movement directions as well as the lack of error propagation between successive targets suggest that the reproduction in this learning task was guided by sequential positions rather than sequential displacements.
展开▼
