### abstract ###
Prior experiences can influence future actions.
These experiences can not only drive adaptive changes in motor output, but they can also modulate the rate at which these adaptive changes occur.
Here we studied anterograde interference in motor adaptation the ability of a previously learned motor task to reduce the rate of subsequently learning a different motor task.
We examined the formation of the motor system's capacity for anterograde interference in the adaptive control of human reaching-arm movements by determining the amount of interference after varying durations of exposure to Task A. We found that the amount of anterograde interference observed in the learning of Task B increased with the duration of Task A. However, this increase did not continue indefinitely; instead, the interference reached asymptote after 15 40 trials of Task A. Interestingly, we found that a recently proposed multi-rate model of motor adaptation, composed of two distinct but interacting adaptive processes, predicts several key features of the interference patterns we observed.
Specifically, this computational model predicts the initial growth and leveling off of anterograde interference that we describe, as well as the asymptotic amount of interference that we observe experimentally.
Understanding the mechanisms underlying anterograde interference in motor adaptation may enable the development of improved training and rehabilitation paradigms that mitigate unwanted interference.
### introduction ###
The history of prior action in the human motor system is known to influence not only future performance through memory, but also the capacity for future learning.
Interference and savings are two oppositely-directed phenomena that produce this effect.
Interference describes the ability of one task to impair the learning of another, while savings describes the ability of previous learning to enhance future learning.
For example, previous work has shown that after initial learning and subsequent washout of a visuomotor rotation task, relearning is faster than the initial learning, even if the performance levels of the learner at the onset of learning and relearning are identical CITATION CITATION.
Similarly, in a saccadic gain adaptation task, after learning and subsequent opposite-learning such that the motor output returns to pre-learning levels, relearning is also observed to be consistently faster than initial learning CITATION .
Other studies have demonstrated that previous learning can hinder or interfere with future learning CITATION CITATION.
An experimental paradigm commonly used to study interference is the A 1BA 2 paradigm, where a subject is instructed to serially learn Task A, Task B, and then Task A again - often with various time delays inserted between tasks.
In this paradigm, Task B is usually made to be the opposite of Task A. Two types of interference can be studied with this paradigm retrograde interference: how Task B interferes with the memory of Task A 1, and anterograde interference: how the memory of Task A 1 interferes with the subsequent learning of Task B. Note that both retrograde and anterograde interference can affect performance in Task A 2.
Although anterograde interference can often have quite substantial effects CITATION CITATION, it has not received as much attention as retrograde interference in the motor adaptation literature.
This is surprising because retrograde interference tends to have a relatively small effect on performance in the studies where it is reported CITATION, CITATION CITATION, whereas anterograde interference often has substantially larger effects CITATION CITATION.
In fact, several interference studies have been specifically designed to minimize the effects of anterograde interference because they recognized the potential it has for masking retrograde interference CITATION, CITATION.
Acquiring a better understanding of the mechanisms underlying anterograde interference is important not merely to provide greater insight into retrograde interference effects, but because the learning phenomenon is significant in and of itself as the primary cause of interference during motor adaptation.
Anterograde interference has been observed in force-field adaptation studies CITATION CITATION, CITATION and visuomotor rotation studies CITATION, and has been shown to weaken as the time between tasks is increased CITATION.
A recently-proposed computational model for motor adaptation has suggested a possible mechanism for anterograde interference CITATION.
In this model, one internal adaptive process responds quickly to motor error, but rapidly forgets, while another adaptive process learns slowly from motor error, but has good retention.
The contributions of these two processes are combined to generate the net motor output.
In the transition from Task A to Task B, the fast process will quickly learn the new task, while the slow process will be reluctant to follow because of its good retention of the previous task.
The multi-rate model predicts that the residual contribution of the slow process would hinder adaptation to Task B, resulting in anterograde interference.
The model also predicts that as training in Task A is extended, the amount of interference will also increase, but then level off beyond 15 40 training trials in Task A. Here, using a simple AB paradigm to avoid retrograde interference effects, we examine for the first time how the duration of exposure to Task A influences the amount of anterograde interference observed in Task B in order to determine how the capacity for interference is built up with practice.
We then use the predictions of the multi-rate model to determine whether anterograde interference stems from interactions between the different timescales of motor learning.
