The learning, transfer, and relearning of movement sequences

我们的研究表明，在学习过程中形成了一个动作结构，为迁移和再学习提供了基础。该结构由连接的子序列（或块）组成，经过大量实践后可以相对无缝地执行。重要的是，在开发一个运动结构之后，参与者已经显示出能够有效地重新调整序列，当运动要求增加或减少幅度，力，运动时间，或者序列必须由一组未练习过的效应器产生，前提是传递条件需要对原始序列进行比例变化。此外，在某些情况下，序列结构为学习新的动作序列或重新学习修改过的序列提供了基础（主动促进），但在其他情况下，序列结构实际上可能成为学习或重新学习修改过的序列的障碍（追溯干扰）。这项研究具有重要的实际意义，例如设计旨在提高运动技能的训练和再训练计划，以及医疗保健专业人员试图设计中风或创伤性脑损伤后有效的运动疗法。此外，了解动作序列的表现、学习和修改过程对一般的动作学习理论很重要，因为动作序列构成了我们习得的动作行为的很大一部分。在另外三个实验中，我们将通过观察复杂运动序列的编码方式来继续这条研究路线。对序列表示方式的洞察可能会提供提出运动序列表现和学习的综合理论所需的缺失环节。

Our research has demonstrated that a movement structure develops during the course of learning a movement sequence that provides the basis for transfer and relearning. The structure is composed of concatenated subsequences (or chunks) that can be executed relatively seamlessly after considerable practice. Importantly, after developing a movement structure, participants have showed to be able to effectively rescale the sequence when movement demands require increases or decreases in amplitude, force, movement time, or the sequence must be produced with an unpracticed set of effectors provided the transfer conditions require proportional changes to the original sequence. Further, the sequence structure in some cases provide the basis for the learning new movement sequences or the relearning of modified sequences (proactive facilitation), but in other situations can actually be a hindrance to learning or relearning modified sequences (retroactive interference). This research is important for practical reasons such as designing training and retraining programs designed to improve motor skill and to health care professionals attempting to design effective movement therapies following stroke or traumatic brain injury. In addition, an understanding of the processes involved in performance, learning, and modification of motor sequences is important to theories of motor learning in general because movement sequences comprise a large proportion of our learned motor behaviors. In three additional experiments we will continue this line of research by looking at the way in which complex movement sequences are coded. Insights into the way in which sequences are represented could potentially provide the missing links required to propose a comprehensive theory of movement sequence performance and learning.