Gupta Rahul, Ashe James
Brain Sciences Center, Veterans Affairs Medical Center, 11B VAMC, One Veterans Drive, Minneapolis, MN 55417, USA.
J Neurophysiol. 2007 Jan;97(1):738-45. doi: 10.1152/jn.00773.2006. Epub 2006 Nov 8.
The concept of internal models has been used to explain how the brain learns and stores a variety of motor behaviors. A large body of work has shown that conflicting internal models could not be learned simultaneously; this suggests either a limited capacity or the unstable nature of short-term motor memories. However, it has been recently shown that multiple conflicting internal models of motor behavior could be acquired simultaneously if associated with appropriate contextual cues and random presentations. We re-examined this issue in a more complex environment in which the magnitude of the conflicting fields could vary randomly. Human subjects failed to show any evidence of learning the force fields themselves or the magnitude of the forces experienced, even with extended practice. Subjects did adapt to the applied perturbation when the field strength was kept constant but still did not form internal models. Our results show that neither random presentation nor specific contextual cues are sufficient for learning conflicting internal models when the magnitude of the forces is also unpredictable. The data suggest that multiple conflicting internal models cannot be learned in all environments, and provide support for the unstable nature or limited capacity of motor memories.
内部模型的概念已被用于解释大脑如何学习和存储各种运动行为。大量研究表明，相互冲突的内部模型无法同时被学习；这表明短期运动记忆要么容量有限，要么性质不稳定。然而，最近有研究表明，如果与适当的情境线索和随机呈现相关联，多个相互冲突的运动行为内部模型可以同时被习得。我们在一个更复杂的环境中重新审视了这个问题，在该环境中，冲突场的大小可能随机变化。即使经过长时间练习，人类受试者也没有表现出任何学习力场本身或所经历力的大小的证据。当场强保持恒定时，受试者确实适应了施加的扰动，但仍然没有形成内部模型。我们的结果表明，当力的大小也不可预测时，随机呈现和特定的情境线索都不足以学习相互冲突的内部模型。数据表明，在所有环境中都无法学习多个相互冲突的内部模型，并为运动记忆的不稳定性质或有限容量提供了支持。
