Nelson D A, Marler P
Animal Communication Laboratory, University of California, Davis 95616.
Proc Natl Acad Sci U S A. 1994 Oct 25;91(22):10498-501. doi: 10.1073/pnas.91.22.10498.
Bird song is a model system for study of the neurobiology, development, and functions of learned vocal communication signals. Research on avian song learning has been dominated by an instructive model of learning--the sensorimotor model. Developmental plasticity is assumed to be based on the incorporation of novel material into the song repertoire. Experimental evidence now indicates an alternative form of plasticity operating in harmony with sensorimotor learning, based on principles of selection, rather than instruction. Song dialects, a common consequence of vocal learning, can be achieved by overproduction of previously memorized songs and selective attrition of those that fail to match the dialect of interacting males. These distinct processes of learning have important implications for study of the neural substrates underlying song production and perception and for efforts to interpret patterns of geographic variation in song.
鸟鸣是用于研究学习性发声交流信号的神经生物学、发育及功能的一个模型系统。对鸟类鸣叫学习的研究一直由一种指导性学习模型——感觉运动模型主导。发育可塑性被认为是基于将新的发声材料纳入鸣叫曲目。现在实验证据表明，基于选择而非指导原则，一种与感觉运动学习协同运作的可塑性的替代形式。方言鸣叫作为发声学习的常见结果，可以通过过度产生先前记忆的鸣叫以及对那些与互动雄性方言不匹配的鸣叫进行选择性淘汰来实现。这些不同的学习过程对于研究鸣叫产生和感知背后的神经基质以及解释鸣叫地理变异模式的努力具有重要意义。
