Unite Mixte de Recherche, Centre National de la Recherche Scientifique, Paris Descartes University, France.
Am Psychol. 2011 Oct;66(7):624-32. doi: 10.1037/a0024038.
Traditionally, characterizations of the macrolevel functional organization of the human cerebral cortex have focused on the left and right cerebral hemispheres. However, the idea of left brain versus right brain functions has been shown to be an oversimplification. We argue here that a top-bottom divide, rather than a left-right divide, is a more fruitful way to organize human cortical brain functions. However, current characterizations of the functions of the dorsal (top) and ventral (bottom) systems have rested on dichotomies, namely where versus what and how versus what. We propose that characterizing information-processing systems leads to a better macrolevel organization of cortical function; specifically, we hypothesize that the dorsal system is driven by expectations and processes sequences, relations, and movement, whereas the ventral system categorizes stimuli in parallel, focuses on individual events, and processes object properties (such as shape in vision and pitch in audition). To test this hypothesis, we reviewed over 100 relevant studies in the human neuroimaging and neuropsychological literatures and coded them relative to 11 variables, some of which characterized our hypothesis and some of which characterized the previous dichotomies. The results of forward stepwise logistic regressions supported our characterization of the 2 systems and showed that this model predicted the empirical findings better than either the traditional dichotomies or a left-right difference.
传统上,对人类大脑皮层宏观功能组织的描述侧重于左右大脑半球。然而,左脑与右脑功能的概念已被证明是过于简化的。我们在这里认为,自上而下的划分,而不是左右划分,是组织人类皮质大脑功能的更有效的方法。然而,目前对背侧(顶部)和腹侧(底部)系统功能的描述都基于二分法,即哪里与什么以及如何与什么。我们提出,描述信息处理系统可以更好地组织皮质功能的宏观层次;具体来说,我们假设背侧系统由预期和序列、关系和运动驱动,而腹侧系统则并行地对刺激进行分类,关注单个事件,并处理对象属性(例如视觉中的形状和听觉中的音高)。为了检验这一假设,我们回顾了人类神经影像学和神经心理学文献中的 100 多项相关研究,并根据 11 个变量对它们进行了编码,其中一些变量描述了我们的假设,而另一些则描述了之前的二分法。逐步向前逻辑回归的结果支持了我们对这两个系统的描述,并表明该模型比传统的二分法或左右差异更能预测经验发现。