Department of General Psychology, University of Padua, Via Venezia 8, Padua, Italy.
Behav Brain Res. 2010 Feb 11;207(1):51-60. doi: 10.1016/j.bbr.2009.09.039. Epub 2009 Oct 2.
Three experiments were designed to investigate visual processing of global and local dimensions of hierarchical stimuli in fish (Xenotoca eiseni). In the first experiment, fish were trained to discriminate between a circle made of circle elements and a cross made of cross elements (consistent stimuli), and tested with a circle made of crosses and a cross made of circles (inconsistent stimuli) to asses their global/local encoding preferences. Fish were also tested for their ability to discriminate single-element shapes. The second and the third experiments manipulated the density of the local elements (Experiment 2) and the size of the global and local shapes of the stimuli (Experiment 3) to assess whether these variables could affect global or local perception of hierarchical visual patterns in fish. In all the experiments, fish showed a global preference irrespective of the density and the size of the stimuli. This preference was not because of an inability to perceive the local constituents of the stimulus, since both fish trained with consistent and fish trained with inconsistent figures showed a clear capacity to discriminate between single-element shapes. Overall, these results suggest that a global preference is not a unique trait of human beings and that differences among different vertebrate species in the global/local strategies of stimulus encoding may be because of different ecological adaptations making initial elaboration of a visual scene in a global or local way more likely.
三个实验旨在研究鱼类(斑鳍蓑鲉)对分层刺激的全局和局部维度的视觉处理。在第一个实验中,训练鱼类辨别由圆形元素组成的圆和由十字形元素组成的十字(一致刺激),然后用由十字组成的圆和由圆形组成的十字测试它们的全局/局部编码偏好,以评估它们辨别单一元素形状的能力。第二个和第三个实验分别操纵局部元素的密度(实验 2)和刺激的全局和局部形状的大小(实验 3),以评估这些变量是否会影响鱼类对分层视觉模式的全局或局部感知。在所有实验中,鱼类表现出全局偏好,而与刺激的密度和大小无关。这种偏好不是因为无法感知刺激的局部组成部分,因为用一致图形和不一致图形训练的鱼类都清楚地表现出区分单一元素形状的能力。总的来说,这些结果表明全局偏好不是人类的独特特征,不同脊椎动物物种在刺激编码的全局/局部策略上的差异可能是由于不同的生态适应,使得以全局或局部方式初始详细描述视觉场景的可能性更大。