Department of Psychology, Rice University.
J Exp Psychol Hum Percept Perform. 2011 Oct;37(5):1331-49. doi: 10.1037/a0024330.
Gestalt phenomena are often so powerful that mere demonstrations can confirm their existence, but Gestalts have proven hard to define and measure. Here we outline a theory of basic Gestalts (TBG) that defines Gestalts as emergent features (EFs). The logic relies on discovering wholes that are more discriminable than are the parts from which they are built. These wholes contain EFs that can act as basic features in human vision. As context is added to a visual stimulus, a hierarchy of EFs appears. Starting with a single dot and adding a second yields the first two potential EFs: the proximity (distance) and orientation (angle) between the two dots. A third dot introduces two more potential EFs: symmetry and linearity; a fourth dot produces surroundedness. This hierarchy may extend to collinearity, parallelism, closure, and more. We use the magnitude of Configural Superiority Effects to measure the salience of EFs on a common scale, potentially letting us compare the strengths of various grouping principles. TBG appears promising, with our initial experiments establishing and quantifying at least three basic EFs in human vision.
格式塔现象通常非常强大，仅通过演示就可以证实其存在，但格式塔很难定义和衡量。在这里，我们概述了一种基本格式塔理论（TBG），将格式塔定义为突现特征（EFs）。这种逻辑依赖于发现比构成它们的部分更具可辨别性的整体。这些整体包含可以作为人类视觉基本特征的 EF。随着上下文被添加到视觉刺激中，EF 的层次结构就会出现。从一个单独的点开始，添加第二个点会产生前两个潜在的 EF：两个点之间的接近度（距离）和方向（角度）。第三个点会引入另外两个潜在的 EF：对称性和线性；第四个点会产生包围感。这个层次结构可能会扩展到共线性、平行性、闭合性等等。我们使用组态优势效应的幅度来在一个共同的尺度上衡量 EF 的显著性，这可能使我们能够比较各种分组原则的强度。TBG 似乎很有前途，我们的初步实验在人类视觉中至少确定并量化了三个基本 EF。
