Project Details
Projekt Print View

Visual and haptic illusions in complex stimulus arrays

Subject Area General, Cognitive and Mathematical Psychology
Term from 2012 to 2022
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 228736743
 
Final Report Year 2022

Final Report Abstract

The T-illusion consists in the overestimation of the length of the T’s undivided line relative to the length of its divided line, when both are equally long, and the Oppel-Kundt illusion consists in the overestimation of textured extents relative to empty or homogeneous ones. Both visual illusions were investigated by means of psychophysical experiments that allow for the separation of response bias – the offset of the point of subjective equality relative to objective zero – and discriminative sensitivity, that is, the ability to nonetheless discriminate lengths correctly. Local symmetries were identified to be essential for the T-effect. As the amount of illusion varied with the type of line from which the T was composed (J-, C-, and S-curves, mimicking handwritten Ts), the origin of the effect does not appear to be adaptation to a letter schema. As the illusion persisted in Ts without expressly drawn T-junctions (e.g., Ts, composed from dashed lines or dots), a more general, nonliterate schema seems to be in operation. Dissecting a T into two separate lines and tilting one of them relative to the other one, depending on the T’s overall orientation, attenuated, annihilated, or reversed the illusion; the two measures added up, which means that they operate independently. Dissection alone was quantitatively equivalent to using the method of adjustment instead of the method of constant stimuli, thus, in terms of cognitive strategies or heuristics, tentatively explaining the dependency of findings obtained with different psychophysical procedures. Adding abutting or nonabutting flanks to a T attenuated the illusion according to the number of gaps in the resulting H- and +-type figures. Arranging Ts in patterns of two enhanced the illusion when the undivided lines were collinear. All of the foregoing effects suggest interactions between orientation-sensitive and end-stopped neurons in primary visual cortex as the neural substrate. In branching patterns, composed from four abutting or intersecting Ts, the illusion vanished when target lengths had to be indicated haptically by spreading thumb and index finger accordingly. Here the suggestion is that the endpoints of target lines are easier to localize in complex patterns as compared to free-floating lines. With bent Oppel-Kundt stimuli, the illusion was attenuated when an empty extent, demarcated by two dots, was oriented vertically, and a dotted extent horizontally, but it was not enhanced in the inverse case – suggesting that the so-called horizontal-vertical illusion (viz., the overestimation of vertical extents relative to horizontal ones) is ineffective in textured extents. The effect replicated in the haptic domain. As the basic illusion in straight extents was at its maximum at a horizontal orientation of the stimulus, it seems to be bound to retinal coordinates. The cause of the illusion may simply be the amount of neural activity elicited by textured versus empty or homogeneous extents.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung