Acceso usuarios
This paper surveys studies on the anisotropic perception of orientation in the visual system. The vertical and horizontal orientations are known to be perceived more accurately than oblique orientations. This anisotropy, which called the oblique effect, was present in a large variety of tasks (detection, discrimination, recognition, memorisation) . The discussion about the visual oblique effect (VOE) concerns the nature and level of the processes that generate this effect. It is still unknown whether the VOE stems from low levels of the visual processing system or/and from higher levels. The « low level hypothesis » explains the VOE by the neural properties existing at low levels of visual processing. This hypothesis is supported by three types of arguments : 1 I the VOE is explained by the axis-dependent change in the properties of orientation- selective neurons early in the visual hierarchy (the neurons that are tuned to horizontal and vertical orientations might be more sensitive, narrowly tuned or/and more numerous than those tuned to oblique orientations) ; 2 I the VOE is not influenced by factors independent of the visual stimulus (e.g. : environment, etc.) ; 3 I the VOE follows a retinal reference frame when the head or body of subjects is tilted with respect to gravity. The hypothesis of a « high level VOE » is supported by opposite arguments : 1 I the observations fail to follow the predictions of neuronal models ; 2 I the magnitude of the VOE is influenced by some factors independent of the visual stimulus ; 3 I the VOE follows a gravitational rather than a retinal reference frame in some experiments. The analyse of contradictory results shows that the origin of the visual oblique effects seems to be multi-component and to occur at different levels of processing according to the nature of task.
