Academic journal article Attention, Perception and Psychophysics

Perceptual Learning of Orientation Judgments in Oblique Meridians

Academic journal article Attention, Perception and Psychophysics

Perceptual Learning of Orientation Judgments in Oblique Meridians

Article excerpt

Abstract Fourteen daily training sessions in orientation discrimination of foveal lines in the 45-deg meridian improved thresholds in the trained meridian by an average of 25 % in five observers. A substantial amount of training transferred to the other obliques, but none to the cardinal meridians, with a consequent reduction in the oblique effect. The data were interpreted as showing perceptual learning at two levels: performance facilitation specific to the trained orientation and improved proficiency globally. The failure of the cardinal orientations to share in the benefit is likely to have its origin in the fact that contour orientation in these meridians is so well established that it had already reached maximum hyperacuity thresholds. The judgment of obliques depends much more than the judgment of cardinals on whether the comparison and test stimuli are shown simultaneously or in succession, but this effect is not changed by perceptual training.

Keywords Oblique effect · Meridional anisotropia · Perceptual learning · Short-term memory

Since Mach first observed that the orientations of vertical and horizontal lines can be judged better than those of oblique lines, the oblique effect, as it has since been called (Appelle, 1972), has been widely demonstrated, yet it has resisted a unitary explanation. The small predominance of neurons in the visual cortex that are tuned to cardinal orientations in the alert primate (Zhang et al., 2010) does not suffice as an explanation, because some very elementary visual tasks do not manifest deficits in oblique meridians, whereas others, whose principal neural substrates are obviously more central and distributed than the primary visual cortex, do (Westheimer, 2003).

Though orientation-specific perceptual learning had been the subject of an earlier inquiry in connection with vernier acuity (McKee & Westheimer, 1978), perceptual learning in the task of discriminating line orientation was first demonstrated by Vogels and Orban (1985), who found evidence for such learning from stimuli with an oblique orientation, but not from those with a cardinal orientation. Improvement through training in an oblique orientation was affirmed by Shiu and Pashler (1992), but their evidence for lack of transfer pertained only to retinal location, not to orientation. A detailed study by Schoups, Vogels, and Orban. (1995) failed to reveal any transfer of learning to other orientations or retinal locations; this, together with binocular transfer, hints strongly that the involved circuits are restricted to the primary visual cortex, though evidence of changes of tuning curves in V4 of monkeys suggests the involvement of higher visual areas (Raiguel, Vogels, Mysore, & Orban, 2006). Recently, Tschopp-Junker, Gentaz, and Viviani (2010) provided further evidence that performance in oblique meridians improves through training, but not enough to eradicate the oblique effect. Though all of these studies involved several thousand responses with feedback, spread over several days and using high-contrast stimuli, the test and training procedures differed among the studies.

The conceptual importance for the oblique effect of transfer of training across orientations motivated us to examine the question of whether substantial training of orientation judgments in a single oblique meridian would result in a performance improvement, and if so, whether the improvement would be specific to the trained meridian or attributable to generalized perceptual learning. Training five observers in one meridian, with detailed pre- and posttraining characterization of their orientation discrimination in eight orientations, enabled us to reach answers to these questions. In the process, we were able to shed some light on the difference in the oblique effect between simultaneous and successive expo- sures of comparison and test stimuli (Heeley & Buchanan- Smith, 1992).



Observers had to detect the direction of misorientation of a foveal line presented for 250 ms, using a pair of flanking lines as a comparison (Fig. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.