Parallel pattern processing and visual agnosia.

Abstract

A series of visual search experiments are reported examining pattern processing in a visual agnosic patient. We examined search for targets defined by: (1) the combination of their features relative to homogeneous distractors; (2) the combination of their features relative to heterogeneous distractors; and (3) a single feature difference relative to the distractors (their orientation). Normal subjects demonstrate evidence of spatially parallel search when combinedfeature targets are detected against homogeneous distractors, and when targets arc defined by a salient feature difference. There are non-linear effects of the number of distractors present, and absent responses can be as fast as present. In contrast, search times for combined-feature targets amongst heterogeneous distractors increase linearly with display size, with the slope for absent responses about twice that for present. The contrast between search for combined-feature targets amongst homogeneous and heterogeneous distractors can be attributed to the effects of grouping between distractors and between distractors and targets (Duncan & Humphreys, 1989 Humphreys & Muller, in press). Grouping between homogeneous distractors facilitates search. An agnosic patient, HJA, showed normal search functions for single-feature targets and for combined-feature targets amongst heterogeneous distractors. However, he was impaired at search for combinedfeature targets amongst homogeneous distractors. This suggests that HJA is selectively impaired at grouping conjunctions of form features. The relations between HJA's agnosia and his problem in the parallel grouping of form conjunctions are discussed, as arc the implications of the work for understanding normal Resume Le document porte sur serie d'experiences dc recherche visuelle et, plus precise'ment, sur le mode dc recherche utilise par un sujet atteint d'agnosie visuelle. Nous nous sommes penches sur la recherche dc cibles selon: (1) la combinaison de traits par rapport a des distractcurs homogfenes; (2) la combinaison de traits par rapport a des distracteurs helcrogcncs et (3) un seul trait distinct par rapport aux distracteurs (I'oricntation). Chez les sujcts normaux, il est evident qu'unc recherche parallele s'effectue sur le plan spatial lorsque des cibles formees

Global shape cannot be attended without object identification.

“…Experiment 4). This result is in agreement with Humphreys and Riddoch's (1987a) finding that HJA manifests no interference effect of the global letter on the identification of local letters; apparently for this patient, local and global forms of information are coded independently (see also Boucart & Humphreys, in press; Humphreys, Freeman, & Müller, in press; Humphreys, Riddoch, Quinlan, Price, & Donnelly, in press).…”

Global shape cannot be attended without object identification.

“…Experiment 4). This result is in agreement with Humphreys and Riddoch's (1987a) finding that HJA manifests no interference effect of the global letter on the identification of local letters; apparently for this patient, local and global forms of information are coded independently (see also Boucart & Humphreys, in press; Humphreys, Freeman, & Müller, in press; Humphreys, Riddoch, Quinlan, Price, & Donnelly, in press).…”

Ventral extra-striate cortical areas are required for optimal orientation averaging

“……”

Ventral extra-striate cortical areas are required for optimal orientation averaging

“…575 It is also interesting to examine the relations between 576 HJA's processing of edge orientations and his brain 577 lesion. Structural and functional imaging indicates that 578 V1 and dorsal extra-striate cortex is relatively spared, 579 and this matches previous behavioural results where 580 HJA shows comparatively normal performance on a 581 range of tests of early vision, including the Efron shape 582 matching test, visual search for a single oriented item 583 and copying tests (Humphreys et al, 1992;Riddoch & 584 Humphreys, 1987). 585 8.…”

Global shape cannot be attended without object identification.

“…Experiment 4). This result is in agreement with Humphreys and Riddoch's (1987a) finding that HJA manifests no interference effect of the global letter on the identification of local letters; apparently for this patient, local and global forms of information are coded independently (see also Boucart & Humphreys, in press; Humphreys, Freeman, & Müller, in press; Humphreys, Riddoch, Quinlan, Price, & Donnelly, in press).…”

Ventral extra-striate cortical areas are required for optimal orientation averaging

“……”

Ventral extra-striate cortical areas are required for optimal orientation averaging

“…575 It is also interesting to examine the relations between 576 HJA's processing of edge orientations and his brain 577 lesion. Structural and functional imaging indicates that 578 V1 and dorsal extra-striate cortex is relatively spared, 579 and this matches previous behavioural results where 580 HJA shows comparatively normal performance on a 581 range of tests of early vision, including the Efron shape 582 matching test, visual search for a single oriented item 583 and copying tests (Humphreys et al, 1992;Riddoch & 584 Humphreys, 1987). 585 8.…”

Global shape cannot be attended without object identification.

“…Experiment 4). This result is in agreement with Humphreys and Riddoch's (1987a) finding that HJA manifests no interference effect of the global letter on the identification of local letters; apparently for this patient, local and global forms of information are coded independently (see also Boucart & Humphreys, in press; Humphreys, Freeman, & Müller, in press; Humphreys, Riddoch, Quinlan, Price, & Donnelly, in press).…”

Ventral extra-striate cortical areas are required for optimal orientation averaging

“……”

Ventral extra-striate cortical areas are required for optimal orientation averaging

“…575 It is also interesting to examine the relations between 576 HJA's processing of edge orientations and his brain 577 lesion. Structural and functional imaging indicates that 578 V1 and dorsal extra-striate cortex is relatively spared, 579 and this matches previous behavioural results where 580 HJA shows comparatively normal performance on a 581 range of tests of early vision, including the Efron shape 582 matching test, visual search for a single oriented item 583 and copying tests (Humphreys et al, 1992;Riddoch & 584 Humphreys, 1987). 585 8.…”