The development of separability in visual perception

Abstract

Kolinsky, R., 1989. The development of separability in visual perception. Cognition, Under conditions that do not allow focused attention, reports of illusory conjunctions (i.e. errors that wrongly recombine the features of different objects) constitute evidence of the separate registering of features at an early processing level. The occurrence of illusory conjunctions was used to determine whether there is preattentive analysis of component dimensions (colour and form) and of parts of shapes (triangles and arrows) in young children aged 5 to 8 years. Evidence of yreattentive analysis was found, even for the youngest children, for colour and form but not for parts of shapes. Although developrxdntal effects hardly reached significance, inspection of abilities assumed to affect the illusory conjunctions phenomenon suggests that at the preattentive stage children can integrate spatially separate segments but lack the crrpaci[y to fully analyse connected segments.A central question in the study of perception is the nature of the units into which the sensory world is analysed. Given that in a hierarchy of processing operations different units may be extracted at different stages, the question of whether perception is analytic or integral must be asked separately for the

Boundaries of Separability between Melody and Rhythm in Music Discrimination: A Neuropsychological Perspective

“…Yet, this claim remains to be demonstrated empirically. At least in the visual domain, some evidence has been found that dimensions described as integral by standard criteria, such as using those of Garner, have to be considered as separable when using a more indirect criterion of separability (Kolinsky, 1989), such as the occurrence of illusory conjunctions (Treisman, 1986;Treisman & Schmidt, 1982). Further exploitation of these concepts and methods in the music domain is certainly worthwhile, for its usefulness has been fully demonstrated in the present study.…”

Illusory conjunctions and the cerebral hemispheres

“…First, individual lines of triangles were shown to be analyzed preattentively. Since they conjoin with other forms to create ICs as often as spatially separate segments, they do not seem to "stick together" at all (Kolinsky, 1989;Treisman & Paterson, 1984). Second, early extraction of some kind of emergent or configurational information may account for the preservation of the relative position of the triangle segments."…”

“…More precisely, given the possibly large number of simple misperceptions, ICs are estimated by comparing target FDs made on target-absent experimental trials, in which all the information necessary to perceive the target is present, although not in the same object, with FDs made on target-absent control trials, in which one of the target's attributes is missing. As can be seen in Figure 1, target-absent experimental displays always contained two differently oriented forms (either triangles, as in Figures Yet, even in such a simple detection situation, both CDs and ICs may result from decision bias (Kolinsky, 1988(Kolinsky, , 1989Treisman & Souther, 1986). Thus, rather than directly comparing raw FD and CD scores with regard to laterality effects, we also used the signal detection theory bias-free index of discriminability, d' (see, e.g., Green & Swets, 1966).…”

“…Since, as already noted, ICs may sometimes result merely from trace interference during recall (Kolinsky, 1988(Kolinsky, , 1989Schooler & Tanaka, 1991;Tippens Reinitz, Lammers, & Pitts Cochran, 1992;Treisman, Sykes, & Gelade, 1977), in the present study, we preferred to use a single target-detection task, which minimized memory load. In such a detection situation, ICs are indicated by the occurrence of incorrect detections (henceforth, false detections, or FDs) of the target on target-absent trials.…”

“…How early these operations are in perceptual processing is a matter for discussion (cf. Houck & Hoffman, 1986;Kolinsky, 1988Kolinsky, , 1989). Yet according to Treisman' s (l986b) theory, ICs arise without the intervention of focal attention, since by definition it is precisely attentional serial scanning that allows correct feature integration.…”