Kaoru Noguchi

The relationship between perceived surface-lightness and perceived illumination

SummaryAn attempt was made to specify the relationship between perceived surface lightness and perceived illumination under the stimulus conditions where different combinations of albedo and illuminance gave 10 approximately equal-luminance levels of a test field (TF) on three different (black-, gray- and white-appearing) backgrounds. Two types of category judgments for TF-lightness and overall illumination were made on the total of 94 TFs by 5 Ss. The results indicated that under the condition where achromatic surface colors appear, the perceptual scission which produces two different perceptual dimensions (lightness and perceived illumination) from one sort of stimulation (luminance) was clearly observed. The relation between the two judgments was consistent with the “lightness-illumination” invariance hypothesis: As lightness judgments (A′) changed from ‘darker’ to ‘lighter’, illumination judgments (I′) shifted form ‘brighter’ to ‘dimmer’, the sum of A′ and I′ being kept nearly invariant for a given luminance of the TF; the psychological relationship between lightness and illumination, A′+I′, changed as a linear function of the photometric combination of albedo and illuminance, log A+log I. It was also found that the albedo of the background was an important factor determining the extent of perceptual scission between lightness and illumination.

Perceptual scission of surface-lightness and illumination: An examination of the gelb effect

SummaryAn attempt was made to examine how the photometric equation: luminance (L)=albedo (A)×illuminance (I) could be solved perceptually when a test field (TF) was not seen as “figure”, but as “ground”. A gray disk with two black or white patches was used as the TF. Illuminance of the TF was changed over 2.3 log units and TF albedo was varied from 2.5 to 8.0 in Munsell value. Albedos of the black- and white-appearing patches were 1.5 and 9.5 in Munsell values, respectively. Two types of category judgments for apparent TF lightness (A′) and apparent overall illumination (I′) were made on the total of 40 TFs (5 illuminances×4 TF-albedos×2 patch-albedos). The results indicated that when the black patches were added to the TF, A′ was indistinguishable from I′ and when the white patches were placed on the TF, A′ and I′ could be distinguished from each other. The Gelb effect was interpreted as a manifestation of such A′–I′ scission. It was concluded, therefore, that as far as the Gelb effect was observed, the perceptual system could solve the equation, L=A×I, in the sense that for a fixed L, the product of A′ and I′ would be constant.

An experimental test of Jameson and Hurvich’s theory of brightness contrast

Two hundred and fifty-two brightness functions for seven simultaneously presented black, gray, and white squares on black, gray, and white backgrounds approximated power curves with positive exponents when illuminance was varied in seven steps over 1.9 log fL. Mean exponents were significantly larger for whiter vs blacker squares and for squares on the white vs the gray background. Mean exponents also increased more for squares on white than on black or gray backgrounds. Finally, a white replacing a black background elicited increasingly larger decremental responses as the squares varied from black to white. Jameson and Hurvich’s opponent-process theory, tested by the experiment, was not predictive.

Illusions of filled lateral and angular extent

SummaryThe common principle of the illusions found by Helmholtz and Botti is the overestimation of a filled extent. They differ in that Helmholtzs illusions of filled angular extent can be accounted for in terms of spatial two-line interactions, whereas Bottis illusions of filled lateral extent cannot. The former observation is consistent with the assumption of interacting orientation-detectors in the visual cortex. The Botti illusion probably depends on global computation in the visual system.

An attempt to unify the size-assimilation and size-contrast illusions

ZusammenfassungAn einer Reihe von Varianten der sog. Kontrast-Täuschung nach Titchener wird nachgewiesen, daß zwischen dieser Täuschung und der sog. Größen-Angleichung eines Kreises an einen etwas größeren konzentrischen Kreis nach Delboeuf ein innerer Zusammenhang besteht. Auch die „Kontrast“-Täuschung scheint in Wirklichkeit vorwiegend auf Angleichungs-Wirkungen zu beruhen, bei denen die gedachten (inneren, aber in gewissem Maß auch äußeren) Ankreise an die umgebenden Figuren maßgeblich sind. Dieselbe angleichende Wirkung haben auch die gedachten kreisförmigen Verbindungen zwischen den Enden gleichgroßer radialer Ansätze an den Hauptkreis.

Perceptual constancy and the system of reference

Summary1.Different conceptions of the perceptual constancy were discussed, and a new methodology was proposed to deal with the constancy in connection with the system of reference.2.Methodologically, two types of constancy were distinguished from each other: (a) constancy obtained by the comparative judgment in the two-stimulus-comparison situation (as has been studied traditionally) and (b) constancy obtained by the absolute judgment in the single-stimulus situation (as is the case in everyday experience). Some studies on the constancy of the latter type were reviewed.3.It was proposed that any change in the condition of stimulus presentation may transform the system of reference in an extent which correponds to the degree of constancy.4.This proposition was supported by experiments which treated the absolute judgment on the length of pencils, the shape of rectangular cards, or the lightness of paper.5.The implications of the constancy were discussed in terms of (i) segregation of stimulation, (ii) perceptual constancy vs. thing constancy, (iii) comparative vs. absolute, and (iv) transient standards vs. lasting norms.Zusammenfassung1.Verschiedene Auffassungen über die Wahrnehmungs-Konstanz werden erörtert, und eine neue Verfahrensweise wird vorgeschlagen, um das Konstanz-Problem im Zusammenhang mit dem Problem der Bezugssysteme zu behandeln.2.Je nach der Art der Feststellung werden zwei Typen der Konstanz voneinander unterschieden: a) Konstanz, die man durch ein Vergleichs-Urteil in der üblichen Versuchs-Situation erhält, in welcher zwei Gegenstände einander gegenübergestellt werden; b) Konstanz, die man durch absolute Urteile über einen einzelnen Gegenstand erhält, wie es im täglichen Leben vorwiegend der Fall ist. Es wird von Untersuchungen über die Konstanz nach der zweiten Art des Vorgehens berichtet.3.Es wird vermutet, daß jede Änderung in der Art der Darbietung der Reiz-Objekte das Bezugssystem verschieben kann in einem Ausmaß, das dem Grad der Konstanz entspricht.4.Diese Vermutung wird unterstützt durch Versuche, in denen absolute Urteile über die Länge von Bleistiften, die Form von Postkarten und die Helligkeit von Papieren verlangt wurden.5.Die Folgerungen für die Theorie der Konstanz werden erörtert unter folgenden Gesichtspunkten: a) „gegabelte Wirkung“, b) „Wahrnehmungs-“ und „Ding“-Konstanz, c) Vergleichsurteil und absolutes Urteil, d) vergängliche (Augenblicks-) Maßstäbe und überdauernde Normen.