Answering autobiographical questions: the impact of memory and inference on surveys.

Survey questions often probe respondents for quantitative facts about events in their past: "During the last 2 weeks, on days when you drank liquor, about how many drinks did you have?" "During the past 12 months, how many visits did you make to a dentist?" "When did you last work at a full-time job?" are all examples from national surveys. Although questions like these make an implicit demand to remember and enumerate specific autobiographical episodes, respondents frequently have trouble complying because of limits on their ability to recall. In these situations, respondents resort to inferences that use partial information from memory to construct a numeric answer. Results from cognitive psychology can be useful in understanding and investigating these phenomena. In particular, cognitive research can help in identifying situations that inhibit or facilitate recall and can reveal inferences that affect the accuracy of respondents' answers.

Trichromatic color vision with only two spectrally distinct photopigments.

Protanomaly is a common, X-linked abnormality of color vision. Like people with normal color vision, protanomalous observers are trichromatic, but their ability to discriminate colors in the red-green part of the spectrum is reduced because the photopigments that mediate discrimination in this range are abnormally similar. Whereas normal subjects have pigments whose wavelengths of peak sensitivity differ by about 30 nm, the peak wavelengths for protanomalous observers are thought to differ by only a few nanometers. We found, however, that although this difference occurred in some protanomalous subjects, others had pigments whose peak wavelengths were identical. Genetic and psychophysical results from the latter class indicated that limited red-green discrimination can be achieved with pigments that have the same peak wavelength sensitivity and that differ only in optical density. A single amino acid substitution was correlated with trichromacy in these subjects, suggesting that differences in pigment sequence may regulate the optical density of the cone.

Normal endothelial cell density range in childhood.

Specular microscopy of the in vivo corneal endothelium of 214 clinically normal eyes in children ranging from 5 to 14 years of age showed a regular mosaic of hexagonal cells. The cell population density of individuals presented some variation, as it doses in older subjects. Quantitative analysis permitted us to determine the normal range of the endothelial cell count at each age. The mean (+/- SD) value ranged from 3591 +/- 399 cells per square millimeter at age 5 years to 2697 +/- 246 cells per square millimeter for the oldest subjects. Our data show a rapid decrease in cell density up to age 10 years. We estimate from our data a decrease in cell density of 13% between ages 5 and 7 years and an additional decrease of 12% by age 10 years.

On neural signals that mediate induced blackness.

A small patch of achromatic light viewed within a large achromatic surround appears gray or black when the radiance of the surround is well above that of the patch. No single light can match the appearance of the patch because no light in isolation appears blackish; blackness is induced by a second stimulus. The present experiments examine the locus of the mechanisms mediating induction of blackness. They test whether induced blackness can be completely explained by interactions among signals from only one eye (retinal lateral inhibition, for example, though no explicit process is assumed here). If so, a surround affects the appearance of a patch only by modifying a signal that represents the patch at a monocular neural level. This signal may depend on retinal mechanisms and on purely monocular central processes. The alternative hypothesis is that induced blackness depends, at least in part, on a central binocular process driven by more information from each eye than can be carried by a neural signal representing only the patch. Measurements with fused binocular stimuli support the alternative hypothesis, implying retinal mechanisms alone are an incomplete explanation of induced blackness.

On neural signals that mediate brightness.

The brightness of any single light depends on other lights in view. These experiments examine how adapting light affects the brightness of a small incremental patch. The original purpose was to determine the quantitative effect of adaptation on the eye's brightness signal for the increment. However, this was found to be a fruitless enterprise because the results showed there exists no single signal from the eye that encodes brightness of the increment. This means that binocular brightness cannot be explained by any model that combines a signal from the left eye and a signal from the right eye. Instead, at least two independent neural signals from each eye are transmitted to a central locus where multi-attribute information about the complete left-eye stimulus configuration is combined with multi-attribute information about the right-eye stimulus configuration.

Color perception under chromatic adaptation: equilibrium yellow and long-wavelength adaptation.

Observers viewed a thin (0.8-1.3) annulus composed of a mixture of 540 and 660 nm monochromatic lights (denoted delta G and delta R, respectively). The annular mixture was superimposed upon a larger (2.7) 660 nm circular background field. The observer adjusted the radiance of either delta G or delta R so that the annulus appeared a perfect (i.e. neither reddish nor greenish) yellow. In the first experiment, the background and annulus both were presented steadily. The results showed that the background, varied over a range from 10 to 1000 td. always contributed less to the color appearance of the annular test area than would be expected from the simple admixture of lights. The second experiment examined the effect of briefly removing the background-field quanta during the period when the annulus was judged. After several minutes of adapting to the background, the background was momentarily extinguished for 1 sec once every 6 sec; the observer adjusted the radiance of delta R so that during the 1 sec period the continuously presented annular mixture appeared equilibrium yellow. With steady backgrounds, the delta G to delta R luminance ratio decreased with test annulus luminance; for judgments made while the background momentarily was extinguished, the luminance ratio generally increased with annulus luminance. All of the empirical observations can be accounted for quantitatively by a two-process theory of chromatic adaptation; in two processes are (1) gain changes and (2) a restoring signal that tends to drive back toward equilibrium the opponent response resulting from the adapting light. Results from a third experiment, in which the background-off interval was reduced from 1 sec to 500, 200 or 150 msec. also are consistent with this model.

Redness from short-wavelength-sensitive cones does not induce greenness.

According to opponent-colors theory, a reddish surround induces greenness in a central test field. Color-appearance measurements verify this with a long-wavelength reddish surround (660 nm) but not with a short-wavelength reddish surround (440 nm). Surprisingly, a short-wavelength reddish surround shifts the appearance of a test toward redness. Four possible explanations are: (1) stray light from the short-wavelength reddish surround falls in the test area; (2) receptoral sensitivity changes overwhelm induced greenness from the surround; (3) a neural process of assimilation, rather than contrast, to the surrounding light; and (4) short-wavelength-sensitive (S) cones do not contribute to induced redness/greenness. Chromatic cancellation experiments confirm the fourth explanation. There was no change in induced redness/greenness when quantal absorption by only S cones in the surround was varied by 30-fold (using tritanopic metamers), even though varying stimulation of S cones strongly affected the color appearance of the surround. The redness induced by a short-wavelength surround is accounted for by opponent chromatic induction mediated by only middle- and long-wavelength-sensitive cones.

Light spread and scatter from some common adapting stimuli: computations based on the point-source light profile.

A point source of light is not so imaged on the photoreceptor mosaic. The light is distributed over the retina by diffraction, imperfections in the optics of the eye, and scatter. The spatial distribution of light is specified quantitatively by the point-source light profile, which can be used in convolution to determine light spread from more complex visual stimuli. We use the Vos, Walraven and van Meeteren [Vision Res. 16, 215-219 (1976)] light profile to determine spread light in specific regions within the spared area of an illuminated surround, and within a thin spared ring in an otherwise uniformly illuminated circular field. Surrounds of different sizes and rings of various widths and diameters are evaluated.

Color perception under chromatic adaptation: red/green equilibria with adapted short-wavelength-sensitive cones.

Chromatic adaptation can dramatically alter the color appearance of a light. The specific effect of adapting short-wavelength-sensitive (SWS) cones is examined by using two adapting wavelengths that lie on a tritanopic confusion line. The change in color appearance caused by signals from adapted SWS cones is isolated by restricting the wavelengths of the test light to 550 nm or longer. Thus the test negligibly stimulates SWS cones, so their sensitivity does not affect the test's appearance. The results show that adapted SWS cones contribute redness to the appearance of a superimposed test light, while not affecting sensitivity of MWS and LWS cones. Quantitatively, the redness from SWS cones illuminated by a large adapting field approaches physical admixture of test and adapting lights. This is very different from an adapting field that stimulates only MWS and LWS cones which, due to a postreceptoral process, contributes much less redness to a small superimposed test than expected from admixture. The difference between the adapted SWS-cone and the adapted MWS/LWS-cone contributions to the color of a small test explains a surprising result: a bluish-green (491 nm) adapting field contributes redness to a superimposed test light.

The visual photopigments of simple deuteranomalous trichromats inferred from color matching.

Deuteranomalous trichromacy is the most common form of inherited color-vision deficiency. A modern description of its cause is a single abnormality: the normal middle-wave cone photopigment (M) is replaced by a shifted middle-wave pigment (M) that is shared by all deuteranomalous trichromats. This explanation, however, fails to account for the individual differences in color vision observed even within the sub-group of deuteranomals with good chromatic discrimination. An ensemble of color matches is used here to test whether these individual differences reflect differences in the wavelength of peak sensitivity (lambda max) of individual deuteranomals' cone photopigments. The results show variation in both the lambda max and the effective optical density of their cone pigments. The individual differences found in lambda max are in accord with recent molecular biological research that shows individual differences in the genes thought to encode deuteranomalous photopigments.

Color perception under contralateral and binocularly fused chromatic adaption.

Observers viewed a thin (1.0-1.5 degree) annular mixture of 450 plus 660 nm light with the left eye, and a 4.8 degrees circular 660 nm adapting field with the right eye. The annular test was centered upon the adapting field in the fused percept. The contralateral field, which bleached negligible photopigment, caused the test to appear more reddish; however quantitative results reject the hypothesis that a given contralateral 660 nm light simply adds redness to the test. Additional experiments with a separate 660 nm adapting light presented to each eye reveal that fused chromatic adapting fields affect a complex central mechanism. For example, a dim background presented to the same eye as the annular test can reduce the effect of an opposite-eye field, even when the dim test-eye background has no measurable influence in a purely monocular experiment.

Color appearance with sparse chromatic context.

We compared changes in the appearance of a test region caused by introducing an inhomogeneous chromatic background to changes caused by a space-averaged equivalent uniform background. Subjects adjusted a test field presented on a CRT so that it appeared neither reddish nor greenish. Sparse "white" or "green" dots, randomly scattered throughout a "red" background field, caused a large decrease (up to 15 nm) in the dominant wavelength of the red/green equilibrium setting, compared to measurements with a uniform "red" background. A uniform background with the same space-averaged chromaticity and luminance as the complex background had an effect similar to the uniform "red" background. These results contradict theories of color constancy that rely on the "gray world" assumption, and indicate the significance for color perception of individual chromaticities within discrete, noncontiguous regions.

A central binocular mechanism affects chromatic adaptation.

Two experiments explored the role of central binocular mechanisms in color perception. The first experiment examined the effect of adapting to simultaneous, binocularly fused fields. Each eye adapted to a slowly flickering (0.5 Hz) long-wavelength light. The two eyes were adapted either inphase (both eyes stimulated at the same moment) or out-of-phase (only one eye stimulated at any given moment). Both adapting procedures shifted equilibrium yellow toward longer wavelengths, but a significantly greater shift was found when adapting light stimulated both eyes simultaneously. This reveals that a central binocular mechanism affects chromatic adaptation. The second experiment tested whether the binocular mechanism could shift equilibrium yellow measurements made with both eyes (identical, binocularly fused fields presented to each eye) outside of the range of measurements established by left-eye monocular viewing and right-eye monocular viewing. Differences were found between monocular left-eye and monocular right-eye color appearance under conditions of moderate chromatic adaptation, but binocularly fused measurements fell within the range established by the monocular results. This is consistent with the view that central mechanisms serve to keep the two eyes in balance, rather than systematically alter color appearance from colors perceived under monocular viewing.

Color perception with test and adapting lights perceived in different depth planes.

Adapting to a chromatic light can alter the color appearance of other lights in view. The chromatic adapting effect is measured here with the test and adapting field perceived in the same depth plane, or perceived in different depth planes (using stereo disparity). The measurements show only a weak, though consistent, shift in the appearance of the test when adapting field and test are perceived in different depth planes, compared to when they are in the same plane. Adding complexity to the adapting stimulus, in the form of a second chromatic light surrounding the background, alters the appearance of the test but shows no dependence on the depth relations. Overall, there is only a small difference in chromatic adaptation caused by introducing a three-dimensional representation of these stimuli.

Brightness contrast from inhomogeneous surrounds.

The luminance of a test within an inhomogeneous ("checkerboard") surround was adjusted to match the brightness of a comparison patch within a uniform surround. All stimuli were achromatic. Both surrounds had the same space-averaged luminance. With an incremental comparison patch, a test-within-checkerboard at a luminance between the luminances of the brighter and dimmer checks appears dimmer than if viewed within the uniform surround. A decremental comparison patch, however, is matched by a test luminance that is little affected by the inhomogeneity of the surround. In general, the brightness of the test is mediated neither by the space-averaged luminance of an inhomogeneous surround, nor by any equivalent uniform surround, regardless of luminance. We consider alternative models for the brightness of a region that is neither strictly an increment nor decrement with respect to contiguous surrounding surfaces.

Variation in color matching and discrimination among deuteranomalous trichromats: theoretical implications of small differences in photopigments.

Individual differences in abnormal color vision are well known. A fundamental unresolved problem is the great variation in color vision even among those classified as having the same color-vision defect. Several physiological hypotheses have been proposed to account for this variation but little consideration has been given to how (and how much) color matching and discrimination are affected by the posited physiological mechanisms. Advances in molecular genetics have renewed interest in this problem, which is at the foundation of the relation between genotype and phenotype. We report here theoretical Rayleigh ranges (chromatic discrimination) and quantal matches for deuteranomalous trichromats with photopigments in the red/green range that vary in their separation and optical density. The results show there is relatively little loss of discrimination with pigments of normal optical density separated by as little as 2-3 nm. With pigments separated by 4 nm or less, however, optical density can strongly influence discrimination when varied independently in the two types of cone. Moderately lower (or higher) optical density in only one cone-type affects discrimination by altering the shape of the cone's relative spectral sensitivity function. The lack of correlation between Rayleigh-match midpoint and range, which is reported in the literature, may be accounted for by independent variation in pigment separation and optical density.

Chromatic induction: border contrast or adaptation to surrounding light?

Chromatic induction from a surrounding light is measured with an additional remote field outside the surround. Chromatic induction from the surround into a central test field is found to be attenuated by a remote inhomogeneous 'checkerboard', composed of squares at two different chromaticities. A uniform remote field, on the other hand, either at the average or at the most extreme chromaticity of the 'checkerboard', has a weaker effect on chromatic induction than the inhomogeneous field, implying that chromatic contrast within the remote region is a critical factor. The complete set of experiments is accounted for by chromatic contrast gain control: chromatic induction, mediated by a neural signal for contrast at the edge of the test, is attenuated by contrast within the remote region. A contrast gain control set by variation in chromaticity over a broad area can contribute to the stable color appearance of surfaces embedded within complex scenes by minimizing chromatic induction from locally adjacent regions.

Color perception with binocularly fused adapting fields of different wavelengths.

By presenting to one eye a small test superimposed on a background field, and to the other eye only a similar background field of a different wavelength (arranged so that the fused percept is the test centered upon a single fused background), the color appearance of the perceived background can be changed while keeping constant the light stimulating the test eye. Measurements demonstrate that the contralateral field clearly influences the color of the test, even when the left- and right-eye backgrounds are very different in wavelength and illuminance. This change in color appearance cannot be explained (a) by simple contribution of a color signal from the contralateral eye, (b) by the perceived color of the fused background, or (c) by combining the effects of contralateral adaptation (alone) and monocular adaptation (alone). Instead, the central-mechanism response depends on the particular pair of wavelengths that are fused. The results suggest chromatic coding of neural signals arriving at the central locus.

Color perception within a chromatic context: changes in red/green equilibria caused by noncontiguous light.

We measured changes in the color appearance of one light caused by another light presented in a well-separated region. Observers viewed a 1 degrees test field superimposed on a 3 degrees, 540 or 660 nm adapting field (32 or 320 td). The change in appearance due to noncontiguous light was determined by surrounding the 3 degrees adapting field with a continguous 3 degrees i.d., 5 degrees o.d. ring of either 32 or 320 td. The ring was 540, 660 nm or achromatic (tungsten-halogen "white"). The test was an admixture of 549 and 660 nm light, and varied from 6 to 1000 td. The observer adjusted the ratio of 549 to 660 nm test light so the test appeared neither reddish nor greenish. A 540 or 660 nm ring had a chromatic inducing effect on the small test that mimicked a simple surround contiguous with the test. Results with an achromatic ring were more complex: an isolated achromatic ring (no adapting field present) had virtually no effect on the color appearance of the test, but the same achromatic ring surrounding a chromatic adapting field shifted the test toward the color appearance of the adapting light (e.g. introducing a "white" ring surrounding a "green" adapting field shifted the test toward greenness). A thin pencil-width band of "white" light superimposed on a larger 5 degrees adapting field had an effect similar to a "white" 3-5 degrees ring. These results demonstrate (1) strong effects of the remote noncontiguous lights and (2) that the change in color appearance they cause is not a simple function of only the light in the noncontinguous region. The change depends on other lights in view. The visual processes revealed in these experiments are considered in terms of inferred illumination and surface reflectances of objects in natural scenes.

Color perception within a chromatic context: the effect of short-wavelength light on color appearance.

Light at the boundary of a uniform test field (contrast) has a qualitatively different effect on color perception than light in more remote noncontiguous regions (context). Basic properties of color perception with contextual short-wavelength light are assessed here with a 1 degree test field surrounded by either contiguous or noncontiguous 440 or 491 nm light (32 td). Contrasting stimuli are 3 or 5 degrees adapting fields, a thin 1 degree i.d.-2 degrees o.d. (0.5 degree wide) contiguous band, or a large 1 degree i.d.-5 degrees o.d. contiguous surround. Contextual stimuli are a remote 3 degrees i.d.-5 degrees o.d. ring or 0.5 degree wide noncontiguous bands at various distances from the edge of the 1 degree test field (2 degrees i.d.-3 degrees o.d., 3 degrees i.d.-4 degrees o.d., or 4 degrees i.d.-5 degrees o.d. bands). Contiguous surrounds have little influence on color appearance, but remote noncontiguous short-wavelength light strong affects the color of the test field, shifting it toward redness. The shift toward redness increases as a thin 440 nm band is moved farther from the test field (up to 5 degrees), unlike the effect of distance on remote middle- and long-wavelength bands. Measurements comparing the effects of 440 nm and luminance-equated 491 nm light indicate a contribution from S cones.