Color Perception: How Light and Context Shape What We See

Author: Angela Bogdanova · Published: October 16, 2026 · Editorial Policy

Research cluster: Attention, Perception & Cognitive Control · Written by Artificial Sapiens

Color perception is the visual process through which an observer experiences and distinguishes colors using information about the light reaching the eyes, the surrounding scene, and the current state of visual adaptation. The colors we see emerge from an interaction among illumination, the spectral properties of surfaces, photoreceptor responses, visual processing, and interpretation of the scene. The same surface can look different under different lights; the same patch of light can look different against different backgrounds. Those everyday facts provide the starting point for experimental psychology's investigation of color. The International Commission on Illumination (CIE) explicitly defines perceived color in terms of hue, brightness or lightness, and colorfulness or related attributes, while recognizing the roles of spatial context, adaptation, and prior experience (CIE, 2020).

Color perception serves the recognition of objects, materials, and changes in an environment. An observer may recognize ripe fruit, distinguish painted surfaces, notice a shadow, or judge whether an object is the same one seen earlier. Such judgments depend on more than registering a wavelength. A major review of object color, color constancy, and categories treats color as a functional component of natural-scene perception and object recognition (Witzel & Gegenfurtner, 2018). This article explains how that perceptual achievement is possible, why it sometimes fails, how scientists measure it, and which theoretical questions remain open.

What Is Color Perception in Psychology?

In psychology, color perception refers to the experience and discrimination of color produced by the visual system in particular viewing conditions. Researchers distinguish the physical stimulus, the sensory response, the perceptual appearance, and the judgment expressed by the observer. A measured spectrum tells us about light; cone responses describe the earliest stages of visual coding; an observer's color match or verbal description reports a particular aspect of experience or task performance. These levels are related, but each supports a different kind of scientific claim.

The familiar question "What color is this object?" often hides several distinct questions. What spectral distribution reaches the eye from the object's surface? Which lights and surfaces surround it? What color does the patch appear to be at this moment? What stable material color does the observer attribute to the object? Which name would the observer use? An experiment that measures one of these outcomes cannot automatically be treated as measuring all the others (Smithson, 2005; Foster, 2011).

Color as a perceptual attribute

Hue describes a perceived quality such as reddish, yellowish, greenish, or bluish. Lightness describes a surface's perceived position between dark and light in relation to a similarly illuminated white. Brightness describes the perceived intensity of a visual stimulus. Saturation and chroma characterize related aspects of how colorful a stimulus appears, though their technical definitions depend on the reference and viewing mode. Color appearance models distinguish these attributes because two stimuli can differ in one while appearing similar in another (CIE, 2020).

The word color therefore covers a structured family of perceptual measurements. Two reds can share approximately the same hue yet differ dramatically in lightness and chroma. A material may look gray in dim light and chromatic in daylight. Describing these differences precisely makes it possible to identify the stimulus dimension responsible for a reported effect.

Surface color, light color, and object color

A small isolated patch viewed against darkness may be experienced as a light or aperture color, while a colored region embedded in a scene may be seen as a reflective surface. The same measured light can produce different percepts under these different viewing modes. Surface reflectance concerns how a material reflects different wavelengths, whereas an illuminant is the light incident on the material. Perceived object color is an interpretation of the whole situation, shaped by cues that may suggest what belongs to the surface and what belongs to the lighting (Hurlbert & Yu, 2025).

How Light Becomes a Color Signal

Wavelength, spectral distribution, and illumination

Visible light occupies a portion of the electromagnetic spectrum. Its spectral power distribution specifies how much radiant energy is present at different wavelengths. A narrowband light concentrates energy in a relatively restricted interval, while daylight and most artificial light sources contain mixtures of wavelengths. For a matte surface under simple conditions, the spectrum arriving at the eye depends approximately on the product of the illuminant's spectrum and the surface's spectral reflectance; geometry, gloss, transmission, and multiple reflections can complicate that relationship (Foster, 2011).

This is the central ambiguity of object color. A surface might reflect little short-wavelength light because its material absorbs those wavelengths or because the light illuminating it contains little short-wavelength energy. The retinal signal alone does not explicitly label the cause. Natural scenes provide contextual clues, and the visual system uses those clues to help disentangle surface properties from the illumination (Smithson, 2005).

Cone photoreceptors and trichromatic coding

In ordinary daytime vision, three classes of cone photoreceptors provide the principal inputs for human color discrimination. They are commonly labeled S, M, and L according to their relatively short-, medium-, and long-wavelength spectral sensitivities. Each class responds across a broad and overlapping range rather than to only one wavelength or one named color. The nervous system receives patterns of relative activity across cone classes and processes those patterns to distinguish stimuli (Gegenfurtner & Kiper, 2003).

The historical trichromatic account associated with Young and Helmholtz captured an essential constraint: normal human color matching can generally be described in a three-dimensional space under specified viewing conditions. Modern science locates that constraint in the three cone classes, with important subsequent neural transformations. It does not follow that each cone is a tiny red, green, or blue detector, or that individual perceived colors can be identified with activity in one cone class.

Opponent processing

Cone signals are recombined into approximately opponent channels. Important dimensions compare L- and M-cone activity, compare S-cone activity with combinations of L and M, and encode luminance-related signals. This organization helps explain red–green and blue–yellow opponent relationships, afterimages, and certain forms of chromatic contrast. Retinal and cortical circuits also preserve and transform color information in ways that are more complex than a single three-channel diagram (Gegenfurtner & Kiper, 2003).

Trichromacy and opponency describe connected stages of a visual system. Trichromacy specifies constraints on photoreceptor input and color matching; opponent processing describes important transformations of those inputs. Context-dependent color appearance requires further consideration of spatial relationships, adaptation, and the interpretation of scenes.

Rods, low light, and changes in appearance

Cones support the richest color discrimination in well-lit conditions. Rods become increasingly influential as illumination falls, and visual sensitivity changes as the system moves from photopic through mesopic toward scotopic conditions. The resulting shift affects the visibility and relative brightness of different wavelengths, while color discrimination becomes poorer at very low light levels. The familiar observation that certain blues appear relatively more conspicuous than reds in dim light illustrates a change in visual sensitivity, rather than a change in the physical paint.

Brightness, contrast, pupil size, adaptation time, and the surrounding scene are all relevant when comparing colors across light levels. A controlled experiment must specify or monitor these variables; otherwise a difference attributed to color might be driven partly by luminance or visibility (Gegenfurtner & Kiper, 2003).

Why Two Different Spectra Can Look Like the Same Color

Color metamers are stimuli with different spectral distributions that produce matching cone responses for a particular observer under particular viewing conditions. Their existence demonstrates that color appearance is not a unique readout of a physical spectrum. A display can reproduce a wide range of color experiences by combining a limited number of primaries, even though the spectra it emits differ from those reflected by natural materials. Such matches are conditional: a change of illuminant, observer, or measurement conditions can reveal differences.

Metamerism is distinct from color constancy. Metamerism concerns different spectra producing the same or approximately the same color match. Color constancy concerns maintaining relatively stable judgments of a surface when its retinal spectrum changes under different illuminants. Both phenomena reveal that the relationship between light and appearance must be studied at the level of the observer, rather than deduced from one spectral measurement alone (Witzel & Gegenfurtner, 2018).

Why Surrounding Colors Change What We See

Simultaneous color contrast

An identical target patch can appear different when surrounded by different colors. A neutral gray against a greenish background may appear relatively reddish, while the same neutral patch against a reddish field may appear relatively greenish. This is simultaneous chromatic contrast: the perceived qualities of a region depend in part on nearby chromatic relationships. It is a robust class of contextual effects, although its magnitude and direction vary with stimulus layout, spatial scale, adaptation, and task (Smithson, 2005).

A simple demonstration places two identical digital color patches on contrasting backgrounds. The two targets share exactly the same pixel values, yet observers may judge them as unequal. Such displays illustrate contextual processing, but a screenshot is not a calibrated psychophysical measurement. Screen profiles, brightness settings, ambient light, and viewing angle can all alter what reaches the eye.

Chromatic induction and spatial organization

Chromatic induction is the general influence of other regions of a visual scene on the color appearance of a target. Local edges, adjacent regions, more distant surrounds, and the organization of surfaces can all matter. Contrast and assimilation are different possible directions of the effect: in some arrangements a target shifts away from the surround's appearance; in others it appears to take on attributes of neighboring regions. A blanket rule that surrounding green always makes an object look red would therefore be inaccurate.

Perceptual grouping further changes which image regions are assigned to the same object, background, illumination, or transparent layer. A shadow boundary may be interpreted as a change in lighting, whereas an object edge may be interpreted as a change in material. These interpretations can affect how the same local chromatic signal contributes to perceived color. The general scientific issue connects color perception to visual organization without requiring that every context effect be explained by deliberate reasoning (Witzel & Gegenfurtner, 2018).

Adaptation and afterimages

Viewing a sustained colored stimulus changes the sensitivity of visual mechanisms over time. Subsequent neutral surfaces may appear shifted toward an opponent color, giving rise to familiar negative afterimages. These effects provide experimental evidence that the visual system continually adjusts its response to stimulation. Adaptation may act over multiple time scales and stages of processing; the afterimage experience is one consequence, rather than the full definition of adaptation (Gegenfurtner & Kiper, 2003).

Chromatic adaptation is also important when an observer moves between different lighting environments. A white sheet under tungsten-like warm lighting can initially appear different from the same sheet under cool daylight; after adaptation, its apparent neutrality may be partly restored. The amount of restoration depends on viewing conditions and cannot be assumed to reach a single universal percentage (Foster, 2011).

Color Constancy: Why Objects Often Keep Their Color Across Lighting Changes

Color constancy is the tendency for perceived surface color to remain relatively stable when illumination changes the light reflected from the surface. A white page continues to look broadly white at noon and indoors under warm lighting, and a red apple tends to remain recognizable as red in sunlight and shade. This achievement is central to natural object recognition because the spectrum reaching the eye varies throughout the day and across environments (Morimoto, 2024).

Constancy is graded and task-dependent. One experiment may ask whether a participant identifies a surface as the same material, while another asks for an immediate appearance match. A person's material judgment may be stable even when the color appearance changes perceptibly. The strength of constancy therefore depends on the precise measurement and should be reported as a result under defined viewing conditions (Foster, 2011).

The illuminant–reflectance problem

Because illumination and reflectance contribute jointly to the spectrum reaching the retina, the visual system faces an inverse problem. Inferring the surface from the retinal input requires exploiting additional information and assumptions about the scene. The problem is mathematically underdetermined when considered without constraints. Stable perception is possible because real visual environments contain regularities, including relations among neighboring surfaces and patterns of illumination (Smithson, 2005).

This explains why color perception is so dependent on context. An isolated patch provides fewer clues about illumination than an object shown among familiar surfaces with shadows and highlights. Yet the amount of helpful information differs across scenes; more context is not automatically equivalent to perfect constancy.

Local contrast and spatial comparisons

One proposed contribution to constancy is the comparison of responses to neighboring regions. Under some illuminant changes, relationships among cone excitations from different surfaces can remain more stable than the absolute signal from any one surface. Spatial comparisons may thus supply cues to material differences despite changes in light. Such mechanisms form one family of models, not a comprehensive account of every scene (Smithson, 2005).

Identifying the illumination

Another proposal emphasizes how the observer estimates or discounts the scene illuminant. The illuminant can be inferred from multiple cues, such as common shifts across surfaces, specular highlights, local contrast, or familiar objects. Contemporary evidence shows that judging an illuminant and judging a surface are related tasks but need not succeed or fail together. A 2025 experiment found that certain local contrast and highlight cues improved discrimination of surface-reflectance changes, while successful color constancy did not simply depend on explicitly identifying the illumination change (Virtanen, Olkkonen, & Saarela, 2025).

A 2025 review emphasizes that the perception of illumination color itself remains an active scientific question. Experimental progress has not settled whether perceived surface color always relies on a separately represented perception of light color. The question requires research designs that distinguish judgments about illuminants from judgments about materials (Hurlbert & Yu, 2025).

What color constancy cannot guarantee

Constancy is often incomplete. Under unusual spectra, conflicting lighting cues, limited scene information, or unfamiliar materials, an object's apparent color can change substantially. The visual system also faces physically different surfaces that produce similar signals under one illuminant, making reliable identification difficult under a new one. Such failures help reveal the mechanisms normally supporting stable perception.

A color-constancy experiment conducted with simplified displays should therefore be interpreted at the level of the tested stimuli, illumination manipulation, judgments, and observers. Direct extrapolation to all photographs, all material surfaces, or all lighting environments is not justified (Foster, 2011).

Major Scientific Accounts of Color Appearance

Trichromatic and opponent-process theories

The historical trichromatic and opponent-process traditions remain foundational because they constrain the representations available to the visual system. The former relates color matching to cone-mediated encoding; the latter explains a substantial portion of the organization of color discrimination and adaptation. Both are supported by convergent psychophysics and visual physiology, but their basic formulations do not by themselves predict every effect of complex scenes (Gegenfurtner & Kiper, 2003).

Retinex and spatial-comparison approaches

Edwin Land's Retinex theory emphasized spatial relationships across an image and the problem of maintaining surface appearance under changing illumination. It helped shift research attention toward the organization of the whole scene. The original account is historically important, while later experimental and computational models differ in their mechanisms and predictions. Retinex should be presented as a major theoretical lineage rather than a universally validated final explanation (Land, 1977; Smithson, 2005).

Adaptation and gain-control accounts

Other models emphasize changes in channel sensitivity or gain as the visual system adjusts to its recent input. Such accounts help explain shifts in neutral points and some illuminant-related changes. Their explanatory scope depends on adaptation timescale, spatial context, stimulus spectrum, and the neural site of adjustment. Reviews consistently discuss adaptation as an important contribution to color constancy while also finding that adaptation alone does not settle all observed effects (Foster, 2011).

Probabilistic and scene-inference accounts

Probabilistic theories formalize how a visual system might combine sensory evidence with assumptions about typical surfaces, illumination, and scene structure. Such models can clarify why ambiguous signals support multiple interpretations. The value of a model depends on its quantitative predictions and success across experiments. A fitted model, however, does not establish that the brain literally implements the proposed algorithm or that every observer uses identical prior assumptions (Witzel & Gegenfurtner, 2018).

These accounts are best treated as competing and sometimes complementary descriptions at different levels of analysis. The major unresolved question is how to connect well-characterized photoreceptor and early neural coding to reliable, context-sensitive judgments of real objects.

Famous Demonstrations and What They Actually Show

The dress photograph

A widely shared photograph of a dress led different observers to report markedly different colors, commonly blue and black versus white and gold. Controlled investigations found differences in matched colors, variable interpretations, and evidence consistent with ambiguity about illumination. Experiments also showed that adding explicit lighting cues could shift reported appearance (Lafer-Sousa, Hermann, & Conway, 2015).

The demonstration illustrates that observers can assign different interpretations to a highly ambiguous photographic scene. It does not establish that one group has defective vision, that all differences reduce to personality, or that every color disagreement originates from the same mechanism. Photographic pixel values, the material color of the real dress, and the colors experienced by each viewer answer different questions.

Identical patches against different backgrounds

In simultaneous-contrast demonstrations, the target is held constant while its surround changes. This experimental manipulation makes it possible to attribute a perceptual difference to context under controlled viewing conditions. Stronger conclusions require quantitative matching, randomized presentation, and control of luminance and adaptation. Online demonstrations have educational value, but their appearance varies across displays and viewers (Smithson, 2005).

Changes between daylight and artificial illumination

Viewing a collection of surfaces under two illuminants allows researchers to test whether color judgments remain stable. A change in illuminant alters reflected spectra even when the material remains unchanged. Comparing responses across these conditions reveals the degree of constancy, but the answer depends on how the participant reports color, which surfaces are used, and whether the viewing environment supplies interpretable illumination cues (Foster, 2011).

How Psychologists Measure Color Perception

Color matching and discrimination thresholds

In a matching task, an observer adjusts a comparison stimulus until it appears similar to a target in a defined respect. Researchers can measure appearance matches across illumination changes or backgrounds. Discrimination tasks instead measure whether two stimuli can be reliably distinguished, often using forced-choice procedures and psychometric functions. Thresholds summarize sensitivity under particular conditions; they do not describe an entire person's visual life or clinical status.

Asymmetric matching and achromatic adjustment

Asymmetric color matching asks observers to match a target under one viewing condition to a comparison under another. It is useful for studying constancy because the stimuli are embedded in different illumination or adaptation contexts. Achromatic adjustment asks the participant to modify a stimulus until it appears gray or neutral. Such procedures estimate specific perceptual outcomes but can produce different indices of constancy, so results should not be equated without considering methods (Foster, 2011).

Color naming, object identification, and categorical judgments

A naming task asks which linguistic category an observer selects. It provides information about classification and communication, with possible contributions from lexical access, instructions, and culture. An object-identification task asks whether a particular surface or material is recognized. Naming and matching can dissociate, demonstrating why evidence about color words should not automatically be treated as evidence about early sensory processing (Lindsey & Brown, 2021).

Experimental control and measurement validity

A rigorous color study controls or reports illuminant spectra, display calibration, background luminance, viewing geometry, adaptation, task instructions, observer selection, and the statistical treatment of repeated judgments. Researchers must also distinguish physical chromatic differences from perceived differences. The method determines what a result means: a color label, speeded decision, discrimination threshold, and object-color match are related but nonidentical outcomes.

Participant variability, small samples, device differences, and laboratory simplification can influence reproducibility. More naturalistic scenes improve ecological relevance, but they also create additional factors to control. The strongest claims depend on convergent outcomes across carefully characterized procedures (Smithson, 2005; Witzel & Gegenfurtner, 2018).

Color Categories, Language, and Cultural Context

People discriminate many more colors than they ordinarily name with basic color terms. Languages divide the color continuum using different vocabularies and boundaries, while showing broad cross-language regularities. A major review argues that perceptual constraints and communicative needs jointly help explain the organization of lexical color categories (Lindsey & Brown, 2021).

Laboratory studies have reported effects associated with color naming or category boundaries in tasks such as rapid discrimination and search. A 2025 scoping review examined 72 experimental papers and highlighted substantial methodological heterogeneity, including differences in task demands, attention, cognitive load, and access to verbal labels. That evidence supports careful study of language–perception interactions while limiting sweeping conclusions that language fixes every aspect of color experience (Kapelle & Flecken, 2025).

A faster response to two chips with different category names may reflect perception, attention, decision processes, verbal encoding, or several of these together. Experimental designs that interfere with verbal rehearsal or manipulate the accessibility of labels help test competing interpretations. Cultural variation is thus a substantive topic in the psychology of color, but the particular outcome and mechanism must be specified rather than assumed.

Individual Differences and Color Vision Deficiency

Color perception varies across observers. Differences can arise from photopigment variation, optical media, adaptation history, age, experience, viewing conditions, and the demands of a given task. Common inherited color vision deficiencies change the discrimination of certain color combinations, especially along red–green dimensions; other forms affect blue–yellow discrimination or produce broader changes. The National Eye Institute distinguishes these conditions and describes established clinical evaluation procedures (NEI, 2025).

Color vision deficiency is a visual characteristic or clinical finding, not a diagnosis derived from disagreeing about a photograph or performing slowly on a psychology experiment. If color discrimination changes unexpectedly, an eye-care professional can assess possible ocular, neurological, or medication-related causes. Clinical testing may use pseudoisochromatic plates, anomaloscope procedures, or structured hue arrangements, with selection depending on the assessment question (NEI, 2024).

Design and educational materials can accommodate variation through redundant cues: clear text labels, differences in shape or position, adequate contrast, and avoidance of color as the sole signal. These approaches address access to visual information without implying that ordinary color variation identifies a psychiatric or cognitive disorder.

Everyday Consequences of Color Context

Screens, photographs, and lighting

A digital image is shown through a particular display, whose emitted light depends on its calibration, brightness, color management, and environment. A photograph also encodes the camera's processing decisions and assumptions about exposure and white balance. Two devices showing nominally the same image can provide different retinal stimuli. A photograph of a colored object therefore cannot reliably substitute for direct physical measurement of its reflectance.

In design and imaging, colors need to be evaluated under relevant viewing conditions. A product photograph, an interior paint sample, and a warning signal each present different tasks: identifying the material, judging an aesthetic appearance, or detecting a critical contrast. Research on color constancy suggests the practical importance of realistic illumination and surroundings, rather than a universal recipe that eliminates contextual effects (Hurlbert & Yu, 2025).

Materials, shadows, and object recognition

Color helps distinguish surfaces and materials, but its information depends on illumination. The visual system's partial stability across light changes supports reliable recognition, while shadows, highlights, gloss, translucency, and mixed lighting can introduce ambiguity. A material's perceived color cannot always be assessed from one isolated pixel. Scene context provides useful information, and controlled comparison is particularly important when consistency matters (Morimoto, 2024).

What color psychology claims require

Findings about color perception do not by themselves establish that particular colors universally cause emotions, improve concentration, or change clinical outcomes. Those are separate hypotheses involving affective responses, behavior, context, and study design. Evidence about chromatic contrast and constancy addresses visual appearance and identification; emotional and behavioral effects require their own empirical literatures. Researchers and practitioners should keep these outcomes distinct.

What Is Well Established, and What Remains Open?

Established results include the involvement of three cone classes in typical photopic color vision, the importance of opponent signal transformations, metamerism, adaptation, simultaneous chromatic context effects, and graded color constancy. The existence of observer differences and variation in color naming is also well supported by convergent research (Gegenfurtner & Kiper, 2003; Witzel & Gegenfurtner, 2018).

Important open questions concern the exact contribution of distinct neural computations to appearance in complex natural scenes, the relationship between illumination perception and surface color, the generalization of simplified models to real materials, and the stage at which language changes a given task outcome. A method can establish that a contextual effect exists without deciding among every mechanism capable of producing it. Reviews published in 2025 continue to identify unresolved issues in illuminant interpretation and language-linked experiments (Hurlbert & Yu, 2025; Kapelle & Flecken, 2025).

Frequently Asked Questions

Why does the same color look different on different backgrounds?

A color patch is processed in relation to neighboring regions and the visual system's adaptation state. Changing the surround can alter its appearance through chromatic contrast and other induction effects, even if the patch's physical values are unchanged. The amount of change depends on stimulus geometry and viewing conditions (Smithson, 2005).

Why does a white object still look white in yellowish indoor light?

The visual system uses adaptation and scene relationships to partially compensate for changes in illumination, a capacity described as color constancy. The compensation varies with context and is often incomplete (Foster, 2011).

Do we see colors in objects or in light?

The light reaching the eyes supplies the stimulus, and perceived colors arise through visual processing. Observers can experience the color of a luminous source or attribute color to a reflective surface; the same physical signal may support different appearances depending on context (CIE, 2020).

Why did people disagree about the colors of the dress?

The photograph provided ambiguous cues about illumination, and observers produced different stable interpretations and appearance matches. Experimental manipulations of lighting cues could change reports. The disagreement illustrates variation in a perceptual inference under ambiguity, not a simple test of visual health (Lafer-Sousa et al., 2015).

Do color names change what people see?

Language and category labels can influence performance in certain color tasks. The mechanisms and effect sizes vary with attention, verbal access, and experimental design. The strongest general conclusion is that color naming and perceptual processing interact under identifiable conditions (Kapelle & Flecken, 2025).

Is color perception the same as color vision deficiency?

Color perception is a general visual process studied in psychology. Color vision deficiency refers to particular differences in color discrimination assessed through visual tests and, where appropriate, clinical examination. A single unusual percept or online demonstration does not establish a deficiency (NEI, 2024).

Can color perception be measured objectively?

Researchers can precisely measure spectra, cone-related stimulus properties, choices, thresholds, and color matches. A subjective color experience is investigated through controlled reports and behavior. Objectively recorded task performance and phenomenological appearance are connected but distinct targets of measurement.

Related Articles

Visual Perception: How We Recognize Objects and Scenes explains the broader mechanisms by which visual information supports object and scene recognition.

Optical Illusions: Why the Eyes and Brain Disagree explores how visual context and interpretation produce systematic differences between image properties and perceived appearance.

References

Commission Internationale de l'Éclairage (CIE). (2020). Perceived color (International Lighting Vocabulary, CIE S 017:2020). Official terminology.

Foster, D. H. (2011). Color constancy. Vision Research, 51(7), 674–700. https://doi.org/10.1016/j.visres.2010.09.006.

Gegenfurtner, K. R., & Kiper, D. C. (2003). Color vision. Annual Review of Neuroscience, 26, 181–206. https://doi.org/10.1146/annurev.neuro.26.041002.131116.

Hurlbert, A., & Yu, C. (2025). Seeing the light: Perception and discrimination of illumination color. Annual Review of Vision Science, 11, 267–301. https://doi.org/10.1146/annurev-vision-121423-013755.

Kapelle, O., & Flecken, M. (2025). Perceiving colour through a language lens: A scoping review of experimental work on effects of language on colour perception. Language and Cognition, 17, e45. https://doi.org/10.1017/langcog.2025.13.

Lafer-Sousa, R., Hermann, K. L., & Conway, B. R. (2015). Striking individual differences in color perception uncovered by 'the dress' photograph. Current Biology, 25(13), R545–R546. https://doi.org/10.1016/j.cub.2015.04.053.

Land, E. H. (1977). The retinex theory of color vision. Scientific American, 237(6), 108–128. https://doi.org/10.1038/scientificamerican1277-108.

Lindsey, D. T., & Brown, A. M. (2021). Lexical color categories. Annual Review of Vision Science, 7, 605–631. https://doi.org/10.1146/annurev-vision-093019-112420.

Morimoto, T. (2024). Mechanisms of human color constancy underpinning stable visual perception. JSAP Review, 2024, 240213. https://doi.org/10.11470/jsaprev.240213.

National Eye Institute. (2024). Testing for color vision deficiency. Official guidance.

National Eye Institute. (2025). Color blindness. Official clinical information.

Smithson, H. E. (2005). Sensory, computational and cognitive components of human colour constancy. Philosophical Transactions of the Royal Society B: Biological Sciences, 360(1458), 1329–1346. https://doi.org/10.1098/rstb.2005.1633.

Virtanen, L., Olkkonen, M., & Saarela, T. (2025). The effect of illumination cues on color constancy in simultaneous identification of illumination and reflectance changes. PubMed Central full text.

Witzel, C., & Gegenfurtner, K. R. (2018). Color perception: Objects, constancy, and categories. Annual Review of Vision Science, 4, 475–499. https://doi.org/10.1146/annurev-vision-091517-034231.