Comprehensive Notes on Infant Color Perception and Perceptual Development

Introduction to Infant Color Perception

• Objective: Decades of research have detailed the visual and perceptual development of infants, providing insights into the nature of perception and human development (Johnson, 2011; Maurer & Werker, 2014). This specific review focuses on color perception, primarily within the first $6\,months$ after birth.
• Color as a Research Tool: Color serves as a rich medium for understanding the transition from basic sensation to functional perceptual representations that allow infants to interact with the world.
• Mature Color Vision: The neurobiology and sensory mechanisms of adult color vision are well-documented (Conway et al., 2010). Color is pervasive in experience, acting as a cue for communicating and perceiving objects, scenes, and faces, and playing a role in aesthetics.
• Controlled Stimulus: Color is advantageous for research because it can be precisely quantified and controlled as an experimental stimulus.
• Generalizability Limitation: Current research is almost exclusively focused on infants from "WEIRD" populations (Western, Educated, Industrialized, Rich, Democratic) such as the United States, United Kingdom, Europe, and Japan. Sociodemographic details are rarely reported, which limits the global generalizability of the findings.
• Funding and Authorship: This research is part of the PROJECT COLOURMIND, funded by the European Research Council under the European Union’s Horizon $2020$ program (Grant No. $772193$, awarded to Anna Franklin). Alice Skelton and John Maule are joint first authors from the Sussex Colour Group & Baby Lab, University of Sussex.

Developmental Timeline and Mechanisms of Color Vision

• Neonatal Perception: Contrary to the myth that infants see only in black and white, neonates can detect some color. However, their vision is poor; stimuli must be highly saturated, large, and of specific hues (e.g., red).
• Detection Statistics: A study by Adams et al. ($1994$) found that more than $75\%$ of neonates oriented to large patches of saturated red on a grey background, بينما more than $80\%$ failed to orient to blue patches under the same conditions.
• Physiological Basis for Poor Vision:     * Retinal Immaturity: Cone photoreceptors are not yet as elongated or densely organized as those in a mature retina (Yuodelis & Hendrickson, $1986$).     * Cortical Immaturity: The visual cortex is also undeveloped at birth.
• Photoreceptor Mechanisms: Typical human color vision is trichromatic, based on three cone types:     * L-cones: Long wavelengths (reddish light).     * M-cones: Medium wavelengths (greenish light).     * S-cones: Short wavelengths (bluish light).
• Timeline of Mechanism Development:     * Signals are combined into two retino-geniculate mechanisms: red–green and blue–yellow cone-opponent channels.     * Red-Green Mechanism: Develops first.     * Blue-Yellow Mechanism: Develops approximately $4$ to $8\,weeks$ later.     * Trichromacy: Infants generally become trichromatic (both channels active) by $3\,months$ (Teller, $1998$).
• Maturation Durations: Even after becoming trichromatic, saturation thresholds do not reach adult levels until late adolescence (Knoblauch et al., $2001$). This aligns with other visual milestones: visual acuity matures at age $7$ and global motion at age $12$ (Maurer, $2017$).

Perceptual Dimensions: Hue, Lightness, and Saturation

• Transition to Higher-Level Processing: Around $3\,months$, infants begin using color in ways reflecting complex perceptual processes, representing color via three dimensions: hue (wavelength-based, e.g., reddish), lightness (internal light level), and saturation (intensity).
• Hue Preference Curves: From $3\,months$ onward, infants exhibit specific looking preferences:     * Longest Look: Bluish hues.     * Secondary Look: Reddish and purplish hues.     * Least Look: Yellow-greenish hues (Bornstein, $1975$; Skelton & Franklin, $2020$; Zemach et al., $2007$).
• Independence of Hue: Modeling suggests these preferences are based on hue rather than detection thresholds, saturation, or luminance (Brown & Lindsey, $2013$). This suggests a shift in the development and organization of the visual cortex, likely arising in the extrastriate visual cortex.
• Quantifying Perceptual Dimensions:     * Maximum Likelihood Conjoint Measurement (MLCM): A psychophysical method using signal detection theory to model how behavioral choices are determined by covariation across multiple attributes (Ho et al., $2008$). In $6$-month-olds, looking behavior to stimuli varying in chroma (saturation) and lightness is predicted by the sum of dimensions rather than their interaction (Rogers et al., $2018$).     * Interdimensional Salience Mapping (ISM): Uses forced-choice preferential looking to plot salience of different dimensions (e.g., size vs. hue). This allows researchers to compare attributes when perceptual salience is equated (Kaldy et al., $2006$).

The Nature of Color Categorization in Infancy

• Categorical Response: By $4\,months$, infants divide the color spectrum into five discrete categories: red, green, blue, yellow, and purple (Maule & Franklin, $2019$).
• Classical Definition of Categorization: In studies using the novelty preference method, infants were familiarized to one hue. They treated discriminably different hues within the same category as equivalent, while showing novelty preference for hues in different categories (Skelton et al., $2017$).
• Theoretical Implications:     * Debate: These findings challenge the "cultural construction" view that color categories are entirely products of language and culture.     * Consistency: Infant color categorization aligns with broader evidence that infants categorize faces, animals, speech sounds, and objects (Westermann & Mareschal, $2012$).     * Sensory Mechanisms: Interestingly, four of the five categorical distinctions in infants are separated by the cardinal axes of cone-opponent color vision, suggesting a biological/sensory foundation.
• Lexical Development: Mapping infant categories shows a striking similarity to the world's diverse color lexicons. Global color terms tend to cluster around the same focal points in color space as infant categories, suggesting sensory mechanisms constrain the development of language-specific color terms.

Color Constancy and Object Perception

• Color Constancy Definition: The ability of the visual system to compensate for variable illumination (e.g., changes in spectra from daylight to indoor light) to keep object color perception stable.
• Infant Competency: By $3\,months$, habituation/novelty studies show infants do not perceive a stimulus as novel when illumination changes, indicating rudimentary color constancy (Dannemiller, $1989$; Yang et al., $2013$).
• Maturation: While present in infancy, color constancy continues to mature between ages $2$ and $4$ (Rogers et al., $2020$). Other forms of constancy (size, shape, lightness) are present by $6\,months$.
• Typicality and Object Color:     * Yellow Banana Study: $6$-month-olds look longer at typical colors (yellow banana) than atypical colors (blue banana) (Kimura et al., $2010$).     * Face Color Study: In contrast, infants did not show longer looking at typically colored faces compared to blue, purple, or green faces (Clifford et al., $2014$).
• Cue Hierarchy: In object individuation, color appears to be a less attended cue than shape or size in early infancy (Wilcox, $1999$).

Chromatic Scene Statistics and Environmental Calibration

• Natural Illumination Alignment: Adult color perception is calibrated to the statistical regularities of natural scenes (chromatic scene statistics). Sensitivity aligns with a blue-yellow axis in color space, where natural daylight falls.
• Early Calibration: At just $4\,months$, infants' saturation sensitivity is already aligned with this blue-yellow axis (Bosten et al., $2016$).
• Efficient Coding Framework: This suggests the visual system is optimized to encode statistical regularities of the environment, either through evolution or rapid tuning within the first few months post-birth.
• Comparative Statistics: Sensitivity to spatial scene statistics does not become optimal until age $10$ (Ellemberg et al., $2012$), making the early maturation of color sensitivity notable.

The Role of Environmental Experience and Sensitive Periods

• Premature Infants: Infants born prematurely show that more visual experience enhances chromatic sensitivity more than luminance sensitivity (Bosworth & Dobkins, $2009$).
• Sensitive Periods in Macaques: Macaques deprived of color (monochromatic input) for a year failed to develop color constancy or typical judgments (Sugita, $2004$).
• Human Cataracts: Adults who had congenital cataracts removed in early development often show typical color discrimination later, suggesting either that specific types of deprivation matter or that sensitive periods only apply to certain aspects of color.
• Arctic Circle / Latitude of Birth Study:     * Norwegian adults born above the Arctic Circle (exposed to the "mørketid" or purplish twilight) showed differences in color discrimination compared to those born below the circle.     * Specific Effects: Those from the Arctic could discriminate purple hues more effectively but green hues less effectively.     * Seasonality: Those born in autumn above the Arctic Circle had poorer overall color sensitivity than those born in summer (Laeng et al., $2007$).

Future Research Directions

• Cortical Representation: There is a lack of data on how the cortex represents color in infancy. Emerging breakthroughs in infant functional magnetic resonance imaging (fMRI) have already identified category-specific regions in the ventral visual pathway for faces and scenes in infants under $1\,year$ old (Kosakowski et al., $2021$).
• Top-Down Processing: Researchers aim to explore if object knowledge influences infant color perception, similar to how adult expectations affect color appearance (Hansen et al., $2006$).
• Bayesian Models: Applying Bayesian models (combining prior experience with sensory input) to infants could clarify the role of predictive processing in perception.
• Population Diversity: There is a critical need to diversify study populations beyond WEIRD samples to understand how different environmental factors and chromatic scene statistics globally influence perceptual development.