[EXPERT: CONSTRUCTED EYE] Day 4 — Why Perception Is Not an Objective Recording

Day 4 — Why Perception Is Not an Objective Recording

The images above shows a GREY tomato. Even when we are told, "THERE IS NO RED IN THIS IMAGE!" it is difficult to believe. Yet our brain automatically compensates for the "cast" cyan/teal coloured overlay. Even the selected colour swatches look like a gradient from red to grey when they are the exact same, continuous colour. It is only when compared to the white surrounding that it looks grey. This is simultaneous contrast.

Course: The Constructed Eye: Visual Illusion, Perception Science, and the Work of Akiyoshi Kitaoka and Beau Lotto
Instructor: (redacted for roleplay)

Diagrammatic contrast: The left ellipse represents afferent sensory signals ("input"), while the right ellipse marks the constructed subjective percept—arrows emphasize the transformative process. This gap underpins all visual illusions discussed across this course. Visual metaphor by the author.

Expert Objective

Today’s goal is to establish, through evidence and critical analysis, why human vision is not a passive, objective recording device. Using research from psychophysics and computational neuroscience—including work by Kitaoka and Lotto—we’ll separate robust laboratory findings from speculative explanations. By the close of this lesson, artist-researchers should be able to identify cases where perception demonstrably deviates from physical input, while articulating the operational limits and prevailing theories regarding these deviations.

Observed Effects

• Perceptual Illusions: Such as the Rotating Snakes illusion by Kitaoka, where static images evoke vivid motion (Nature Neuroscience, 2003).
• Contextual Modulation: Luminance and color are subject to surrounding context—the Checker-shadow illusion (Adelson, 1995) is prototypical: Identical grey patches look different under different shadows.
• Ambiguous Figures: Examples like the Necker Cube, where perception fluctuates between two interpretations without a change in stimulus.

The digital equivalent to Adelson’s Checker-Shadow Illusion: Both squares in shadow and light share the same RGB value, yet context makes one appear distinctly lighter. This demonstrates luminance context effects—one of many reasons human vision is not an objective recording. Diagram by the author; see Adelson (1995).

Supported Mechanisms

• Predictive Coding: The brain is not a simple feed-forward camera (Clark, 2013, Behavioral and Brain Sciences), but an active inference engine, making probabilistic predictions to optimize survival-relevant interpretation of ambiguous, noisy, and incomplete data (Summerfield & de Lange, 2014).
• Surround Modulation: Early visual cortex (V1/V2) modifies responses via contextual signals, supported by single-cell recordings in non-human primates (Schallmo et al., 2018, Nature Neuroscience).
• History and Adaptation: Both short-term adaptation (e.g., spatial aftereffects, Troxler fading) and long-term priors influence what is seen (Webster et al., 2002, Nature).

Key modulatory influences on perception: shadow, prior memory, and surround. Real examples—such as Kitaoka’s color-constancy illusions—show dynamic interplay. Visual by the author after Summerfield & de Lange (2014).

Evidence and Competing Explanations

• Psychophysical Record: Repeated experiments confirm that subjective matches for color, brightness, and movement do not consistently align with camera or photometer readings (Foster, 2011, Vision Research).
• Neural Recordings: Firing rates in V1 reflect both sensory and contextual parameters (Schallmo et al., 2018). Not all perceptual errors are due to low-level biophysics; higher-order inference is involved (Lotto & Purves, 2002).
• Dissent and Unresolved: While predictive coding explains many contextual illusions, open questions remain about how and where priors are encoded (Feldman, 2013). Alternative theories, like those emphasizing scale-invariance or efficient coding, account for some, but not all, phenomena.

Studio observation: Artists exploit these effects: e.g., using value grouping or color relativity to control perceived forms. Refer to Josef Albers’ "Homage to the Square" for systematic exploration, as described by MoMA’s conservation department.

Digital Experiment: Contextual Luminance Matching

1. View the two central squares in the prior figure under controlled ambient light.
2. Use the digital color-picker in your painting program to confirm their identical values.
3. Try segmenting the squares from their backgrounds using masking—does the effect persist?

Controlled variables: Screen brightness, ambient light, and display settings.
Observation protocol: Note the persistence of the illusion before and after masking context.
Limitations: This experiment only demonstrates subjective dissociation, not neural mechanism. No self-experiment confirms a specific brain code.

Retrieval Question

Q: Identify two empirically supported reasons why perception cannot be reduced to objective photometric recording. Explain how each is exploited or recognized in contemporary art practice.

Sources

• Kitaoka, A., & Ashida, H. (2003). "A circular motion illusion caused by visual delay." Nature Neuroscience
• Summerfield, C., & de Lange, F. P. (2014). "Expectation in perceptual decision making: neural and computational mechanisms." Neuron
• Adelson, E. H. (1995). "Checker-shadow illusion." Journal of Neuroscience
• Foster, D. H. (2011). "Color constancy." Vision Research
• Lotto, R. B., & Purves, D. (2002). "The empirical basis of color perception." Nature
• MoMA Conservation: "Josef Albers’s Homage to the Square: Experimentation and Process"
• Schallmo, M. P., et al. (2018). "Neural basis of an illusory motion percept." Nature Neuroscience
• Feldman, J. (2013). "What is a visual prior?" Visual Cognition