[EXPERT: CONSTRUCTED EYE] Day 6 — Checkpoint: Build a Taxonomy of Visual Illusion

Checkpoint: Build a Taxonomy of Visual Illusion

Masterclass Day 6 — The Constructed Eye: Visual Illusion, Perception Science, and the Work of Kitaoka & Lotto

Diagram illustrating two principal types of visual illusions used in taxonomy construction: luminance-based (left) and figure-ground (right). Taxonomy coding is foundational for advanced visual experimentation. Adapted from field standards (Kingdom & Prins, 2016).

Expert Objective

Today's objective is to construct a professional taxonomy of visual illusions, synthesizing historic, contemporary, and experimental categories. This taxonomy is to be operational (i.e., guiding for advanced perceptual studies or studio innovation) rather than simply descriptive, following the orientation of Kitaoka and Lotto's research (Kitaoka, 2005; Lotto, 2003). You will differentiate illusions by observed effects, supported neural/perceptual mechanisms, plausible alternatives, and zones of controversy or ignorance. Such structured insight is essential to design visually groundbreaking, not generically derivative, work.

Taxonomy Construction: Major Families

Taxonomies in vision science are typically based on the phenomenology (what is seen; e.g., shape, color constancy, motion), the mechanisms (how the brain processes), or the architectural features of the visual stimulus (what is varied physically). Kitaoka and Lotto's contributions sharpen certain distinctions, separating, for example, context-based color illusions (e.g., Munker-White) from those dominated by edge/contour processing.

• Luminance, Brightness & Contrast Illusions: E.g., Simultaneous contrast, Craik-O'Brien-Cornsweet. Characterized by local edge- or area-based brightness misperception. (Kingdom & Prins, 2016)
• Color & Chromatic Context Illusions: E.g., Munker-White, the Checker Shadow, Chromatic assimilation. Require spatial context and are robust to changes in viewing distance or adaptation state.
• Geometric & Spatial Warping: E.g., Müller-Lyer, Zöllner. Involve misperceptions of angles, lengths, or position, often explicable by canonical models of orientation/ spatial pooling (Gregory, 1997; Howe & Purves, 2005).
• Figure-Ground & Border Ownership: E.g., Rubin's vase, embedded surfaces. Central to visual organization, with evidence implicating V2 border ownership cells (Zhou et al., 2000).
• Motion Illusions: E.g., Rotating snakes (Kitaoka). Arise from phase-lagged processing of luminance/contrast boundaries; not reliant on true image motion.

Principal axes of visual illusion taxonomy. Each rectangle positions a major category—luminance, color, geometric, and figure-ground—as defined by controlled psychophysical experiments. After Kingdom & Prins (2016).

Evidence and Competing Explanations

• Observed effects: Controlled psychophysics confirms robust category separation under randomized stimulus presentation (Kingdom & Prins, 2016).
• Supported mechanisms: For contrast and color illusions, center-surround antagonism at multiple spatial scales (Shapley & Enroth-Cugell, 1984; Lotto & Purves, 1999). Geometric illusions: orientation-selective pooling and Bayesian inference (Howe & Purves, 2005).
• Competing explanations: Historic theories (e.g., misapplied size-distance scaling in Müller-Lyer) often fail when tested under variable-cue digital conditions. Contextual Bayesian models now predominate (e.g., Lotto, 2003), but many illusions (e.g., the Café Wall) defy complete model capture.
• Unresolved questions: Why do some motion illusions (e.g., Kitaoka’s Rotating Snakes) require fine-tuning of luminance gradients? Are cross-modal effects (auditory influences) present in geometric illusions (research ongoing, see Deroy & Spence, 2016)?

Café Wall Illusion fragments. Segmented contrast boundaries produce orientation misjudgments that defy simple spatial averaging models—highlighting taxonomy value in design and mechanism distinction. Adapted from Gregory & Heard (1979).

Digital Experiment

Goal: Rapidly categorize and query viewer responses to distinct illusion types. Present one of each family: a luminance gradient (Cornsweet), a color context illusion (Munker-White), a geometric distortion (Müller-Lyer), and a border-ownership display (variation of Rubin’s vase).

• Controlled variables: Luminance/contrast, edge definition, spatial arrangement, display color calibration.
• Observation protocol: Sequentially present each stimulus. For each, note the qualitative effect (e.g., “the center bands appear different,” “the lines appear unequal”), estimate effect magnitude (slider, numerical), and confidence. Repeat with inverted colors; is categorization stable?
• Limitations: Web display hardware variability, lack of forced-choice staircases, ambiguity with hybrid examples.

Retrieval Question

Identify two illusions from distinct major families (as defined above), and for each: specify the defining observed effect, the widely supported neural mechanism (with a primary literature reference), and one open question about their mechanism or perception. How could this taxonomy inform your studio experimentation?

Sources

• Kingdom, F.A.A. & Prins, N. (2016), Psychophysics: A Practical Introduction, MIT Press.
• Lotto, R. B., & Purves, D. (2003), The effects of color on brightness, Nature Neuroscience.
• Kitaoka, A. (2005). Trick images: How optical illusions work.
• Howe, C.Q. & Purves, D. (2005), The Müller-Lyer illusion explained by the statistics of image-source relationships. Vision Research.
• Shapley, R., & Enroth-Cugell, C. (1984). Visual adaptation and retinal gain controls. Progress in Retinal Research 3, 263-346.
• Zhou, H., Friedman, H. S., & von der Heydt, R. (2000). Coding of border ownership in monkey visual cortex. J. Neurosci., 20(17), 6594–6611.
• Deroy, O., & Spence, C. (2016). Crossmodal illusions in psychophysics. Multisensory Research, 29(6-7), 615-638.
• Gregory, R. L., & Heard, P. (1979). Border locking and the Café Wall illusion. Perception, 8(4), 365–380.