Perception is often assumed to be a direct, passive readout of external reality, yet human vision frequently constructs interpretations that transcend sensory data. When presented with visual stimuli that support two or more mutually exclusive perceptual states, the brain exhibits an extraordinary property known as perceptual multistability. Studying these intriguing phenomena reveals that perception is an active, inferential dialogue between sensory input and neural modeling rather than a static recording of the visual world.
Ambiguous Figure
1. Concise Definition
An ambiguous figure (also known as a reversible figure or bistable stimulus) is a visual image engineered to elicit two or more distinct, mutually exclusive interpretations despite presenting constant sensory stimulation. Although the physical properties of the image on the retina remain entirely static, the observer's conscious perception spontaneously alternates between incompatible perceptual organizations over time.
In cognitive psychology and visual neuroscience, ambiguous figures serve as standard empirical tools for dissociating objective retinal input from subjective conscious experience. Because the physical stimulus never alters during observation, any perceptual shift reflects internal, top-down neural restructuring, visual inference, or spontaneous neural adaptation rather than external ecological variation. Consequently, these stimuli provide a critical window into the neural correlates of consciousness, cognitive modeling, and subjective perceptual organization.
Furthermore, ambiguous figures challenge classical naive realist conceptions of perception. They demonstrate that visual perception is fundamentally an interpretive act governed by stochastic processes, visual scene statistics, and constructive cognitive constraints rather than a deterministic, bottom-up registration of the physical environment.
2. Etymology & Linguistic Origin
The term ambiguous originates from the Latin adjective ambiguus, meaning "shifting, doubtful, uncertain, or wavering," which derives from the verb ambigere ("to wander about, go about, or debate"), constructed from the prefix ambi- ("both, around, on both sides") and agere ("to drive, lead, or act"). The word emphasizes an inherent dualism or multiplicity of trajectories, capturing the idea of being pulled in two directions at once.
The noun figure traces its roots to the Latin figura, denoting "form, shape, configuration, or appearance," derived from the Proto-Indo-European root *dheigh- ("to form, fashion, or knead"). In psychological discourse, the compound phrase entered widespread academic nomenclature during the late 19th and early 20th centuries, heavily influenced by the emergence of psychophysics and Gestalt psychology in German-speaking Europe, where terms such as Kippfigur (reversing or tilting figure) and Vexierbild (puzzle picture) were systematically translated into English as reversible or ambiguous figures.
3. Pronunciation & Grammatical Form
Pronunciation: /æmˈbɪɡ.ju.əs ˈfɪɡ.jər/ (Received Pronunciation) or /æmˈbɪɡ.ju.əs ˈfɪɡ.jɚ/ (General American).
Grammatical Form: Noun phrase (compound noun). Plural: ambiguous figures. Adjectival usage is typically framed as ambiguously figured, while the phenomenon itself is referred to as figural ambiguity, perceptual bistability, or multistable perception.
4. Detailed Conceptual Explanation
The conceptual framework underlying an ambiguous figure revolves around the computational problem of inverse optics. The human visual system faces the daunting task of inferring a three-dimensional, structured reality from a noisy, flat, two-dimensional array of photons cast upon the retina. Because infinite configurations of the physical world could generate identical retinal images, visual processing relies on probabilistic Bayesian inferences and structural assumptions to arrive at the most likely interpretation. In the case of an ambiguous figure, the computational visual engine encounters a rare, mathematically balanced equilibrium: two distinct environmental interpretations share roughly equal posterior probabilities.
When an observer looks at an ambiguous figure, the visual cortex does not register an uncommitted blur or a simultaneous superposition of both possibilities. Instead, conscious visual awareness collapses into a single, coherent, unified percept. After maintaining this interpretation for a variable interval—typically ranging from two to several seconds—the visual system experiences an involuntary switch, rapidly transitioning to the alternative interpretation. This ongoing alternation continues indefinitely so long as gaze and attention remain fixed upon the image, a dynamic process known as spontaneous perceptual reversal.
The boundaries of an ambiguous figure are defined by its capacity to present incompatible Gestalt structures without providing definitive corroborating evidence that favors one structure over the other. If an artist or experimenter systematically alters the visual cues—such as adding depth lighting, asymmetric line thickness, or chromatic contrast—the ambiguity is broken, and perception stabilizes around the biased percept. The pure ambiguous figure exists precisely at this precarious threshold where bottom-up sensory information is insufficient to resolve perceptual competition, forcing the brain to oscillate between competing hypotheses.
Crucially, bistable ambiguity can manifest across different sensory and conceptual dimensions. In figure-ground ambiguity, the contest centers on which portion of the image is perceived as an occluding foreground object and which as an undifferentiated background. In perspective ambiguity, the competition involves depth orientation, such as whether a geometric structure faces toward or away from the viewer. In semantic ambiguity, high-level categorical identities clash, forcing the brain to alternate between entirely disparate lexical and conceptual representations, such as recognizing an animal versus a human face.
5. Historical Development
The systematic study of ambiguous figures began in the early 19th century as natural philosophers and visual physiologists sought to map the relationship between physical geometry and mental representation. In 1832, the Swiss crystallographer Louis Albert Necker published his observations on the apparent spontaneous perspective reversals of rhomboid crystal line drawings, introducing what is now universally known as the Necker cube. Necker noted that when staring at an isometric wireframe drawing of a cube, the apparent front face would suddenly recede to the back, demonstrating that three-dimensional depth is an internal construction rather than an intrinsic physical property of the drawing.
In 1870, German visual physiologist Ludimar Hermann examined reversible figures, while Joseph Jastrow popularized ambiguous drawings in mainstream psychology during the late 1890s. In 1899, Jastrow adapted an anonymous caricature from a German humor magazine into the famous Duck-Rabbit illusion, using it to demonstrate how expectation, suggestion, and context influence visual perception. Around the same time, in 1896, the Danish psychologist Edgar Rubin conducted rigorous phenomenological investigations into figure-ground segregation, producing the iconic Rubin vase/faces illusion and formulating fundamental principles of visual organization.
During the 1910s and 1920s, Gestalt psychologists—including Max Wertheimer, Wolfgang Köhler, and Kurt Koffka—elevated ambiguous figures to central evidentiary artifacts in their campaign against structuralist atomism. They argued that sensory experience could not be deconstructed into isolated atomic sensations because the visual system spontaneously organizes sensory elements according to holistic rules (Prägnanz, proximity, similarity, and closure). The ambiguous figure demonstrated that the overarching organization of the whole dictates the interpretation of the local parts.
In the mid-20th century, Edwin G. Boring introduced the young girl/old woman figure (originally drawn by cartoonist W. E. Hill in 1915) into academic psychology, prompting deeper explorations into perceptual set and priming. By the late 20th and early 21st centuries, the advent of functional neuroimaging (fMRI), magnetoencephalography (MEG), and single-unit neurophysiology allowed neuroscientists such as Nikos Logothetis, Randolph Blake, and Geraint Rees to track the precise neural dynamics that trigger perceptual switches, firmly establishing ambiguous figures as key paradigms in the modern neuroscience of consciousness.
6. Theoretical Foundations
The mechanisms driving the spontaneous reversal of ambiguous figures have been the subject of an enduring theoretical debate in visual psychology, traditionally framed as a competition between bottom-up satiation models and top-down cognitive models.
The bottom-up neural exhaustion (satiation) hypothesis, originally championed by Wolfgang Köhler and later formalized in neural adaptation models, posits that localized neural populations in early visual areas (such as V1, V2, and V4) code for a specific perceptual interpretation. When an observer views an ambiguous figure, the neural assembly representing Percept A fires vigorously, inhibiting the assembly representing Percept B via reciprocal lateral inhibition. Over time, the active neurons experience metabolic fatigue or synaptic adaptation. As their firing rates decline, their inhibitory influence wanes until the unadapted, rested neural assembly representing Percept B breaks through the inhibition threshold, taking command of conscious awareness. This cycle of adaptation, release, and resurgence explains the continuous, quasi-periodic alternations observed during prolonged viewing.
In contrast, the top-down cognitive hypothesis, advocated by Richard Gregory and Irvin Rock, contends that perceptual reversals reflect executive hypothesis testing and attentional reallocation mediated by higher-order frontal and parietal cortical networks. According to this perspective, perception is an act of unconscious inference; the brain evaluates sensory data against prior models. When faced with an ambiguous figure, the supervisory attentional system detects an unresolved ambiguity and proactively shifts spatial attention or exploratory mental templates, intentionally driving the perceptual switch. Proponents of this model emphasize that voluntary effort, semantic context, and saccadic eye movements can substantially modulate reversal rates.
In contemporary cognitive neuroscience, these competing perspectives are unified within the framework of predictive coding and hierarchical Bayesian inference. Under this model, the brain functions as a prediction machine that generates top-down generative hypotheses to account for bottom-up sensory prediction errors. In an ambiguous figure, two distinct generative models yield comparable error minimization. As the currently dominant model persists, internal adaptation causes unexplained residual errors to accumulate, eventually tipping the Bayesian likelihood balance and prompting the brain to switch to the alternative perceptual hypothesis.
7. Key Components, Types & Dimensions
Ambiguous figures can be categorized into distinct structural classes based on the perceptual transformations they evoke:
- Figure-Ground Ambiguities: Stimuli in which identical contours form the common boundary between two distinct shapes, forcing visual competition over which side of the contour is assigned as the bounded figure and which is relegated to the shapeless background (e.g., the Rubin vase/faces illusion).
- Perspective / Reversible Depth Figures: Geometric, skeletal, or isometric two-dimensional line drawings that can be perceived in two mutually incompatible spatial orientations along the z-axis (e.g., the Necker cube, the Schröder stair).
- Semantic / Meaning Content Ambiguities: Complex representational drawings that share structural features which can be categorized as two entirely different real-world objects, entities, or living beings (e.g., the Duck-Rabbit, the Boring Young Girl/Old Woman figure).
- Orientation / Inversion Ambiguities: Visual compositions designed to be perceived as one coherent image when viewed right-side up, but reveal an entirely distinct, coherent image when physically rotated 180 degrees (e.g., upside-down face-landscape caricatures by Giuseppe Arcimboldo or Rex Whistler).
- Motion Multistability (Dynamic Ambiguities): Kinetic displays where identical physical movements can be resolved into divergent directional trajectories or three-dimensional rotations (e.g., the kinetic depth effect, the spinning dancer illusion, or ambiguous apparent motion lattices).
8. Examples & Illustrative Cases
Several canonical ambiguous figures have defined perceptual psychology and cognitive science across two centuries of research:
The Necker Cube: Published in 1832 by Louis Albert Necker, this isometric drawing of a wireframe cube lacks perspective foreshortening. When observed, the central junction appears alternately to face downward and toward the left, or upward and toward the right. Observers report that the perceived front face spontaneously flips to become the back face without any physical movement of the diagram, demonstrating that depth perception involves internal coordinate assignment.
Rubin's Vase / Two Faces: Developed in 1915 by Danish psychologist Edgar Rubin, this design features a central symmetrical goblet flanked by two opposing silhouetted profiles. The visual system cannot perceive both the vase and the two profiles as simultaneous foreground objects; the contour belongs either to the white central object or the black outer faces, illustrating the fundamental Gestalt law of unilateral contour ownership.
The Duck-Rabbit: Popularized by Joseph Jastrow in 1899, this line drawing depicts an animal head that can be interpreted either as a rabbit facing right (with the long appendages serving as ears) or as a duck facing left (with the appendages serving as a beak). This figure elegantly demonstrates semantic ambiguity and highlights how linguistic priming or environmental context (such as viewing the image at Easter) can bias which interpretation surfaces first.
The Young Girl / Old Woman (My Wife and My Mother-in-Law): Originally published by British cartoonist William Ely Hill in 1915 and studied extensively by Edwin Boring, this complex illustration contains dual semantic features: the young woman's jawline serves as the old woman's nose, her ear serves as the old woman's left eye, and her necklace acts as the old woman's mouth. This classic example shows how local features are systematically reinterpreted depending on the global semantic schema adopted by the observer.
9. Measurement & Assessment
Empirical research on ambiguous figures utilizes several psychophysical, behavioral, and neurophysiological measurement paradigms to track the temporal dynamics of multistability:
Bistable Reversal Rates: The foundational behavioral metric is the alternation frequency, quantified as the number of perceptual switches reported per unit of time (typically reversals per minute). Participants view an ambiguous stimulus under fixed viewing conditions and press specific response keys to signal the onset of each perceptual state or the occurrence of a transition. The distribution of dominance durations typically conforms to a right-skewed gamma or log-normal distribution, characteristic of stochastic renewal processes.
Voluntary Control Paradigms: Researchers assess top-down cognitive flexibility by instructing participants to either actively maintain a given percept (hold condition) or accelerate the switching rate between percepts (switch condition). The difference in reversal frequency between these conditions relative to a passive viewing baseline provides a quantitative index of voluntary attentional modulation.
Objective Physiological Markers: Because manual button presses introduce motor reaction delays and potential reporting biases, modern laboratories rely on non-invasive physiological indicators of perceptual switching. These include tracking microsaccades and spontaneous blink rates, recording optokinetic nystagmus (OKN), measuring pupil diameter dilations (which consistently expand immediately prior to a perceptual switch), and capturing event-related potentials (ERPs)—particularly the reversal-related negativity occurring approximately 200–300 milliseconds before conscious awareness of a perceptual transition.
10. Applications & Practical Significance
Beyond theoretical visual science, ambiguous figures have found meaningful applications across multiple disciplines:
Clinical Neuropsychology and Psychiatry: Reversal dynamics of ambiguous figures serve as diagnostic and exploratory probes for psychiatric and neurodevelopmental conditions. Individuals on the autism spectrum often exhibit reduced reversal rates and longer dominance durations, reflecting atypical predictive processing and altered local-to-global perceptual integration. Conversely, patients with schizophrenia often display erratic switching dynamics and impaired top-down voluntary control over reversals, pointing to disrupted GABAergic inhibitory signaling and destabilized attractor dynamics in prefrontal and parietal cortices.
Cognitive Ergonomics and Interface Design: In aviation, military defense, and user-experience engineering, designers must avoid unintentional figural ambiguity. High-stress user interfaces, heads-up displays, and radar screens must eliminate visual geometries that could accidentally induce bistability or figure-ground confusion, which could lead to fatal misreadings of altitude, orientation, or navigational vectors.
Visual Arts and Graphic Architecture: Fine artists, graphic designers, and architects deliberately deploy ambiguous figures to provoke sustained intellectual engagement. The woodcuts of M. C. Escher, the surrealist paintings of Salvador Dalí, and modern corporate identity logos use deliberate figure-ground and perspective ambiguities to create visually memorable, intellectually stimulating imagery.
11. Research & Empirical Evidence
Centuries of empirical investigation have shed light on the neural substrates that govern multistable perception. Neuroimaging studies utilizing functional magnetic resonance imaging (fMRI) have revealed that while representations of the competing percepts alternate within early visual cortices (such as the lateral geniculate nucleus and visual areas V1 through V5), the neural trigger for the perceptual transition itself originates elsewhere.
Groundbreaking work by Kleinschmidt et al. (1998) and Sterzer et al. (2009) demonstrated that transient bursts of activation in the right inferior parietal lobule and the right dorsolateral prefrontal cortex reliably precede perceptual reversals. This finding provided strong evidence that high-level frontoparietal networks monitor sensory conflict and coordinate restructuring of visual consciousness.
Single-unit electrophysiology in non-human primates, pioneered by Nikos Logothetis and colleagues using binocular rivalry and ambiguous motion displays, showed that while early sensory neurons in primary visual cortex (V1) remain largely tied to physical retinal inputs, neurons in higher-order ventral stream areas—such as the inferotemporal (IT) cortex—fire in direct lockstep with the monkey's reported subjective percept. This demonstrated that perceptual ambiguity is progressively resolved along the hierarchical visual processing stream.
Furthermore, pharmacobehavioral studies have linked reversal rates to chemical neurotransmission. Van Loon et al. (2013) demonstrated through magnetic resonance spectroscopy that concentrations of gamma-aminobutyric acid (GABA)—the primary inhibitory neurotransmitter in the visual cortex—directly predict individual perceptual switch rates. Higher concentrations of GABA in visual visual cortex correlate with longer dominance durations and slower reversal rates, confirming that mutual inhibitory circuits are essential for sustaining stable perceptual representations.
12. Cultural & Cross-Cultural Considerations
Although the basic neurobiological machinery of bistable perception is universal across humans, cultural environment and ecological visual experience exert surprising influences on how ambiguous figures are perceived.
Cross-cultural investigations into semantic ambiguous figures show that cultural familiarity strongly shapes initial perceptual categorization. In classic cross-cultural studies utilizing the Duck-Rabbit drawing, populations residing in arid or rural regions with high ecological exposure to hares but minimal exposure to aquatic birds overwhelmingly report seeing the rabbit first, with some observers failing to detect the duck entirely until prompted. Conversely, urban populations or communities living near waterfowl habitats show balanced or duck-biased initial percepts.
Moreover, differences in perceptual style between Western and East Asian populations influence figure-ground ambiguous displays. Research by Nisbett and colleagues on cultural visual cognition suggests that individuals from East Asian cultural backgrounds exhibit more holistic perceptual processing, attending preferentially to context and background relationships. In contrast, Western participants tend to demonstrate analytic visual processing, focusing focal attention on central foreground objects. Consequently, in ambiguous figure-ground tasks, East Asian observers often detect background silhouettes and alternative contextual configurations more rapidly than their Western counterparts.
13. Criticisms, Debates & Limitations
Despite their ubiquity in cognitive science, ambiguous figures have faced theoretical criticisms and experimental challenges:
A major critique focuses on cognitive penetration and demand characteristics. Critics point out that in typical laboratory settings, participants are explicitly informed that the stimulus is ambiguous and are shown both interpretations beforehand. Under these conditions, it is difficult to determine whether spontaneous reversals reflect purely natural sensory processes or are driven by conscious expectation and demand characteristics. Rock and Mitchener (1992) demonstrated that naive observers who have never been informed of a figure's ambiguity frequently view figures like the Necker cube for extended durations without reporting a single reversal, suggesting that conceptual awareness of ambiguity is often a prerequisite for perceptual bistability.
Another longstanding controversy surrounds the equivalence of ambiguous figures and other multistable phenomena, such as binocular rivalry. While some researchers argue for a universal, centralized "multistability processor" in the human brain, empirical studies reveal distinct differences between the two paradigms. Voluntary attentional control can modulate ambiguous figures far more effectively than binocular rivalry, and the two phenomena show divergent developmental trajectories and unique neurochemical profiles, casting doubt on the existence of a single, unified perceptual switching mechanism.
14. Related Terms & Distinctions
To prevent conceptual confusion, ambiguous figures should be carefully distinguished from related perceptual and artistic phenomena:
- Optical Illusions: Broad phenomena in which an observer's perception systematically departs from physical reality (such as misjudging size in the Müller-Lyer illusion or color in the checker shadow illusion). In standard illusions, perception remains stable and static, whereas ambiguous figures produce ongoing dynamic alternations between two equally valid interpretations.
- Binocular Rivalry: A form of multistable perception that occurs when two entirely different physical images are presented simultaneously to each eye (e.g., vertical stripes to the left eye, horizontal stripes to the right eye). In ambiguous figures, both eyes view the exact same static, physically unified stimulus.
- Pareidolia: The psychological tendency to detect meaningful, recognizable patterns (such as human faces) in random, unstructured visual noise (such as cloud formations or rocky surfaces). Ambiguous figures are deliberately engineered compositions with dual structured interpretations, not random sensory noise.
- Visual Puzzles (Hidden Pictures): Graphic compositions containing a hidden figure embedded within a complex background. Once the hidden figure is discovered, visual organization typically locks onto it permanently, unlike ambiguous figures, which continue to alternate spontaneously between interpretations.
15. Summary / Key Takeaways
Ambiguous figures represent a cornerstone of cognitive science, perceptual psychology, and visual neuroscience. By presenting the visual system with sensory inputs that comfortably support multiple, mutually exclusive interpretations, they prove that conscious vision is fundamentally an active, inferential construction rather than a passive, deterministic recording of the external world.
The temporal dynamics of ambiguous figures are governed by an intricate interplay between bottom-up synaptic adaptation in early visual cortices and top-down attentional modulation orchestrated by frontoparietal networks. As computational models like predictive coding continue to advance, these intriguing stimuli remain vital empirical tools for unraveling the neural foundations of visual inference, sensory organization, and conscious awareness.
Ultimately, ambiguous figures show that when confronted with uncertainty, the human brain refuses to settle for ambiguity; it actively creates coherent interpretations, alternating between complementary truths in its search for meaning.
References
- Blake, R., & Logothetis, N. K. (2002). Visual competition. Nature Reviews Neuroscience, 3(1), 13–21. https://doi.org/10.1038/nrn701
- Gregory, R. L. (1997). Knowledge in perception and illusion. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 352(1358), 1121–1127. https://doi.org/10.1098/rstb.1997.0095
- Kleinschmidt, A., Büchel, C., Zeki, S., & Frackowiak, R. S. (1998). Human brain activity during spontaneously reversing perception of ambiguous figures. Proceedings of the Royal Society of London. Series B: Biological Sciences, 265(1413), 2427–2433. https://doi.org/10.1098/rspb.1998.0594
- Rock, I., & Mitchener, K. (1992). Further evidence of failure of reversal of ambiguous figures without knowledge. Perception, 21(1), 39–45. https://doi.org/10.1068/p210039
- Sterzer, P., Kleinschmidt, A., & Rees, G. (2009). The neural bases of multistable perception. Trends in Cognitive Sciences, 13(7), 310–318. https://doi.org/10.1016/j.tics.2009.04.006
- van Loon, A. M., Knapen, T., Scholte, H. S., John-Saaltink, E. S., Donner, T. H., & Lamme, V. A. (2013). GABA shapes the dynamics of bistable perception. Current Biology, 23(9), 823–827. https://doi.org/10.1016/j.cub.2013.03.067