The Competitive Paradigm for Chimpanzee Theory of Mind – Brian Hare, Josep Call, and Michael Tomasello

For more than four decades, comparative psychology and cognitive ethology have grappled with one of the most conceptually profound questions in the study of animal minds: do non-human primates possess a theory of mind? Initially coined in 1978, the phrase denotes the capacity to attribute unobservable mental states—such as desires, intentions, perceptions, and beliefs—to oneself and others, using those attributions to predict and interpret behavior. What began as an experimental investigation into whether a chimpanzee could deduce the intentions of a human actor soon degenerated into an intractable empirical deadlock. Throughout the late 1980s and 1990s, an increasingly influential line of research contended that non-human primates were fundamentally behavior-readers rather than mind-readers: organisms remarkably adept at processing external cues, surface bodily postures, and statistical regularities, yet profoundly blind to the internal, mentalistic lives of their social companions.

This negative consensus was not merely an empirical assertion; it was the product of a specific, human-centric methodological paradigm. Experimental designs during this era routinely placed great apes in cooperative-communicative encounters with human experimenters. Captive chimpanzees were tasked with interpreting subtle referential gestures, such as pointing or declarative gaze cues, to locate hidden food items. When apes repeatedly failed these cooperative tests, leading researchers concluded that the evolutionary origins of mentalizing were uniquely human, emerging only after the divergence of the hominin lineage from the last common ancestor shared with chimpanzees. This view, however, neglected a fundamental tenet of evolutionary biology: cognitive adaptations do not evolve in an ecological vacuum. They are forged by the selective pressures of a species’ natural socioecology.

At the turn of the twenty-first century, a revolutionary methodological and theoretical realignment swept through comparative cognition. Spearheaded by Brian Hare, Josep Call, and Michael Tomasello at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, this shift challenged the ecological validity of the prevailing cooperative paradigms. By reconceptualizing chimpanzee social intelligence through the lens of conspecific resource competition, the researchers constructed a novel experimental framework: the competitive paradigm. By pitting subordinate chimpanzees against dominant conspecifics in ecologically authentic contests over food, Hare, Call, and Tomasello uncovered an astonishing array of socio-cognitive abilities that cooperative tasks had systematically obscured. This article provides an exhaustive, critical analysis of the competitive paradigm, tracing its historical origins, experimental architectures, theoretical controversies, philosophical implications, and lasting legacy in cognitive science.

1. Historical Antecedents: The Search for Primate Theory of Mind Before the Paradigm Shift

1.1 Premack and Woodruff’s Foundational Inquiries

The empirical quest to determine whether non-human animals mentalize formally originated with David Premack and Guy Woodruff’s seminal 1978 paper published in Behavioral and Brain Sciences, titled “Does the chimpanzee have a theory of mind?” Premack and Woodruff sought to investigate whether an adult, language-trained chimpanzee named Sarah could comprehend the internal problem-solving states of a human actor. Sarah was shown a series of videotaped vignettes depicting a human experimenter struggling to resolve various physical dilemmas. These situations included an actor attempting to escape from an inaccessible, locked cage, shivering uncontrollably next to an unplugged electric heater, or trying to sweep dirty paper out of reach behind an imposing physical barrier.

Following each videotaped presentation, Sarah was presented with paired photographic alternatives, only one of which represented a viable solution to the problem depicted in the video. For instance, the solution to the locked cage was a key; the solution to the unlit heater was a power cord plugged into an electrical socket; the solution to the inaccessible debris was an elongated pole or broom. Across extensive testing sessions, Sarah consistently and accurately selected the photograph showing the correct instrumental resolution to the actor’s predicament. Premack and Woodruff interpreted these choices as evidence that Sarah attributed internal, epistemic states to the human actor—specifically, an understanding of the human’s underlying desires, goals, intentions, and motivational states.

Despite the revolutionary impact of Premack and Woodruff’s study, philosophers and cognitive scientists quickly highlighted severe epistemological limitations and interpretive ambiguities within the experimental design. Commentaries by Daniel Dennett, Jonathan Bennett, and Gilbert Harman pointed out that Sarah’s success did not necessitate the attribution of mental states. Instead of mentalizing, Sarah could have operated via lower-level, associative heuristics: pairing elements based on learned functional associations or perceptual affinity (e.g., associating the image of a key with a physical lock, or an electric plug with a heater). Sarah possessed extensive training with human artifacts, tools, and visual materials, making it impossible to disentangle true mental state attribution from high-level instrumental problem-solving and environmental pattern recognition. Consequently, the field required designs capable of disentangling perceptual-associative heuristics from genuine social cognition.

1.2 The Rise of Cooperative-Communicative Paradigms in Primate Testing

In response to the conceptual challenges leveled against Premack and Woodruff, researchers shifted toward communicative and cooperative testing scenarios during the late 1980s and early 1990s. This methodological program sought to test whether primates understood communicative intent by designing tasks centered on human-experimenter cueing. The preeminent experimental framework of this era was the object-choice paradigm. In a typical iteration, an ape sat across from a human experimenter, with two opaque containers placed between them, one of which concealed a food reward. The human experimenter, who possessed knowledge of the food’s location, provided the ape with various communicative cues, including declarative pointing gestures, direct visual orienting (gazing), or exaggerated head-tilts directed toward the baited container.

These paradigms rested on a profound, unexamined assumption: that captive great apes, having been raised in proximity to humans, would inherently possess the socio-communicative competence required to interpret cooperative referential gestures. Researchers assumed that pointing and gaze-following in an object-choice task were ecologically transparent, universal mechanisms of social transmission. If a chimpanzee truly understood what an experimenter saw or intended, it should readily comprehend that the experimenter was offering valuable, cooperative information designed to facilitate food retrieval.

The empirical reality was a persistent, confounding failure. Across hundreds of trials conducted by multiple independent laboratories, captive chimpanzees, bonobos, and other great apes performed near chance levels in standard cooperative object-choice tasks. When an experimenter pointed directly at a cup concealing food, apes frequently failed to exploit the cue, often choosing randomly or requiring hundreds of reinforced conditioning trials to associate the static physical finger with the reward location. In food-begging tasks, where chimpanzees were given the opportunity to solicit food from two human experimenters—one attentive and one inattentive—the animals routinely begged indiscriminately. The systemic failure of great apes to decode cooperative-communicative gestures led researchers to conclude that these animals possessed a profound cognitive deficit regarding communicative intent and visual attention.

1.3 The Emergence of the Negative Consensus in Primate Social Cognition

The theoretical synthesis of these negative findings culminated in the work of Daniel Povinelli and his colleagues at the University of Louisiana at New Iberia. Throughout the 1990s, Povinelli conducted an exhaustive series of controlled experiments designed to test whether chimpanzees understood seeing as an attention-mediated mental state. In their classic 1996 monograph, What Chimpanzees Know About Seeing, Povinelli and Eddy tested young captive chimpanzees who had the opportunity to beg for food from two human experimenters. The experimenters adopted varying postures to disrupt visual perception: one wore a blindfold over their eyes while the other wore a blindfold over their mouth; one wore a bucket over their head while the other held a bucket on their shoulder; one sat with their back turned to the ape while the other faced forward; or one held their hands over their eyes while the other held their hands over their ears.

If chimpanzees understood seeing as an epistemic conduit connecting an external object to an internal mental representation, they should selectively solicit food from the individual whose visual field was unobstructed. Instead, Povinelli’s chimpanzees directed their begging gestures almost equally toward blindfolded individuals, individuals with buckets over their heads, and individuals who could visually perceive them. While the apes eventually learned to avoid individuals with their backs turned, fine-grained temporal analysis revealed that this distinction was driven by a gross behavioral heuristic—approaching an actor whose front body surface or face was oriented toward them—rather than an appreciation of the experimenter’s open gaze or visual attention.

From these data, Povinelli formulated the behavior-reading hypothesis (or the low-level reinterpretation model). He posited that the cognitive architecture of non-human primates is fundamentally devoid of mentalistic concepts. Chimpanzees, Povinelli argued, are biological automatons of extreme behavioral sophistication: they represent complex spatio-temporal contingencies, body orientations, postural cues, and dynamic behavioral histories, yet never posit intermediate psychological constructs like “seeing,” “knowing,” or “intending.” By the close of the twentieth century, an entrenched academic consensus had formed: Theory of Mind was an evolutionary novelty unique to the genus Homo, leaving chimpanzees stranded on the non-mentalizing side of an insurmountable phylogenetic chasm.

2. Ecological and Evolutionary Foundations of the Competitive Paradigm

2.1 Socioecology of Pan troglodytes and Selection Pressures

The foundational flaw of the late-twentieth-century negative consensus lay in its complete detachment from the evolutionary ecology of the study organism. The cognitive mechanisms of any species are tailored by natural selection to resolve specific challenges encountered within its ancestral environment. For Pan troglodytes, that environment is characterized by a fission-fusion social system marked by perpetual within-group competition for calorie-dense, spatially unpredictable resources, such as ripe fruit and vertebrate meat.

In wild chimpanzee communities, daily social life is structured by strict, highly linear dominance hierarchies. High-ranking, despotic individuals routinely monopolize preferred resources through physical intimidation, agonistic displays, and overt violence. When a high-value resource appears, dominant chimpanzees exercise immediate priority of access. For a subordinate individual, approaching a contested resource in plain view of a dominant carries substantial fitness costs, ranging from severe physical wounding to socio-political demotion within the group. To survive and secure nutrition under such despotic conditions, subordinates cannot rely on cooperative communication. Rather, they face intense selection pressures to engage in tactical circumvention: foraging covertly, monitoring the visual access of dominant competitors, and exploiting occlusions in the forest canopy.

This reality directly aligns with the Machiavellian Intelligence Hypothesis, formulated by Richard Byrne, Andrew Whiten, and Nicholas Humphrey. The hypothesis posits that higher primate intelligence evolved not to solve non-social ecological problems (such as extractive foraging or mental mapping), but to master the shifting dynamics of intra-group competition, deception, manipulation, and political maneuvering. To place a chimpanzee in an artificial laboratory environment and demand that it process cooperative pointing from a human experimenter who willingly hands over food violates every ecological norm of Pan troglodytes. In the wild, chimpanzees almost never point cooperatively to declare the presence of food to a group-mate. Food sharing is heavily contested, begrudging, or mediated by complex reciprocal coalitions. The cooperative-communicative paradigm had asked the right cognitive question through the wrong evolutionary interface.

2.2 The Tomasello and Call Theoretical Realignment

Recognizing the deep ecological invalidity of Povinelli’s paradigms, Michael Tomasello and Josep Call, working at the newly established Max Planck Institute for Evolutionary Anthropology in Leipzig, executed an ambitious theoretical realignment of primate cognitive testing. In their 1997 treatise, Primate Cognition, Tomasello and Call argued that comparative psychology had systematically conflated an organism’s general social-cognitive capacity with its species-specific communicative habits. Human infants develop within an ontogenetic framework dominated by cooperative caregiving, shared intentionality, and joint attention; consequently, human infants thrive in cooperative-communicative scenarios.

Chimpanzees, conversely, operate via a social dynamic where information is guarded rather than freely shared. If a chimpanzee possesses the capacity to understand what another individual perceives or intends, that capacity would not manifest as a benign willingness to read cooperative pointing gestures. Rather, it would emerge as a weapon of tactical exploitation during aggressive resource competition. In a cooperative context, a subordinate has no evolutionary expectation that a dominant or human partner will assist it in acquiring food. In a competitive context, however, calculating what a rival can and cannot see is a matter of direct survival.

Along with their doctoral student Brian Hare, Tomasello and Call developed the structural blueprint for intraspecific competitive testing protocols. Instead of using human experimenters wielding artificial communicative gestures, the competitive paradigm paired two conspecific chimpanzees—a dominant and a subordinate—in direct, physical, zero-sum competition over food rewards. By manipulating the geometric relationship between the food, the physical environment, and the visual access of both competitors, the researchers constructed an experimental crucible capable of determining whether chimpanzees could mentalize when it mattered most: under the intense, ecologically authentic pressure of social competition.

3. The Seminal 2000 Study: Chimpanzees Know What Conspecifics Do and Do Not See

3.1 Experimental Architecture of Hare, Call, Agnetta, and Tomasello (2000)

In 2000, Brian Hare, Josep Call, Bryan Agnetta, and Michael Tomasello published their watershed study in Animal Behaviour: “Chimpanzees know what conspecifics do and do not see.” The experimental apparatus was designed to model the exact physical dynamics of natural foraging under competitive social constraint. The testing environment consisted of two opposing indoor holding cages separated by a larger, rectangular testing arena. One holding cage contained an established, socially dominant chimpanzee, while the opposing cage housed a subordinate individual. Both cages were fitted with vertically sliding, transparent mesh doors and overhead overhead-controlled opaque doors, allowing experimenters to regulate the visual and physical access of each ape independently.

The experimental procedure began by baiting the central arena with high-value food items, typically pieces of banana or apple, placed between the two sliding doors. The strategic crux of the design was the systematic introduction of visual barriers: physical occluders constructed from opaque wooden boards or inverted PVC cylinders. Depending on their spatial orientation, these barriers could obstruct the direct line of sight between the dominant chimpanzee’s holding cage and a specific piece of food, while leaving the food completely visible to the subordinate on the opposite side of the arena.

Methodological precision was maintained regarding the temporal mechanics of door operation. The overhead doors were initially raised to allow both the dominant and subordinate individuals to inspect the layout of the arena through clear mesh. Once baiting was complete, the subordinate’s mesh door was released slightly ahead of the dominant’s door (or simultaneously, under specific delay conditions). This slight temporal advantage was critical: it forced the subordinate to make an immediate, predictive behavioral decision before the dominant entered the arena, eliminating the possibility that the subordinate was simply reacting to the dominant’s overt pursuit behavior or spatial trajectory.

3.2 Key Conditions: Opaque Versus Transparent Obstacles

The 2000 study implemented a series of conditions designed to systematically parse the cognitive calculations governing the subordinate’s foraging choices. In the primary test condition—the Opaque Barrier Condition—two pieces of food were placed in the central arena. One food item was positioned out in the open, fully visible to both the dominant and the subordinate. The second food item was positioned behind an opaque, upright wooden barrier, deliberately arranged so that it was visible to the subordinate from its vantage point, but visually occluded from the dominant’s vantage point across the arena.

If the subordinate was incapable of visual perspective taking and relied solely on basic self-centered heuristics (e.g., “I see food, so I will take the closest piece”), it should target both food items at equal rates, or preferentially target the open food if it sat along a more direct approach trajectory. If, however, the subordinate understood that the barrier blocked the dominant’s line of sight—and consequently recognized that the dominant would immediately claim the visible food—the subordinate should selectively target the occluded piece, securing the uncontested reward without provoking an attack.

The results provided empirical confirmation of perspective tracking: subordinate chimpanzees overwhelmingly approached and retrieved the food item hidden behind the opaque barrier, while consistently avoiding the food out in the open. They demonstrated a sophisticated tactical restraint, recognizing that visible food was de facto dominant food. In control trials where both food items were out in the open, subordinates almost never ventured into the arena, recognizing that the dominant’s presence eliminated their chances of retrieval. When both items were placed behind barriers of unequal heights or orientations, subordinates systematically targeted the specific piece that was geometrically concealed from the dominant’s angle of vision.

3.3 Excluding Low-Level Explanations: The Transparent Barrier Control

While the initial results from the opaque barrier conditions were compelling, they were instantly vulnerable to a serious low-level behavioral counter-argument. Did the subordinates target the hidden food because they understood the dominant could not see it, or were they simply adhering to a basic, non-mentalistic environmental association: “Always retrieve food that sits adjacent to a physical wall or barrier”? Physical barriers might provide an unconditioned sense of security, serving as an artificial “safety haven” or a mechanical edge that reduced perceived vulnerability, independent of any social perspective taking.

To eliminate this non-mentalistic explanation, Hare and colleagues introduced the critical Transparent Barrier Control. In this condition, the physical architecture of the arena remained identical, with one key modification: the opaque wooden barrier was replaced with an identical barrier constructed of clear, transparent Plexiglas. Under this configuration, the physical, spatial, and mechanical properties of the environment were preserved. The food sat directly adjacent to an upright, rigid surface, exactly as it had in the opaque condition. Crucially, however, the Plexiglas allowed light to pass through: the dominant possessed an unobstructed visual line of sight straight through the clear barrier to the food reward.

The behavioral results were definitive. When faced with the transparent Plexiglas barrier, subordinate chimpanzees ceased their preference for the obstructed food. They treated food behind the transparent barrier exactly as they treated food out in the open: as a dangerously exposed, high-risk target that the dominant could visually detect and claim. Subordinates only demonstrated a robust, statistically significant preference for barrier-adjacent food when the barrier was physically opaque. This control proved that spatial proximity to a physical object was not driving the apes’ choices. Instead, their decisions were dictated by the geometric interruption of their competitor’s line of sight—providing strong empirical evidence that chimpanzees understand the perceptual boundaries of others.

4. Temporal Dynamics and Epistemic States: Visual History and Knowledge Attribution (2001)

4.1 Experimental Design of Hare, Call, and Tomasello (2001)

Demonstrating that an animal understands visual line of sight in the perceptual present represents a critical step forward, yet it addresses only the baseline tier of mentalizing: perceptual perspective taking (often termed Level 1 visual perspective taking). A more advanced tier of Theory of Mind involves tracking informational access across time: understanding that past visual perception produces current knowledge, while the absence of past perception results in ignorance. In 2001, Hare, Call, and Tomasello published their next landmark study in Nature: “Do chimpanzees know what conspecifics know and do not know?”

This investigation moved beyond instantaneous line-of-sight geometry to examine whether chimpanzees could maintain an epistemic record of another individual’s visual history. The physical layout mirrored the 2000 setup, with two opposing chimpanzees competing for food in a central arena featuring opaque occluders. The crucial experimental variable was the manipulation of the dominant’s visual access during the prior baiting process, well before the doors were opened for physical competition.

In these designs, the subordinate individual was consistently granted full, unobstructed visual access to the entire baiting sequence. The subordinate could observe the food being carried into the arena, placed behind one of two opaque occluders, and, critically, could see whether the dominant was watching the baiting event or was prevented from doing so. By decoupling what the dominant could see in the moment of competition from what the dominant had seen in the preceding minutes, the experimenters isolated memory-based mental state attribution from instantaneous perceptual cues.

4.2 The ‘Seen’ Versus ‘Unseen’ Experimental Conditions

Hare, Call, and Tomasello established three foundational conditions to assess the tracking of epistemic states. In the first condition, the Dominant-Seen Condition, the dominant individual’s transparent door was open during the baiting process, allowing the dominant to watch the human experimenter place a piece of food behind one of two opaque occluders. In the second condition, the Dominant-Unseen Condition, the dominant’s opaque door was closed during the baiting process; the subordinate watched the food being placed behind an occluder, fully aware that the dominant was cut off from this visual event.

Once the food was hidden, the dominant’s opaque door was opened, presenting both contestants with an identical perceptual scene: two opaque occluders concealing food, with neither individual capable of seeing the food at that specific moment. If subordinates were restricted to evaluating instantaneous perceptual states, they should show no behavioral difference between conditions, since the dominant could not see the food in either scenario. If, however, the subordinates tracked what the dominant knew based on its visual history, they should anticipate that the dominant in the “Seen” condition would head directly to the baited occluder, whereas the dominant in the “Unseen” condition would be ignorant of the food’s location.

Subordinates demonstrated an acute sensitivity to this visual history. In the Dominant-Unseen condition, subordinates entered the arena and retrieved the hidden food with significantly higher frequency and shorter latencies than in the Dominant-Seen condition. They accurately deduced that a dominant who had not witnessed the baiting event posed minimal threat, because the dominant did not know where the food was hidden.

To provide an even more rigorous test of conspecific knowledge attribution, the researchers introduced the Dominant-Switch Control. In this condition, a specific dominant individual (Dominant A) watched the food being baited behind an occluder. However, prior to releasing the contestants, Dominant A was led out of the holding area and replaced with a naive individual (Dominant B), who had witnessed nothing. The subordinate observed this switch. When the door opened, subordinates treated Dominant B as completely ignorant, aggressively targeting the food that Dominant A had seen hidden. The subordinates were not merely tracking whether “a dominant” was present; they tracked the specific epistemic profile of individual conspecifics over time.

4.3 The Shifted Food Paradigm: Updating Epistemic States

To further test the boundaries of chimpanzee epistemic tracking, Hare and colleagues devised the Shifted Food Paradigm. This design addressed whether chimpanzees could track the dynamic updating of visual information, testing the threshold between understanding ignorance (non-knowledge) and attributing explicit false belief. In the primary shifted condition, both the subordinate and the dominant initially witnessed a piece of food being hidden behind Occluder 1. Following this initial placement, the dominant’s visual access was blocked by closing its opaque door. With the dominant unable to see, the human experimenter entered the arena, retrieved the food from behind Occluder 1, and relocated it behind Occluder 2, all within clear view of the subordinate.

When the dominant’s door was reopened, the two apes faced each other. Perceptually, both occluders were closed. Epistemically, however, the dominant possessed outdated visual information: it had seen the food placed at Occluder 1, but was ignorant of its relocation to Occluder 2. The subordinate, meanwhile, held a complete and updated informational account of the event. If the subordinate possessed the ability to anticipate the dominant’s behavior based on outdated information, it should predict that the dominant would search at the original location (Occluder 1), thereby leaving the relocated food at Occluder 2 open for safe retrieval by the subordinate.

The empirical findings demonstrated that subordinate chimpanzees selectively targeted the relocated food item at Occluder 2, taking strategic advantage of the dominant’s outdated informational state. However, whether this behavior constitutes evidence of false belief understanding—the representation of another agent’s internally held, counterfactual representation of reality—remained a subject of intense debate. While the data decisively proved that chimpanzees update their understanding of what others have and have not registered, Tomasello and Call remained cautious. They argued that subordinates might not be computing a subjective “false belief,” but rather operating on a model of “ignorance of recent updates,” predicting that a dominant will target only those spatial coordinates it directly registered in the past.

5. Auditory Perspective Taking and Concealment Strategies

5.1 Sensory Modality Expansion: Audition in Competition

If Theory of Mind represents a generalized, domain-general understanding of another individual’s perceptual states, it should not be constrained to the visual modality. Auditory perception represents an equally vital channel through which conspecifics acquire information about their environment. In dense tropical rainforests, where visual lines of sight are routinely fractured by thick vegetation, acoustic information often serves as the primary sense alerting dominants to the presence of food or the actions of competitors.

In 2006, Alicia Melis, Josep Call, and Michael Tomasello extended the competitive paradigm into the auditory domain in a study published in the Proceedings of the Royal Society B: “Chimpanzees (Pan troglodytes) conceal visual and auditory information from others.” The experimental apparatus tested whether chimpanzees, when attempting to steal food from a human competitor, were sensitive to the acoustic consequences of their own actions. Chimpanzees faced two distinct tunnels leading toward a contested food reward. One tunnel was rigged with silent, soft rubber flaps, while the other tunnel was adorned with metallic hanging bells and rigid plastic baffles that rattled loudly when moved.

The experimenters systematically manipulated the visual attention of the human competitor, who sat between the two tunnels guarding the food. When the human competitor was facing forward, looking directly at the ape, the chimpanzees demonstrated no tunnel preference, recognizing that visual detection rendered auditory stealth irrelevant. However, when the human competitor had their back turned—and was therefore relying exclusively on acoustic cues to detect intruders—the chimpanzees selectively approached the food through the silent tunnel, actively suppressing the production of acoustic cues that could betray their presence. This finding proved that chimpanzees do not merely calculate geometric lines of sight; they monitor acoustic environments, anticipating how mechanical noise generates awareness in a potential competitor.

5.2 Tactical Deception and Stealth in Naturalistic Settings

The laboratory findings of Melis, Call, and Tomasello corroborated an extensive corpus of naturalistic observations collected by primatologists studying wild chimpanzee communities. Decades of field research in Gombe, Taï Forest, and Kibale had documented instances of tactical deception that hinged on the active suppression of sensory cues. For example, during inter-group border patrols—where chimpanzee parties invade the territory of neighboring communities—males display sustained, collaborative silence. They walk single-file along ridges, carefully place their feet to avoid snapping dry twigs, suppress species-typical pant-hoots, and manually stifle loud vocalizations from excitable juveniles.

Tactical silence is equally apparent during intra-group social interactions. Subordinate males attempting to solicit illicit copulations with oestrus females under the watchful eye of an alpha male engage in nuanced, multisensory concealment. An alpha male exercises despotic mating privileges; a subordinate caught copulating faces brutal retaliatory violence. Subordinate males observe profound acoustic discretion, forgoing the loud, species-typical courtship vocalizations (“staccato panting”) that usually precede copulation. Instead, they position themselves behind natural buttress roots or dense foliage, out of the alpha’s line of sight, and rely entirely on silent visual gestures, such as gently extending a hand or peeling bark, to summon the female into an occluded gully.

Beyond acoustic silence, field ethologists like Frans de Waal and Jane Goodall documented instances of anatomical concealment. Subordinate males experiencing sudden penile erections in front of a dominant will actively adjust their posture or place both hands over their genitalia to physically conceal their arousal until the dominant looks away. This field evidence, combined with rigorous laboratory data from the competitive paradigm, confirmed that sensory concealment in chimpanzees is not a laboratory artifact. Rather, it represents an evolved, multisensory adaptation for navigating the treacherous social landscape of chimpanzee dominance hierarchies.

5.3 Auditory Occlusion vs. Visual Occlusion: Comparative Cognitive Load

Comparing auditory perspective taking with visual perspective taking reveals profound differences in sensory dynamics and cognitive load. Visual perspective taking is inherently spatial, geometric, and directional. To determine whether an individual sees an object, an observer must construct an imaginary linear vector connecting the competitor’s eyes to the target, verifying that no opaque barriers intersect the trajectory. This requires spatial-relational computation, yet the physical properties of the occluder remain static throughout the encounter.

Auditory perspective taking, by contrast, requires the continuous monitoring of physical output across time. Sound is non-directional in its dissemination; acoustic waves travel omnidirectionally through physical environments, reflecting off surfaces and bypassing visual occluders. Consequently, suppressing sound requires an animal to maintain constant self-monitoring: modulating locomotion speed, applying delicate tactile pressure to mechanical substrates, and inhibiting vocalizations. The animal must predict how its own bodily interactions with physical objects will sound to an external observer who lacks visual contact.

Despite these differences in sensory mechanics, empirical testing reveals that chimpanzees master both visual and auditory perspective taking with comparable efficiency within competitive contexts. In both domains, the cognitive load is managed through an intuitive, goal-directed heuristic: subordinate chimpanzees act to disrupt the perceptual channels of their competitors. Whether extinguishing a visual line of sight via an opaque barrier or neutralizing an acoustic channel via silent tunnels, the underlying mental architecture remains unified: the chimpanzee understands that external sensory input informs the competitor’s immediate behavior.

6. The Behavior-Reading Challenge: Daniel Povinelli and the Logical Problem

6.1 Formulation of ‘The Logical Problem’ in Comparative Cognition

The empirical triumphs of Hare, Call, and Tomasello did not go unchallenged. The most rigorous and persistent critique came from Daniel Povinelli and his collaborator Jennifer Vonk in a series of highly influential papers, most notably their 2003 paper in Trends in Cognitive Sciences, “Facing up to the animal mind: The ‘behavior reading’ vs. ‘mind reading’ debate,” and their 2004 follow-up, “We agree that we can go further: An answer to Call et al.” Povinelli and Vonk argued that the competitive paradigm, despite its methodological sophistication, had failed to provide definitive proof of Theory of Mind. They grounded this critique in a profound epistemological dilemma known as The Logical Problem.

The Logical Problem posits that because mental states (such as “seeing,” “believing,” or “knowing”) are inherently unobservable, an organism cannot perceive them directly. Instead, mental states are always causally linked to overt, observable behaviors and physical configurations (such as eye gaze, head orientation, body posture, and line of sight). Consequently, for any hypothesis which claims that an animal is reading an internal mental state ($MS$), one can formulate an alternative, functionally identical hypothesis which states that the animal is simply reading a complex, observable behavioral cue ($B$).

Povinelli and Vonk asserted that all the experimental successes observed in Hare et al.’s competitive paradigms could be fully explained without attributing any mentalistic concepts to the chimpanzee. In their view, the subordinate chimpanzee does not represent the internal proposition: “The dominant does not see the food.” Rather, the subordinate relies on an acquired or innate behavioral rule grounded in environmental invariants: “Do not go for food if there is an unobstructed, straight-line spatial trajectory between that food and the dominant’s eyes/face.” Because the behavioral rule predicts the dominant’s actions with absolute fidelity, postulating an intermediate psychological representation of “seeing” is, according to Povinelli, theoretically superfluous.

6.2 Low-Level Postulates: Postural Cues, Head Orientation, and Body Micro-Movements

To substantiate the behavior-reading alternative, skeptics formulated a battery of low-level mechanisms that could theoretically account for the performance of subordinates across the Leipzig experiments. One prominent hypothesis centered on the detection of postural micro-cues and intent movements. In natural social groups, a dominant individual preparing to claim a food item exhibits subtle bodily signals: tensing of the shoulder musculature, micro-adjustments of head orientation, pupil dilation, and changes in respiratory cadence. Skeptics argued that subordinates might monitor these micro-movements through the mesh partitions, using them as discriminative stimuli to abort approaches toward open food.

A second low-level explanation relied on the “Evil Eye” Hypothesis or conditioned gaze avoidance. Throughout primate ontogeny, direct visual fixation from a dominant individual functions as an explicit threat display. Subordinate primates learn early in life that being caught within a dominant’s direct gaze leads to immediate physical aggression. Consequently, subordinates develop a conditioned, somatic aversion to regions of space intersected by a dominant’s forward eye-gaze. When a piece of food is placed behind an opaque barrier, that specific region represents an “unwatched” zone, safe from gaze-elicited punishment. The subordinate does not need to understand what the dominant *experiences*; it simply moves away from the spatial vector of the dominant’s eyes to avoid triggering an attack.

Finally, critics pointed to potential statistical artifacts and peripheral scanning models. They argued that subordinates, when released into the testing arena, engage in rapid saccadic scanning. If an occluder is present, the physical board visually divides the arena, altering how the ape tracks the dominant’s face. Critics argued that subtle variations in latency, spatial layout, or habituation across successive trials could generate statistically significant differences in retrieval rates without requiring the attribution of perceptual perspective taking.

6.3 The Counter-Rebuttal by Tomasello, Call, and Hare (2003a, 2003b)

Tomasello, Call, and Hare mounted a systematic defense against the behavior-reading challenge in their 2003 companion papers in Trends in Cognitive Sciences: “Chimpanzees understand seeing: The end of a controversy?” and “Chimpanzee social cognition: Many levels, many tracks.” Their central counter-argument addressed scientific parsimony. While it is theoretically possible to devise a specific, post-hoc behavior-reading rule to explain any individual condition, the behavior-reading hypothesis collapses under its own complexity when forced to account for the entire, interlocking body of empirical findings.

To illustrate this, consider the cognitive gymnastics required of a behavior-reading chimpanzee across the suite of competitive experiments. The animal would require:

• Rule 1 for static opaque barriers: “Target food where a wooden board intersects the eye line of the dominant.”
• Rule 2 for transparent barriers: “Disregard Rule 1 if the board is transparent Plexiglas, even though the spatial geometry and board proximity are identical.”
• Rule 3 for auditory tasks: “Do not use noisy tunnels if the competitor’s head is turned 180 degrees, but use either tunnel if the head is oriented toward you.”
• Rule 4 for temporal knowledge tasks: “Avoid food if the door was open during baiting two minutes ago, but target it if the door was closed, despite the current scene being identical.”
• Rule 5 for switched-dominant tasks: “Disregard the prior visual history if the physical individual behind the door has been replaced by another dominant.”

Tomasello and colleagues argued that relying on an ever-expanding, ad-hoc catalog of hyper-specific behavioral rules violates the principle of parsimony (Occam’s Razor). The far more economical, coherent, and predictive explanation is that chimpanzees possess a single, integrated psychological construct: a concept of seeing. By attributing to the competitor an internal capacity to perceive the external world, the chimpanzee unifies visual line of sight, acoustic output, temporal history, and individual identity under a parsimonious socio-cognitive framework. In comparative cognitive ethology, intervening psychological variables are just as methodologically legitimate as intervening physical variables in mechanics.

7. Methodological Rigor and Experimental Controls in the Competitive Framework

7.1 Controlling for Dominant Behavior: The Delay-Release Mechanism

To permanently resolve the critique that subordinates were reacting to the overt behavioral cues or micro-movements of their rivals, Hare and colleagues introduced critical refinements to the physical operation of their testing paradigms. Central among these was the Delay-Release Mechanism. In earlier iterations, the doors of both contestants were raised nearly simultaneously. In refined iterations, the experimenters instituted a strict temporal offset: the subordinate was released into the arena while the dominant’s door remained securely locked and opaque, with the dominant completely invisible behind a secondary barrier.

Under this delay design, the subordinate was forced to enter the arena, scan the occluders, and commit to an approach trajectory before the dominant entered the space. By introducing this temporal lag, the subordinate could not use the dominant’s bodily cues, speed of approach, head orientation, or intent movements to guide its choice. The subordinate’s decision-making relied entirely on an internal, predictive representation of what the dominant *would* do once released, based entirely on the spatial configuration of the occluders or the prior visual history of the baiting sequence.

To confirm that dominant individuals were not providing subtle auditory or vibrational cues through the cage walls, researchers implemented high-resolution video recording, infrared movement sensors, and acoustic decibel meters inside the holding quarters. Frame-by-frame analysis demonstrated that dominants remained stationary behind their doors during the subordinate’s initial approach phase. Furthermore, quantitative eye-tracking and gaze-trajectory analyses of the subordinates revealed that their visual attention was directed toward the food and occluders, not focused on inspecting the dominant’s door seams for micro-movements. The subordinate’s foraging choices were proactive, predictive, and cognitive, rather than reactive.

7.2 Controlling for Spatial and Environmental Artifacts

A rigorous comparative psychology paradigm must account for non-social spatial biases that can inadvertently dictate animal behavior. In competitive testing environments, animals frequently display natural spatial perseveration (e.g., a baseline preference to turn left rather than right), positional attachments, or neophobic reactions to novel physical materials introduced into their living quarters.

To eliminate spatial artifacts, Hare, Call, and Tomasello implemented strict counterbalancing across all testing blocks. The spatial positioning of the opaque occluders, transparent controls, and food items was randomized on a trial-by-trial basis using predetermined pseudo-random matrices. If an occluder stood on the left side in Trial 1, it was shifted to the right side in Trial 2; the distance between the occluder, the arena walls, and the holding cages was varied systematically to prevent subordinates from relying on simple egocentric motor pathways.

Critically, the experimenters disentangled the physical distance to the barrier from the visibility parameters relative to the competitor. In specific control conditions, food items were positioned close to barriers that offered no visual protection from the dominant’s vantage point (e.g., barriers angled parallel rather than perpendicular to the line of sight). Subordinates systematically avoided these barrier-proximal items, proving that spatial proximity to a board was irrelevant; what mattered was the geometric shadow cast by the occluder across the dominant’s visual field. Generalized linear mixed models (GLMMs) isolated the main effect of visibility, confirming that neither spatial side biases nor proximity preferences accounted for the statistical distribution of subordinate choices.

7.3 Eliminating Experimenter Clever Hans Effects

In any cognitive study involving non-human animals, the threat of the Clever Hans effect—the inadvertent cueing of the animal by human handlers through unconscious postural adjustments, gaze direction, or breathing rhythms—must be systematically eliminated. Early ape language studies and communicative tasks were frequently criticized for lax experimental controls, where human experimenters were fully aware of the correct choice and unwittingly directed the ape’s attention toward it.

The competitive paradigm eliminated experimenter cueing through automation, physical partitioning, and double-blind testing protocols. The sliding mesh and opaque doors were operated via remote overhead mechanical pulleys, ropes, and counterweights controlled from behind opaque testing blinds or adjacent rooms. During baiting, human handlers followed strict, algorithmic choreography: entering the arena, placing the food items at millimeter-precise coordinates, aligning the occluders, and immediately exiting the testing area before any visual doors were opened. The handlers were isolated from the chimpanzees during the execution of the trials.

To guarantee inter-observer reliability, all competitive trials were recorded from multiple synchronized camera angles. The resulting digital video footage was subjected to blind coding by independent researchers who were completely unaware of the specific hypotheses, trial conditions (in cases where camera angles could isolate the subordinate’s approach without showing the occluder type), or baiting histories. Inter-observer reliability metrics routinely achieved near-perfect Cohen’s kappa coefficients ($kappa > 0.90$), confirming that the behavioral responses were unpolluted by human communicative cues.

8. The Structural Anatomy of Primate Social Knowledge: Seeing vs. Believing

8.1 Perceptual Perspective Taking (Level 1 and Level 2)

To contextualize the findings of the competitive paradigm within developmental psychology, researchers refer to John Flavell’s classical taxonomy of visual perspective taking. Flavell distinguished between two distinct stages in human ontogeny:

• Level 1 Perspective Taking: The capacity to compute what an agent can and cannot see (an understanding that an object is either visible or occluded from an observer’s line of sight). This ability emerges in human infants around 18 to 24 months of age.
• Level 2 Perspective Taking: The capacity to compute how a visible object appears to an observer from a specific, distinct spatial angle (an understanding that an object can present differing visual aspects, such as looking upside down or appearing partially obscured, to two observers who both see it simultaneously). This emerges in human children between 4 and 5 years of age.

The competitive paradigm provided conclusive proof that chimpanzees possess fully functional Level 1 Visual Perspective Taking. Subordinates clearly compute whether an item is within another individual’s visual field. They predict the behavioral consequences of visual occlusion with remarkable speed and flexibility across visual and auditory modalities, requiring no explicit behavioral shaping.

Evidence for Level 2 Perspective Taking in chimpanzees, however, remains elusive. When tasked with paradigms that require understanding how an object appears to another individual when the object’s identity, orientation, or color is altered by visual filters or viewing angles, great apes fail to show consistent performance. Chimpanzee social cognition operates on an object-centered, binary framework: an entity is either visually accessible to an agent, or it is blocked. The nuance of subjective internal representations—the understanding that a single physical object can be seen under different psychological descriptions—appears to be a uniquely hominin developmental milestone tied to the emergence of linguistic representation.

8.2 The Epistemic Boundary: Knowledge Versus False Belief

In 2008, thirty years after Premack and Woodruff launched the search for animal Theory of Mind, Josep Call and Michael Tomasello published a defining synthesis in Trends in Cognitive Sciences: “Does the chimpanzee have a theory of mind? 30 years later.” In this review, Call and Tomasello took stock of three decades of empirical research, dividing primate social cognition into distinct mentalizing components:

Mental State Category	Empirical Status in Chimpanzees	Representative Experimental Paradigms
Goals and Intentions	Robustly Present	Unwilling vs. Unable paradigms; Imitation of intended actions
Perception and Attention	Robustly Present	Competitive occluder tasks; Auditory concealment; Gaze-following around barriers
Knowledge and Ignorance	Robustly Present	Visual history tracking; Dominant-switch tasks; Delay-release baiting paradigms
False Belief	Consistently Absent (Traditional Behavioral Tasks)	Change-of-location tasks; Misleading appearance tasks; Deceptive displacement

The line drawn by Call and Tomasello was unambiguous: chimpanzees possess an understanding of perception-goal psychology, but lack a belief-desire psychology. Chimpanzees track what others see, what they know, and what their immediate behavioral goals are. However, they consistently failed traditional change-of-location false belief tasks (analogous to the classic Sally-Anne task used with human children). If a dominant chimpanzee possessed an explicitly false, counterfactual representation of the world, subordinates did not predict its actions based on that mistaken mental state. The epistemic boundary of the non-human hominid mind was bounded by reality: apes track knowledge (a true mental relation to a physical state of affairs) and ignorance (the absence of a mental relation), but could not represent subjective error.

8.3 Anticipatory Looking and Implicit Mental Attribution

For nearly a decade, Call and Tomasello’s 2008 conclusion stood as the canonical consensus of comparative psychology. Yet, in 2016, this boundary was upended by Christopher Krupenye, Fumihiro Kano, Satoshi Hirata, Josep Call, and Michael Tomasello in a study published in Science: “Great apes anticipate that other individuals will act according to false beliefs.” This breakthrough did not rely on physical food-retrieval competition; instead, it married the conceptual framework of the competitive paradigm with state-of-the-art, non-invasive infrared eye-tracking technologies.

In this study, chimpanzees, bonobos, and orangutans watched video dramas depicting a human actor engaged in a fierce, competitive conflict with an agent dressed in an ape suit (known as “King Kong”). In the critical false-belief condition, King Kong attacked the human, hid a stone inside one of two large haystacks, and then chased the human out of the testing arena. Once the human was out of sight, King Kong retrieved the stone and fled entirely, leaving both haystacks empty. The human then re-entered the scene to search for the hidden object. The human possessed a false belief: he believed the stone was still in Haystack A, where he had last seen it hidden, whereas the viewer (the ape) knew both haystacks were empty.

By measuring the apes’ anticipatory looking—where their eyes fixated on the screen *before* the human reached a haystack—the researchers revealed that great apes looked selectively at the haystack where the human falsely believed the item to be. Even though the apes knew the item was no longer there, they anticipated that the human would act based on an outdated, counterfactual mental representation. This finding demonstrated that while great apes struggle to navigate false beliefs in complex, active behavioral tasks that demand high levels of executive function and behavioral inhibition, they possess an implicit capacity for false belief attribution. This discovered continuity blurred the once-rigid epistemic boundary between human and non-human minds.

9. Comparative Primate and Mammalian Implementations of the Paradigm

9.1 Extensions to Bonobos (Pan paniscus) and Other Great Apes

Following the success of the competitive paradigm with Pan troglodytes, comparative researchers extended the framework across the entire hominid superfamily, most notably to bonobos (Pan paniscus). Bonobos and chimpanzees represent sister taxa, having diverged approximately 1.5 to 2 million years ago, yet their socio-sexual structures diverge substantially. Wild bonobo societies are characterized by female dominance or co-dominance, elevated levels of social tolerance, frequent non-conceptive socio-sexual interactions used for tension regulation, and reduced rates of lethal intra-group aggression.

When tested in competitive occluder tasks, bonobos demonstrated visual perspective-taking capabilities fully comparable to those of chimpanzees. However, their behavioral deployment of this cognitive capacity was heavily modulated by their unique emotional and motivational profiles. In semi-cooperative or co-foraging contexts, bonobos exhibited significantly higher levels of food-sharing and mutual tolerance near occluders, frequently allowing subordinates to feed peacefully. However, when strict competition was experimentally enforced, bonobos exploited opaque occluders and visual blind-spots with the same tactical precision as chimpanzees. This confirms that Level 1 visual perspective taking is a robust, phylogenetically conserved cognitive trait shared across the genus Pan.

Comparative testing across western lowland gorillas (Gorilla gorilla gorilla) and Sumatran and Bornean orangutans (Pongo abelii and Pongo pygmaeus) yielded complementary results. While gorillas often show reduced performance due to pronounced neophobia and strict spatial-social taboos surrounding direct eye contact with dominants, orangutans—solitary arboreal specialists—excelled in visual occlusion paradigms. The presence of visual perspective taking across all four extant great ape genera demonstrates that the evolutionary origin of perceptual mentalizing dates back at least 14 million years to the common ancestor of the family Hominidae.

9.2 Monkeys: Macaques, Baboons, and Capuchins Under Competition

Does perceptual perspective taking extend beyond the hominid radiation into Old World and New World monkeys? To answer this question, Jonathan Flombaum and Laurie Santos executed a brilliant, ecologically grounded series of competitive experiments with free-ranging rhesus macaques (Macaca mulatta) at the Cayo Santiago field station, published in 2005 in Current Biology: “Rhesus monkeys attribute perceptions in others.”

Flombaum and Santos developed a naturalistic food-stealing paradigm. An unrestrained macaque encountered two human experimenters sitting along a trail, each guarding a grape placed on a small platform. The experimenters adopted subtle anatomical postures to disrupt their gaze: one experimenter turned their head 45 degrees away while the other looked forward; one had their entire torso turned away; or one held a small opaque card directly over their eyes while the other held the card over their mouth. Across hundreds of trials, the rhesus macaques demonstrated a clear tactical preference: they selectively approached and stole food from the human who could not see them, navigating subtle variations in eye visibility, head tilt, and bodily orientation.

Studies with olive baboons (Papio anubis) and New World brown capuchin monkeys (Sapajus apella) have produced more variable results. While baboons demonstrate tactical deception and visual concealment in wild competitive contexts, capuchin monkeys often struggle in laboratory occluder tasks. This difficulty is frequently attributed to limitations in inhibitory control: capuchins exhibit high impulsivity, often prioritizing rapid food-retrieval attempts over strategic, occluded approach pathways. Nonetheless, the rhesus macaque data definitively confirms that the cognitive capacity to compute another agent’s perceptual line of sight is not unique to great apes, but evolved deep within the anthropoid primate lineage at least 25 to 30 million years ago.

9.3 Convergent Evolution: Corvids and Domestic Canines

The evolutionary pressures governing social competition are not confined to primates. In a striking example of convergent cognitive evolution, members of the corvid family—specifically common ravens (Corvus corax) and Western scrub-jays (Aphelocoma californica)—have demonstrated socio-cognitive abilities that rival, and in some domains surpass, those of non-human primates. Corvids survive by scatter-hoarding thousands of food caches across vast landscapes, facing constant competitive pressure from conspecific pilferers who observe caching events and subsequently steal the hidden food.

Pioneering investigations by Nicola Clayton, Nathan Emery, and Thomas Bugnyar revealed that food-caching scrub-jays and ravens are master tacticians of perceptual occlusion. When caching food in the presence of an observing conspecific, birds selectively cache behind physical barriers, exploit ambient shadows to obscure cache sites, and use acoustic suppression, caching in gravel substrates only when loud ambient noise masks their activity. If a scrub-jay is watched by a competitor while caching, it will return later, when alone, to dig up the food and re-cache it in a completely new, secure location. Corvids track the individual identity of observing birds, demonstrating a sophisticated memory-based model of conspecific visual history.

Domestic dogs (Canis lupus familiaris) also show acute sensitivity to human visual and auditory attention. In the classic “forbidden food paradigm,” dogs commanded by an owner not to eat a treat reliably obey as long as the human maintains direct eye contact. The moment the human turns their back, closes their eyes, or reads a book, the dog approaches and consumes the treat. Furthermore, dogs selectively use silent pathways when attempting to steal forbidden food, confirming that cross-species competitive pressures consistently drive the emergence of perceptual monitoring across phylogenetically distant mammalian and avian taxa.

10. Critiques, Alternative Paradigms, and Internal Debates

10.1 The Ecological Validity Critique of the Competitive Paradigm Itself

While the competitive paradigm liberated comparative psychology from the constraints of human-centric cooperative designs, it soon generated its own theoretical critiques. Foremost among these was the contention that Hare, Call, and Tomasello had overcorrected. Primatologists like Frans de Waal argued that by viewing chimpanzee cognition exclusively through the lens of cutthroat competition, the Leipzig school had constructed an overly agonistic, hyper-Machiavellian caricature of Pan troglodytes.

De Waal pointed out that wild chimpanzees engage in sophisticated forms of naturalistic cooperation, including collaborative hunting of red colobus monkeys, reciprocal grooming coalitions, political alliance formation, consolation of distressed victims of aggression, and targeted food sharing. By designing experimental apparatuses centered strictly around zero-sum resource monopolization, critics argued that the competitive paradigm might fail to measure the prosocial and cooperative cognitive architecture that chimpanzees genuinely possess.

The synthesis that has emerged does not invalidate the competitive paradigm; rather, it contextualizes it. Competition did not eliminate cooperation from chimpanzee psychology; rather, competitive motivation acted as a cognitive catalyst that unlocked latent socio-cognitive capacities that artificial cooperative designs had masked. Chimpanzees possess the cognitive machinery to understand seeing and knowing, but they prioritize its deployment in scenarios where social stakes are high and resource acquisition is actively contested. In recent years, researchers have designed balanced paradigms showing that when the cooperative task is ecologically intuitive (e.g., pulling heavy ropes collaboratively to drag large food trays), chimpanzees demonstrate impressive social coordination and mutual monitoring.

10.2 Replications, Non-Replications, and Methodological Sensitivities

The definitive test of any scientific breakthrough is empirical replication across independent laboratories. In 2002, Derek Karin-D’Arcy and Daniel Povinelli published an attempted direct replication of Hare et al.’s seminal 2000 occluder experiment in Animal Behaviour: “Do chimpanzees know what conspecifics see? A closer look.” Karin-D’Arcy and Povinelli reported that their cohort of chimpanzees at the New Iberia Research Center failed to demonstrate a statistically significant preference for food hidden behind opaque occluders, targeting open and hidden food items at nearly equal rates. They concluded that Hare et al.’s original findings were statistical anomalies or experimental artifacts.

Tomasello and Call immediately delivered a comprehensive rejoinder, identifying critical methodological discrepancies between the Leipzig setup and the New Iberia replication attempt. The most prominent flaws in Karin-D’Arcy and Povinelli’s study involved spatial geometry, temporal pacing, and social dynamics:

• Cage Dimensions and Spatial Spacing: In the New Iberia setup, the testing arena was cramped. The physical distance between the dominant’s door and the food was so short that the dominant could cover the distance in a fraction of a second, rendering any subordinate detour toward an occluded piece of food a high-risk liability.
• Dominance Relationship Integrity: Tomasello and Call discovered that many of the chimpanzee pairings used by Karin-D’Arcy and Povinelli lacked established, linear dominance relationships. In several pairings, the “subordinate” was actually dominant over or socially tolerant of the other ape, eliminating the competitive motivation required to drive tactical occlusion.
• Delay Timing: The New Iberia study failed to implement the precise delay-release protocols required to grant the subordinate the spatial head start necessary to make a deliberate, strategic foraging choice.

Subsequent independent replications by researchers at the Wolfgang Köhler Primate Research Center in Leipzig, the Yerkes National Primate Research Center in Atlanta, and the Primate Research Institute in Kyoto, Japan, robustly vindicated Hare and colleagues. Meta-analytic evaluations confirmed that when proper spatial distancing, clear social dominance gradients, and adequate delay-release mechanics are maintained, the competitive paradigm reliably yields robust, statistically significant evidence of visual perspective taking.

10.3 Human vs. Conspecific Competitors

An enduring methodological question within the competitive framework is the differential performance observed when chimpanzees compete against conspecifics versus human experimenters. While apes readily calculate visual lines of sight when facing a dominant chimpanzee, their performance often becomes erratic or attenuated when competing against a human handler adopting an adversarial posture.

Comparative psychologists attribute this disparity to differences in emotional arousal, social communication, and ontogenetic conditioning. Captive chimpanzees spend their lives interacting with human caretakers who provide food, enrichment, and medical care. These long-term associative histories can lead to confusion when a human suddenly adopts an agonistic role: the ape may struggle to interpret whether the human is behaving as an aggressive dominant, a negligent caregiver, or an ambiguous partner. The human body also exhibits morphological differences in gaze orientation, facial musculature, and sclera visibility (humans possess white sclera that make gaze direction conspicuous, whereas chimpanzees possess dark, pigmented sclera that mask direct eye direction).

Consequently, the competitive paradigm established a critical methodological principle: conspecific testing represents the gold standard for evaluating animal social intelligence. Pitting an animal against its ecological peers eliminates cross-species communicative interference, taps directly into innate behavioral responses, and guarantees that the cognitive challenges presented map cleanly onto the species’ evolutionary history.

11. Philosophical and Theoretical Implications for the Philosophy of Mind

11.1 Deconstructing Anthropocentric Conceptions of Theory of Mind

The empirical findings generated by the competitive paradigm forced a fundamental philosophical reassessment of Theory of Mind. Throughout the twentieth century, analytic philosophy of mind and cognitive psychology operated under an anthropocentric bias: mentalizing was conceptualized almost entirely as a linguistic, proposition-tracking capacity. Under this classical framework, to have a Theory of Mind was to hold mental representations of propositional attitudes: Agent X believes that [p], where [p] is a truth-evaluable linguistic statement.

The discovery that chimpanzees understand seeing, knowing, and intending in competitive environments without possessing human language broke this paradigm. Philosophers such as Stephen Stich, José Luis Bermúdez, and Ian Apperly argued for a radical decomposition of Theory of Mind from a monolithic, all-or-nothing cognitive module into a multi-tiered, modular cognitive architecture:

Cognitive Tier	Representational Content	Taxonomic Distribution	Primary Sensory/Cognitive Mechanism
Minimal / Perceptual ToM	Relational tracking of perception, gaze, attention, and spatial occlusion	Apes, Monkeys, Corvids, Canines	Geometric vector tracking; sensory channel monitoring
Epistemic / Informational ToM	Tracking perceptual history, individual familiarity, and ignorance	Great Apes, Corvids	Memory-based integration of past perceptual access
Full Propositional / Counterfactual ToM	Attribution of subjective false beliefs, meta-representation, and recursive perspectives	Human Children (3-5+ yrs), Adults	Linguistic proposition-embedding; dual-level counterfactual modeling

This taxonomy validates the concept of Minimal Theory of Mind. Animals do not require linguistic propositions to mentalize. Instead, they can construct non-propositional, relational representations: representing a relationship between an agent, an object, and a perceptual field ($Agent \rightarrow Sees \rightarrow Object$). By demonstrating that chimpanzees manipulate these relational representations during competition, the competitive paradigm established that the evolutionary origins of social cognition are non-linguistic, perceptual, and pragmatically embedded in action.

11.2 The Evolution of Intersubjectivity

The success of the competitive paradigm served as the empirical foundation for Michael Tomasello’s influential evolutionary framework: the Shared Intentionality Hypothesis. Tomasello argued that while the competitive paradigm proved that chimpanzees understand basic perception and knowledge, it also illuminated the precise cognitive chasm separating non-human great apes from modern humans.

Chimpanzees use their understanding of seeing and knowing almost exclusively for individualistic, competitive ends: to exploit blind spots, steal food, conceal copulations, and out-maneuver rivals. What they lack, according to Tomasello, is the motivation and cognitive architecture for shared intentionality: the capacity to participate in collaborative activities with shared goals, joint attention, and mutual commitments. In Tomasello’s view, early hominins encountered an ecological shift (such as obligate cooperative foraging or cooperative infant care) that transformed human psychology. Individuals who could not coordinate their mental states toward cooperative ends were selected against.

The evolutionary trajectory can thus be mapped with clarity:

1. Ancestral Baseline (Pan/Homo LCA): Possessed an individualistic, competitive Theory of Mind: capable of Level 1 visual perspective taking, auditory concealment, knowledge tracking, and tactical exploitation of conspecific perception.
2. Hominin Divergence (Early Homo): Evolved cooperative breeding, collaborative foraging, and shared intentionality, transforming perceptual perspective taking into recursive mentalizing: “I know that you know that I see…”
3. Modern Human Specialization: Evolved full linguistic propositional mind-reading, enabling explicit false belief attribution, cultural transmission, moral norms, and formal institutional reality.

The competitive paradigm did not merely reveal the chimpanzee mind; it mapped the ancestral baseline from which human intersubjectivity emerged.

12. Legacy of the Competitive Paradigm in Contemporary Cognitive Science

12.1 Transformation of Comparative Psychology Methodologies

The legacy of Brian Hare, Josep Call, and Michael Tomasello extends far beyond their original empirical discoveries. Their work precipitated a paradigm shift that permanently revolutionized comparative psychology. The era of evaluating non-human intelligence solely through cooperative, human-centric tasks—where captive animals were treated as deficient human children—came to a close.

The competitive paradigm established a new methodological golden rule: cognitive testing protocols must reflect the evolutionary ecology and natural motivational structures of the study organism. If a researcher wishes to measure the cognitive limits of an animal, the experimental apparatus must be designed around challenges that the animal evolved to overcome. This principle has been embraced across the cognitive ethology landscape. Cetacean researchers evaluate dolphin cognition through hydroacoustic tracking and fission-fusion alliance monitoring; elephant researchers utilize olfaction-based cooperative pulling rigs; and rodent researchers employ burrow-like, tactile navigation spaces rather than visually driven mazes.

Furthermore, the competitive paradigm established unprecedented standards of experimental rigor. The meticulous deployment of delay-release mechanisms, transparent controls, spatial counterbalancing, blind coding, and automated delivery systems established a high bar for contemporary animal cognition research, dismantling the casual anthropomorphism and low-level behavioral ambiguities that had hindered the discipline for decades.

12.2 The Modern Consensus on Chimpanzee Theory of Mind

As cognitive science navigates the twenty-first century, the empirical consensus regarding chimpanzee Theory of Mind has largely stabilized around the discoveries first unearthed by the competitive paradigm. Today, few comparative psychologists defend the old behavior-reading hypothesis in its extreme, radical form. The scientific community broadly accepts that chimpanzees possess a sophisticated, multi-layered social intelligence:

• They understand the visual channel: computing lines of sight, exploiting physical occluders, and differentiating between transparent and opaque materials.
• They understand the acoustic channel: actively concealing noise in the presence of competitors while ignoring acoustic output when competitors are already visually aware.
• They track epistemic states over time: distinguishing between conspecifics who witnessed an event and those who did not, updating these records when social partners are swapped.
• They understand goals and intentions: differentiating between human or conspecific partners who are unwilling to assist them versus those who are physically unable to do so.

The remaining frontiers of research no longer debate whether chimpanzees mentalize, but focus on the exact representational format of that mentalizing. Ongoing research programs explore the neurological substrates of primate social cognition using functional neuroimaging, high-speed eye-tracking, and computational models of social interaction. The boundary between implicit and explicit mentalizing remains an active arena of investigation, as does the degree to which non-human primates can simulate the counterfactual perspectives of others during high-stakes strategic games.

Ultimately, the competitive paradigm accomplished what decades of cooperative-communicative testing could not: it lifted the veil on the inner mental lives of our closest evolutionary relatives. By understanding the chimpanzee on its own terms—as a competitive, highly political primate navigating a complex social landscape—Brian Hare, Josep Call, and Michael Tomasello permanently restructured our understanding of the primate mind, demonstrating that the roots of social cognition run deep into our shared evolutionary history.

Conclusion

The transformation of primate cognitive science over the past quarter-century stands as a testament to the profound importance of evolutionary ecology in the study of animal minds. For over two decades following Premack and Woodruff’s initial inquiry, the field languished within an artificial, anthropocentric framework. By asking chimpanzees to engage in cooperative-communicative gestures foreign to their wild socioecology, researchers inadvertently engineered a negative consensus that relegated our closest evolutionary kin to the status of non-mentalizing biological automatons.

The competitive paradigm dismantled this negative consensus. By restructuring experimental architectures around conspecific competition, Brian Hare, Josep Call, and Michael Tomasello demonstrated that chimpanzees understand seeing, knowing, and intending when those calculations are anchored in evolutionary relevance. In doing so, they not only answered the empirical question first posed in 1978, but transformed the theoretical foundations of comparative psychology. They revealed that the human mind does not sit alone across an unbridgeable cognitive chasm; rather, it represents a cooperative, linguistic expansion of a deeply ancient, perceptual social intelligence forged millions of years ago in the competitive social arenas of the African forest canopy.

References

• Apperly, I. A. (2011). Mindreaders: The cognitive basis of “theory of mind”. Psychology Press. https://doi.org/10.4324/9780203834015
• Bermúdez, J. L. (2003). Thinking without language. Oxford University Press. https://doi.org/10.1093/0195159691.001.0001
• Bugnyar, T., & Kotrschal, K. (2002). Observational learning and the raiding of food caches in ravens, Corvus corax: Is it ‘tactical’ deception? Animal Behaviour, 64(2), 185–195. https://doi.org/10.1006/anbe.2002.3056
• Byrne, R. W., & Whiten, A. (Eds.). (1988). Machiavellian intelligence: Social expertise and the evolution of intellect in monkeys, apes, and humans. Oxford University Press.
• Call, J., & Tomasello, M. (2008). Does the chimpanzee have a theory of mind? 30 years later. Trends in Cognitive Sciences, 12(5), 187–192. https://doi.org/10.1016/j.tics.2008.02.010
• Clayton, N. S., Dally, J. M., & Emery, N. J. (2007). Social cognition by food-caching corvids. The western scrub-jay as a natural psychologist. Philosophical Transactions of the Royal Society B: Biological Sciences, 362(1480), 507–522. https://doi.org/10.1098/rstb.2006.1998
• de Waal, F. B. M. (1982). Chimpanzee politics: Power and sex among apes. Johns Hopkins University Press.
• Flavell, J. H. (1992). Perspectives on perspective taking. In H. Beilin & P. Pufall (Eds.), Piaget’s theory: Prospects and possibilities (pp. 107–139). Lawrence Erlbaum Associates.
• Flombaum, J. I., & Santos, L. R. (2005). Rhesus monkeys attribute perceptions in others. Current Biology, 15(5), 447–452. https://doi.org/10.1016/j.cub.2005.01.061
• Hare, B., Call, J., Agnetta, B., & Tomasello, M. (2000). Chimpanzees know what conspecifics do and do not see. Animal Behaviour, 59(4), 771–785. https://doi.org/10.1006/anbe.2000.1518
• Hare, B., Call, J., & Tomasello, M. (2001). Do chimpanzees know what conspecifics know and do not know? Animal Behaviour, 61(1), 139–151. https://doi.org/10.1006/anbe.2000.1517
• Karin-D’Arcy, M. R., & Povinelli, D. J. (2002). Do chimpanzees know what conspecifics see? A closer look. Animal Behaviour, 64(5), 789–799. https://doi.org/10.1006/anbe.2002.3059
• Krupenye, C., Kano, F., Hirata, S., Call, J., & Tomasello, M. (2016). Great apes anticipate that other individuals will act according to false beliefs. Science, 354(6308), 110–114. https://doi.org/10.1126/science.aaf8170
• Melis, A. P., Call, J., & Tomasello, M. (2006). Chimpanzees (Pan troglodytes) conceal visual and auditory information from others. Proceedings of the Royal Society B: Biological Sciences, 273(1594), 1649–1655. https://doi.org/10.1098/rspb.2005.3371
• Povinelli, D. J., & Eddy, T. J. (1996). What chimpanzees know about seeing. Monographs of the Society for Research in Child Development, 61(3), 1–189. https://doi.org/10.1111/j.1540-5834.1996.tb00534.x
• Povinelli, D. J., & Vonk, J. (2003). Facing up to the animal mind: The ‘behavior reading’ vs. ‘mind reading’ debate. In F. B. M. de Waal & P. L. Tyack (Eds.), Animal social complexity: Intelligence, culture, and individualized societies (pp. 379–403). Harvard University Press.
• Povinelli, D. J., & Vonk, J. (2004). We agree that we can go further: An answer to Call et al. Trends in Cognitive Sciences, 8(1), 8–9. https://doi.org/10.1016/j.tics.2003.11.002
• Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1(4), 515–526. https://doi.org/10.1017/S0140525X00076512
• Tomasello, M., & Call, J. (1997). Primate cognition. Oxford University Press.
• Tomasello, M., Call, J., & Hare, B. (2003a). Chimpanzees understand seeing, revisited. Trends in Cognitive Sciences, 7(4), 153–154. https://doi.org/10.1016/S1364-6613(03)00035-9
• Tomasello, M., Call, J., & Hare, B. (2003b). Chimpanzee social cognition: Many levels, many tracks. Trends in Cognitive Sciences, 7(9), 379–380. https://doi.org/10.1016/S1364-6613(03)00104-3
• Tomasello, M., Carpenter, M., Call, J., Behne, T., & Moll, H. (2005). Understanding and sharing intentions: The origins of cultural cognition. Behavioral and Brain Sciences, 28(5), 675–691. https://doi.org/10.1017/S0140525X05000129
• Whiten, A., & Byrne, R. W. (1988). Tactical deception in primates. Behavioral and Brain Sciences, 11(2), 233–244. https://doi.org/10.1017/S0140525X00048457
• Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children’s understanding of deception. Cognition, 13(1), 103–128. https://doi.org/10.1016/0010-0277(83)90004-5