The Helper/Hinderer Infant Experiment – Kiley Hamlin, Karen Wynn, and Paul Bloom

For centuries, the human moral faculty was widely assumed to be the late-emerging product of cultural transmission, verbal education, and disciplinary socialization. From the theological doctrines of original sin to the Enlightenment tabula rasa posited by John Locke, Western intellectual traditions consistently framed the neonate as an amoral or premoral creature—an anarchic bundle of vegetative drives lacking any intrinsic apprehension of good and evil. In developmental psychology, this presumption calcified through the twentieth-century constructs of Jean Piaget and Lawrence Kohlberg, who argued that genuine moral discernment emerges only after complex cognitive stages, linguistic competence, and the gradual de-centering of childhood egocentrism. Under this orthodoxy, the preverbal infant was seen as utterly devoid of socio-moral judgment, capable of perceiving the physical world only as a “blooming, buzzing confusion” and social interactions through the narrow lens of bodily gratification and reflexive distress.

This long-standing paradigm was radically upended in November 2007 with the publication of a landmark paper in Nature by J. Kiley Hamlin, Karen Wynn, and Paul Bloom, titled “Social evaluation by preverbal infants.” Conducted at the Yale Infant Cognition Center, this investigation presented an deceptively simple experimental paradigm: preverbal infants observed a brightly colored wooden character attempting and failing to climb a steep hill, only to be subsequently aided by an altruistic “Helper” or assaulted and obstructed by a malicious “Hinderer.” Through rigorous measurements of looking time and unprompted manual reaching choices, Hamlin and her colleagues demonstrated that infants as young as six months old systematically preferred the cooperative agent over the disruptive one. The implications reverberated across developmental psychology, evolutionary biology, and moral philosophy, suggesting that humans possess an innate, evolutionarily hardwired “moral sense” that operates long before the acquisition of language, explicit cultural instruction, or adult socialization.

The “Hill Climber” experiment did not merely challenge the chronological timeline of socio-cognitive development; it destabilized foundational tenets regarding the architecture of the human mind. By deploying non-verbal behavioral metrics to reveal sophisticated third-party moral appraisals in infants who could neither speak nor walk, the study catalyzed a profound nativist paradigm shift. Over the ensuing decades, the Helper/Hinderer paradigm expanded into a sprawling research program probing mentalistic attribution, the roots of retributive justice, the dark contours of in-group parochialism, and the neurobiological substrates of early social cognition. This comprehensive treatise offers an exhaustive analysis of the Helper/Hinderer infant experiments: tracing their theoretical antecedents, dissecting their methodological apparatus, evaluating empirical variations across age cohorts, examining fierce replication controversies, and exploring their enduring philosophical and evolutionary ramifications.

1. Historical and Theoretical Foundations of Moral Development

1.1 The Pre-Nativist Consensus: Piaget, Kohlberg, and the Tabula Rasa

Throughout the greater part of the twentieth century, developmental psychology was dominated by theoretical frameworks asserting that moral reasoning is a sophisticated cognitive achievement arrived at only after years of experiential scaffolding. Jean Piaget’s foundational treatise, The Moral Judgment of the Child (1932), established a constructivist model wherein children progress from an initial stage of “heteronomous morality” (or moral realism) toward “autonomous morality.” In Piaget’s constructivist paradigm, the young child is fundamentally egocentric, viewing rules as immutable, sacred fiats imposed by adult authority. Under heteronomous morality, children evaluate actions exclusively on the basis of their objective, physical consequences rather than the subjective intentions of the actor. An agent who accidentally breaks fifteen ceramic cups while opening a door is deemed categorically more culpable than an agent who maliciously shatters a single cup while attempting to steal jam. For Piaget, the preverbal infant existed in an undifferentiated sensorimotor stage, completely lacking the cognitive apparatus required to process abstract social relations, intentionality, or normative valences.

This constructivist trajectory was systematized and extended by Lawrence Kohlberg, who proposed a six-stage cognitive-developmental trajectory of moral growth organized into three overarching levels: preconventional, conventional, and postconventional. In Kohlberg’s paradigm, young children inhabit the preconventional level, wherein moral evaluation is dictated entirely by an obedience-and-punishment orientation (Stage 1) followed by naive instrumental hedonism (Stage 2). According to Kohlberg, an action is understood as “bad” solely if it elicits physical chastisement, and “good” if it serves the actor’s immediate self-interest. Like Piaget, Kohlberg maintained that moral cognition requires sophisticated linguistic competence, perspective-taking skills, and formal logical operations. Consequently, the infant was conceptually excluded from moral psychological inquiry. The infant mind was regarded as a blank slate—a passive, premoral tabula rasa upon which culture, parental discipline, and social conditioning would progressively inscribe ethical norms.

Concurrently, the behaviorist paradigm championed by John B. Watson and B.F. Skinner attributed prosocial conduct and moralized behavioral patterns exclusively to external conditioning, associative reinforcement, and environmental contingencies. Prosocial habits—such as sharing, taking turns, or refraining from violence—were conceptualized as conditioned operants shaped through continuous schedules of positive reinforcement, classical conditioning, and behavioral extinction. Within this radical empiricist framework, altruistic impulses and socio-moral evaluations were deemed fundamentally unnatural. Without systemic external intervention, an infant would remain entirely governed by biological appetites, immediate drive reduction, and unmediated hedonistic self-interest. Thus, across both cognitive-developmental and behaviorist schools of thought, the suggestion that a six-month-old infant could form third-party appraisals of social benevolence or malevolence was considered theoretical heresy.

1.2 The Nativist Counter-Revolution in Cognitive Science

The theoretical architecture began to fracture during the latter decades of the twentieth century under the impetus of the cognitive revolution, sparked in large measure by Noam Chomsky’s theory of Universal Grammar. Chomsky demonstrated that human infants acquire complex linguistic competencies with astonishing rapidity despite the poverty of the stimulus—the incomplete, noisy, and unannotated linguistic input provided by their environment. Chomsky posited the existence of an innate Language Acquisition Device (LAD), a biologically dedicated mental organ that equips the human infant with pre-existing, domain-specific computational constraints. This revolutionary epistemological stance provided a potent theoretical template for moral nativism. If the human mind possesses innate, specialized modules for processing the syntax of natural language, it became profoundly plausible that our species might also harbor an innate “moral grammar”—a suite of evolutionarily evolved cognitive heuristics designed to process the complex grammar of human social interaction.

Simultaneously, the nascent discipline of evolutionary psychology, spearheaded by figures such as Leda Cosmides, John Tooby, and Robert Trivers, provided the ultimate adaptive rationale for an innate moral architecture. Within social species, survival and reproductive success are intimately contingent upon navigating intricate webs of cooperation, reciprocal altruism, and coalitional alliances. Trivers’ theory of reciprocal altruism (1971) illustrated that for cooperative strategies to remain evolutionarily stable against free-riders and parasitic defectors, organisms must possess sophisticated cognitive mechanisms for social vigilance. They must be capable of rapidly detecting cheaters, remembering past interactions, and forming selective alliances with trustworthy partners. In an ancestral environment fraught with lethal inter- and intra-group hazards, an infant or young hominin capable of distinguishing benevolent cooperators from dangerous disruptors would enjoy a profound selective advantage. Socio-moral evaluation was thus reframed not as an arbitrary cultural artifact, but as a primary adaptive imperative encoded directly into human biology.

This evolutionary framework converged with the “Core Knowledge” theory pioneered by developmental cognitive scientists such as Elizabeth Spelke and Susan Carey. Through the inventive utilization of non-verbal, visual habituation methodologies, Spelke and Carey overturned Piagetian assumptions regarding infant object permanence and physical reasoning, demonstrating that infants possess innate, foundational systems of core knowledge regarding inanimate physical objects (object mechanics, cohesion, continuity), spatial geometry, and basic numerical values. If infants enter the world pre-equipped with intuitive physics and intuitive mathematics, scholars hypothesized that they must also possess a dedicated core knowledge system for “intuitive sociology”—a domain-specific capacity to conceptualize other organisms as teleological agents possessing mental states, goals, and normative relationships.

1.3 The Genesis of the Yale Infant Cognition Center Research Program

The direct institutional lineage of the Helper/Hinderer experiment resides within the Yale Infant Cognition Center, co-directed by developmental psychologist Karen Wynn and cognitive scientist and philosopher Paul Bloom. In 1992, Wynn had achieved international renown through her groundbreaking paper in Nature demonstrating that five-month-old infants possess preverbal mathematical competence—specifically, the capacity to track simple addition and subtraction operations (e.g., 1 + 1 = 2 and 2 – 1 = 1) enacted by small physical puppets hidden behind an occluding screen. Wynn established that infant looking-time measures could serve as a profoundly sensitive, high-resolution window into the latent conceptual competence of non-verbal minds, proving that infants form precise numerical expectations and experience measurable cognitive surprise when those expectations are violated.

Paul Bloom brought to this empirical crucible an expansive philosophical inquiry into the innate architecture of the mind, the cognitive origins of religion, and the foundations of morality. In seminal theoretical works such as Descartes’ Baby (2004), Bloom examined the intuitive dualism inherent to human psychology—the intuitive separation between physical bodies and immaterial minds—and began formulating hypotheses regarding the evolutionary origins of human ethical judgment. Bloom postulated that our sophisticated cultural moralities are not built from scratch upon a foundation of blank empiricism, but are instead biological elaborations constructed atop an evolutionarily ancient, unlearned moral core. What was desperately needed, however, was an empirical paradigm capable of testing whether this putative moral core was operational prior to the acquisition of language and culture.

The critical empirical breakthrough occurred with the arrival of J. Kiley Hamlin, then a doctoral student working under the joint mentorship of Wynn and Bloom. Hamlin possessed the experimental vision to translate abstract philosophical questions regarding moral agency, altruism, and malice into a tangible, non-verbal kinetic drama that could be presented to infants who possessed no linguistic capacity. Wynn’s methodological mastery of looking-time and reaching protocols, Bloom’s theoretical insights into evolutionary psychology and moral philosophy, and Hamlin’s inventive experimental design coalesced into a rigorous research program. Their central, audacious research question was straightforward: When completely stripped of linguistic labels, cultural symbols, and explicit pedagogical cues, can the preverbal human infant evaluate third-party agents solely based on the prosocial or antisocial nature of their unprovoked actions toward an unrelated peer?

2. Experimental Architecture: The Landmark 2007 Climber Paradigm

2.1 Physical Apparatus, Visual Stimuli, and Kinematics

To interrogate the moral evaluations of preverbal infants without introducing confounding social variables, Hamlin, Wynn, and Bloom engineered a highly standardized physical apparatus known colloquially as the “Hill Climber” paradigm. The experimental stage comprised an inclined ramp constructed from dense, green-foam-covered wood, measuring approximately 90 centimeters in length and angled upward at a consistent slope of roughly 30 degrees. The incline was explicitly structured with distinct topographic features: a broad, horizontal lower base, an uninterrupted ascending slope presenting a substantial gravitational challenge, and an elevated summit plateau representing the successful culmination of the climb.

The focal protagonist of this dynamic display was the “Climber”—a smooth, circular wooden disc, painted a uniform bright red, measuring roughly 10 centimeters in diameter. To imbue this inert geometric object with unmistakable cues of psychological agency without invoking complex, familiar human phenotypes, the researchers affixed large, plastic, movable “googly eyes” (approximately 2.5 centimeters in diameter) to its front face. The secondary interacting entities—the “Helper” and the “Hinderer”—were constructed from the identical wooden base material and bore the same prominent googly eyes, but were sharply differentiated from the Climber and from one another along both geometric and chromatic axes: one was an equilateral triangle, while the other was a perfect square, rendered in contrasting primary colors (typically saturated lemon yellow and cobalt blue).

Crucially, the kinematics of all characters were executed manually by a hidden experimenter operating through a concealed slot at the rear of the apparatus, manipulating the characters via rigid, unobtrusive rods. The movements were meticulously choreographed to violate standard Newtonian mechanics, displaying what developmental psychologist David Premack termed “self-propelled, non-Newtonian motion.” The characters accelerated spontaneously without external physical propulsion, changed velocity abruptly, altered their trajectories intentionally, and maintained their posture against gravity. These kinematic signatures function as universal, hardwired perceptual cues that immediately trigger the infant brain to categorize an entity not as an inert inanimate object, but as a living, autonomous intentional agent endowed with internal mental states, desires, and teleological goals.

2.2 The Prosocial Script: The Helper Event

The experimental script began with a standardized sequence establishing the Climber’s intentional goal. The infant, seated comfortably on their parent’s lap approximately 100 centimeters from the miniature stage, observed the red circular Climber positioned at the base of the green hill. In the opening baseline sequence, the Climber engaged in clear, teleological action: it faced upward toward the elevated plateau, oscillated gently side to side as if preparing for physical exertion, and then embarked on an ascent of the incline. After smoothly traversing roughly one-third of the slope, the Climber’s upward momentum abruptly stalled. It vibrated slightly, slid back down to the baseline plateau, paused momentarily, and initiated a second, identical ascent. On this second attempt, the Climber struggled upward slightly farther—reaching approximately the midpoint of the hill—before failing once more, tumbling backward to the base, and remaining motionless, thereby unambiguously communicating an unfulfilled, thwarted goal to the infant observer.

At this precise narrative juncture, the prosocial “Helper” event commenced. While the Climber rested at the midpoint of its third arduous attempt, the Helper agent (e.g., the yellow triangle) entered the stage from the bottom of the incline. The Helper moved upward along the slope with decisive, smooth kinematics until it positioned itself directly beneath the stalled Climber. Once situated at the Climber’s lower perimeter, the Helper delivered a series of synchronized, upward nudges. With this collaborative physical support, the Climber overcame the gravitational resistance of the slope and ascended directly onto the summit plateau.

Upon facilitating the Climber’s successful arrival at the summit, the Helper’s behavior reinforced the prosocial narrative. The Helper did not usurp the plateau or attack the protagonist; instead, it descended smoothly back down the hill and exited the stage entirely. Meanwhile, the Climber remained at the apex plateau, executing a subtle, rhythmic side-to-side oscillation—a kinematic script designed to simulate a jubilant, celebratory dance signifying a successfully resolved goal. The visual display concluded, the stage was occluded by a soft curtain, and the infant was allowed a brief cognitive pause before the next presentation.

2.3 The Antisocial Script: The Hinderer Event

In alternating presentation blocks, infants were exposed to the diametrically opposed antisocial “Hinderer” event. The foundational premise of this script was identical in its opening kinematics: the red circular Climber was positioned at the base of the incline, displayed the same preparatory oscillations, and initiated its first and second failed ascents, sliding backward down the incline to establish its unambiguous directional teleology. The infant observer was once again primed with the visual understanding that the Climber’s unyielding internal desire was to reach the summit of the hill.

The divergence occurred during the Climber’s third attempt. As the Climber struggled against gravity at the hill’s midpoint, the Hinderer agent (e.g., the blue square) entered the stage, not from the bottom, but dramatically from the summit plateau above. The Hinderer descended rapidly down the slope, positioning itself directly in front of the ascending Climber, physically barring its path. Instead of assisting, the Hinderer executed an aggressive, downward collision, physically bumping and shoving the Climber from above. Under the mechanical force of this hostile collision, the Climber was forcibly knocked backward, tumbling completely down the 30-degree incline until it landed inert at the very foot of the hill.

Following this destructive intervention, the Hinderer remained atop the slope momentarily before ascending back to the summit plateau and exiting the visual field. The Climber remained motionless at the base of the hill, its upward trajectory completely thwarted, its explicit goal violently subverted. The total duration of both the Helper and Hinderer events was rigorously standardized to ensure that infants in both conditions were exposed to precisely equivalent amounts of visual movement, spatial translation, and character presence, isolating the sociomoral valence of the interaction as the single manipulated independent variable.

3. Methodological Paradigms: Eliciting Choice in Preverbal Subjects

3.1 The Habituation and Familiarization Protocols

To accurately assess infant social evaluation without the confounding variable of novelty bias, Hamlin, Wynn, and Bloom employed a rigorous familiarization and habituation methodology. Because human infants exhibit an innate orienting reflex toward novel or unfamiliar visual stimuli, immediately presenting them with a choice between the Helper and Hinderer could result in idiosyncratic selections driven simply by which shape or color caught their eye last. To extinguish superficial perceptual interest, infants underwent alternating familiarization exposures prior to testing, observing repeated iterations of the Helper and Hinderer events in counterbalanced sequences (e.g., Helper-Hinderer-Helper-Hinderer or vice versa).

The habituation criteria were defined operationally using computerized gaze-tracking software monitored by hidden observers who were blind to the identity of the specific character roles. An infant’s looking time was tracked from the moment the curtain opened until the infant averted their gaze continuously for more than two consecutive seconds, or until a maximum ceiling duration of 30 seconds elapsed. Habituation was formally achieved when an infant’s total visual fixation across three consecutive trials dropped below 50% of their total fixation time during the first three trials, signifying that the physical mechanics of the climb, the hill, and the characters had been cognitively processed and integrated into their working memory.

Balancing these exposure cycles required exceptional experimental finesse. Preverbal infants possess limited cognitive stamina; excessive habituation leads inevitably to motor restlessness, fussiness, and experimental attrition, while insufficient habituation leaves perceptual novelties unattenuated. By implementing counterbalanced trial presentations and strictly limiting the total exposure protocol to an average of six to eight cycles, the researchers successfully mitigated fatigue while ensuring that the deep semantic meaning of the interactions—the socio-moral contrast between collaborative facilitation and hostile obstruction—was thoroughly registered by the infant’s cognitive architecture.

3.2 The Violation of Expectation (VoE) Looking-Time Measure

The first empirical metric utilized by Hamlin and colleagues was the classic Violation of Expectation (VoE) paradigm, a methodological gold standard developed in infant cognitive science by researchers like Renée Baillargeon. The theoretical premise of VoE rests upon a robust evolutionary principle: infants, like adult scientists, direct prolonged visual attention toward events that violate their intuitive models of the world. If an infant observes a physical object floating unsupported in mid-air or passing through a solid wall, their gaze lingers significantly longer on that scene compared to a physically normative event. When transposed to the social domain, VoE posits that if infants infer intentionality and social relationships, they will form distinct expectations regarding how agents ought to interact in the future.

In the VoE test trials of the 2007 study, infants who had been fully familiarized with both the Helper and Hinderer events were presented with a novel interaction scene. The green hill was present, and the red Climber was positioned at the base, flanked symmetrically by the Helper and the Hinderer, who sat motionless on either side. In one test event (the Expectation-Congruent event), the Climber ascended the slope and moved toward and snuggled next to the Helper. In the alternative test event (the Expectation-Incongruent event), the Climber moved toward and snuggled next to the Hinderer—the very entity that had violently knocked it down the hill moments earlier.

The empirical results were striking: infants looked significantly longer at the test event where the Climber approached the Hinderer than when it approached the Helper. This statistical elevation in fixation time demonstrated that preverbal infants were not merely processing the physical movements as disjointed mechanical animations. Rather, they had constructed a mental model of social affiliation: they expected the victim of social obstruction to display an avoidance response toward its tormentor and an affiliative approach toward its benefactor. When the Climber embraced the hostile agent, it violated the infant’s socio-cognitive expectations regarding third-party relationships, commanding prolonged, puzzled visual scrutiny.

3.3 The Two-Alternative Forced-Choice Manual Reaching Paradigm

While looking-time measures provided profound insights into the infant’s passive expectations, Hamlin, Wynn, and Bloom sought an even more decisive, ecologically valid behavioral metric: active, unprompted physical selection. To achieve this, they pioneered the application of the two-alternative forced-choice (2AFC) manual reaching paradigm to the domain of socio-moral evaluation. Looking time, while revolutionary, had long faced philosophical and methodological criticisms from skeptics who argued it merely reflected low-level perceptual salience, visual complexity, or ambiguous attentional drift. Active motor reaching, by contrast, represents an unambiguous, volitional behavioral commitment—an infant physically reaching out, grasping, and claiming a specific object represents an explicit assertion of positive valence or preference.

Following the familiarization phase, the apparatus was removed, and an experimenter who was entirely blind to the identities of the Helper and Hinderer entered the room. To prevent parental cueing—a fatal methodological artifact colloquially known as the “Clever Hans effect”—the infant’s parent was instructed to sit directly behind the child, close their eyes tightly, and maintain an entirely neutral, motionless posture throughout the assessment. The experimenter presented the infant with a broad, neutral white tray upon which the actual three-dimensional wooden characters—the Helper and the Hinderer—were mounted side by side, equidistant from the infant’s midline (approximately 30 centimeters apart).

The experimenter placed the tray just outside the infant’s reaching range, established eye contact with the infant’s forehead to avoid directional gaze cueing, lowered their head to occlude their own gaze, and slowly pushed the tray forward into the infant’s reaching space, vocalizing a neutral prompt: “Look, who’s here?” The infant’s behavioral response was recorded via high-definition cameras positioned at multiple angles. An independent coder scored the infant’s choice based on the character that was first physically contacted, touched, or grasped. This manual reaching protocol yielded an extraordinary window into preverbal moral evaluation, forcing the infant to translate their internal social judgments into a definitive, physical behavioral choice.

4. Experimental Controls and Mitigating Confounding Variables

4.1 Counterbalancing Morphological and Chromatic Variables

To establish beyond empirical reproach that infants were responding exclusively to the moral valence of the social interactions—rather than superficial sensory characteristics—Hamlin, Wynn, and Bloom instituted an exhaustive, multi-layered counterbalancing architecture. In any infant cognition paradigm involving colorful visual stimuli, researchers must contend with intrinsic sensory biases. Human infants frequently exhibit spontaneous, unconditioned perceptual preferences: they may preferentially orient toward higher-frequency wavelengths (such as red or yellow over blue), high-contrast geometric vertices (favoring sharp triangles over rounded circles or squares), or particular spatial orientations within their visual field.

To nullify these confounding variables, the researchers systematically crossed and rotated all physical parameters across their subject pool. Half of the participating infants observed an interaction where the Helper was a bright yellow triangle and the Hinderer was a deep blue square; for the other half of the cohort, these identities were precisely inverted, rendering the blue square as the Helper and the yellow triangle as the Hinderer. Furthermore, alternative color and shape pairings—including red and green hues, as well as circular versus rectangular configurations—were systematically deployed in supplementary control cohorts to ensure that no single physical phenotype was inherently tethered to a specific behavioral role.

Spatial presentation biases were similarly controlled with mathematical rigor during the two-alternative forced-choice manual reaching phase. The physical placement of the Helper and Hinderer characters on the presentation tray—left versus right side—was randomized and counterbalanced across trials. This precluded the possibility that an infant’s reaching choice was dictated by motoric handedness, unilateral hemispheric visual bias, or spatial perseveration. Through these rigorous controls, any statistical divergence in reaching preference could be definitively uncoupled from low-level morphological, chromatic, or spatial preferences, isolating social valence as the sole causal driver of infant behavior.

4.2 The Inanimate Control Condition

The most profound methodological challenge to the 2007 findings was the mechanistic critique: Could the infants’ reaching behavior be explained entirely by low-level physical dynamics rather than high-level social appraisals? In the visual displays, the Helper pushed an object upward, whereas the Hinderer knocked an object downward. Perhaps infants simply possessed an unlearned perceptual preference for objects associated with upward visual trajectories, or an aversion to objects that collided violently with other surfaces, independent of any socio-moral meaning.

To decisively dismantle this non-social interpretation, Hamlin and colleagues devised the “Inanimate Control Condition.” In this control experiment, the physical apparatus, the green foam hill, the kinematics, the velocity curves, the spatial paths, and the collisions were replicated with mechanical exactitude. However, the researchers made one single, critical modification: they stripped the googly eyes entirely from all three wooden characters. By removing the eyes, the visual stimuli were transformed from anthropomorphic entities into inanimate geometric shapes. The red disc no longer appeared as a living agent struggling to achieve a personal goal; instead, it was simply an inert wooden circle sliding mechanically up an incline due to unperceived physical forces, where it was subsequently bumped upward by an inanimate yellow triangle or knocked downward by an inanimate blue square.

The results of this control were definitive: When presented with the eye-less geometric shapes on the forced-choice tray, the infants showed no statistically significant reaching preference whatsoever. They selected the “pushing-up” shape and the “knocking-down” shape at absolute chance levels (50% vs. 50%). This dramatic attenuation of preferential reaching in the absence of eyes proved that infants were not computing simple physical kinematics, vectors of motion, or collisions. The preferential reaching observed in the experimental condition was fundamentally predicated upon the perception of agency. Only when the entities were recognized as psychological beings engaged in intentional, goal-directed behavior did the infant’s socio-moral evaluation system activate.

4.3 Movement Direction and Perceptual Motion Trajectories

Even with the inanimate control, lingering concerns remained regarding whether the infants were reacting to specific motion trajectories when embodied by agents. For instance, did infants possess an unconditioned socio-perceptual preference for agents that move in an ascending, upward path versus those that descend? In the natural environment, upward motion might be subconsciously associated with growth, triumph, or levity, while downward motion could evoke falls, danger, or failure. To disentangle this confound, the researchers analyzed and manipulated the baseline trajectory profiles of the agents independent of their interaction with the Climber.

In supplementary experimental variations, the timing and spatial vectors of the agents were decoupled from the direct assistance or obstruction of the Climber. In these baseline trajectory controls, infants were familiarized with agents that traversed the green hill in upward or downward trajectories without ever making contact with, or altering the trajectory of, another character. When subsequently offered a manual choice between an agent that had moved upward along the slope and an agent that had moved downward, infants demonstrated no systematic preference for the upward-moving agent. Their choices hovered strictly around the null hypothesis.

These trajectory controls confirmed that the infant’s evaluation was inherently relational and social. The infant did not care whether an agent moved up or down the hill in an absolute physical sense; the infant cared about the impact of that movement upon the intentions of another sentient agent. The Helper was valorized not because it ascended the slope, but because its physical contact facilitated the Climber’s internal goal; the Hinderer was condemned not because it descended, but because its intervention actively thwarted that goal. The infants were parsing the drama through the sophisticated lens of intentional teleology and relational morality.

5. Empirical Findings Across Developmental Age Cohorts

5.1 The Baseline 10-Month-Old Cohort Findings

The foundational empirical data reported in the 2007 Nature paper focused initially on a cohort of 10-month-old infants—an age where gross motor reaching is well established, but expressive linguistic capability remains completely absent. In the two-alternative forced-choice manual reaching task, the behavioral preference for the prosocial character was overwhelming. Out of a total sample of twelve 10-month-old infants tested in the primary experimental condition, every single infant (12 out of 12, or 100%) systematically reached for the Helper agent over the Hinderer agent (binomial test, p < 0.001).

This striking uniformity was virtually unprecedented in developmental psychology, where behavioral noise, infant distractibility, and individual variation routinely yield far more modest statistical margins. The robust 100% preference held steady regardless of whether the Helper was the yellow triangle or the blue square, and irrespective of which side of the presentation tray the character was positioned. When replication cohorts were added and procedural variations explored, the overall reaching preference among 10-month-olds reliably settled between 80% and 90%, maintaining overwhelming statistical significance across diverse testing environments.

Furthermore, these behavioral reaching selections were corroborated by the Violation-of-Expectation looking-time metrics recorded from an independent group of 10-month-olds. As detailed previously, 10-month-old infants looked significantly longer when the Climber subsequently approached the Hinderer compared to when it approached the Helper. This dual-methodological convergence provided robust, multi-modal validation: 10-month-old infants not only preferred prosocial individuals in their own direct behavioral choices, but they also held abstract third-party expectations that other agents would—and should—affiliate with those who had helped them rather than those who had harmed them.

5.2 Replicating Success in the 6-Month-Old Cohort

The most profound theoretical contribution of the 2007 study emerged when the research team extended the identical paradigm downward to a cohort of 6-month-old infants. At six months of age, human infants are extraordinarily neurologically immature. They cannot walk, crawl, or produce words; they have barely mastered independent sitting, and their manual reaching capabilities are newly emergent, characterized by unrefined gross motor swipes rather than precise grasping. If moral evaluations were the downstream consequence of prolonged cultural socialization, linguistic instruction, or advanced peer interactions, six-month-olds should have performed at absolute chance.

To accommodate the motoric limitations of this younger cohort, Hamlin and colleagues refined the forced-choice protocol: rather than demanding a definitive grasp, coders scored the infant’s choice based on the character that the infant first made clear, sustained tactile contact with using an open palm or finger swipe. The empirical findings were extraordinary: 14 out of 16 six-month-old infants (87.5%) reliably reached for or touched the Helper over the Hinderer (binomial test, p = 0.004). The prosocial preference was virtually as robust in infants who had spent only half a year on Earth as it was in the older cohort.

Intriguingly, however, a critical developmental dissociation emerged in the looking-time data. While 6-month-olds exhibited a robust manual reaching preference for the Helper, their Violation-of-Expectation looking times failed to show the systematic difference observed in the 10-month-olds; they did not look significantly longer when the Climber approached the Hinderer versus the Helper. Hamlin and colleagues interpreted this dissociation as revealing an ontogenetic progression in social-cognitive complexity. At six months, infants possess the capacity to form personal socio-evaluative representations—”I like the one that helped; I dislike the one that hindered”—which guides their direct motor reaching. However, the higher-order capacity to project those socio-relational expectations onto third parties—anticipating that an independent agent will selectively affiliate with its benefactor—develops slightly later, crystallizing between six and ten months of age. Most importantly, the success of the 6-month manual reaching task dealt a catastrophic blow to the tabula rasa consensus, precluding long-term cultural conditioning as a necessary prerequisite for moral appraisal.

5.3 Extending the Paradigm to 3-Month-Olds via Preferential Looking

Emboldened by these discoveries, Hamlin, Wynn, and Bloom sought to locate the absolute ontogenetic emergence of social evaluation. Can socio-moral parsing be detected in neonates who are entirely incapable of manual reaching? In a landmark follow-up study published in 2010 in Developmental Science, titled “Three-month-olds show a negativity bias in their social evaluations,” the researchers adapted the Hill Climber paradigm for infants aged merely 12 to 14 weeks. Because 3-month-olds lack the physical neuromuscular maturity to reach for objects on a tray, the team abandoned the manual reaching task in favor of the high-resolution Preferential Looking paradigm.

In this adapted protocol, 3-month-old infants were familiarized with the identical animated Hill Climber dramas. Following familiarization, the Helper and Hinderer characters were placed side-by-side on a large visual screen, and an automated corneal reflection eye-tracking system, complemented by blind human video coders, recorded the exact duration of the infant’s visual fixations toward each character over extended, multi-second presentation intervals. Decades of cognitive research had firmly established that in the absence of active reaching, preverbal infants direct their visual attention preferentially toward social entities that they find appealing, positive, or rewarding.

The experimental results extended the boundaries of moral psychology to the very dawn of human infancy: 3-month-old infants looked significantly longer at the Helper character than at the Hinderer character. When individual looking distributions were categorized, a commanding majority of the 3-month-old cohort exhibited a pronounced visual preference for the prosocial agent. The demonstration that rudimentary socio-moral valence parsing operates at twelve weeks of postnatal life provided arguably the most powerful empirical evidence to date in favor of evolutionary moral nativism. At three months, an infant has had virtually negligible social interaction outside of immediate maternal feeding and caretaking routines. The cognitive capacity to parse a dynamic third-party conflict, identify the victim’s thwarted goal, evaluate the moral valence of two novel geometric agents, and form an evaluative preference based entirely on their benevolence or malevolence appears to be a biological component of the human mind’s core architecture.

6. Valence Parsing: Positive Attraction Versus Negative Avoidance

6.1 Incorporating the Neutral Agent Paradigm

The initial 2007 findings established an unequivocal behavioral divergence: infants preferred the Helper over the Hinderer. However, this binary choice revealed a profound theoretical ambiguity that struck at the heart of evolutionary moral psychology: What was the primary psychological engine driving this preference? Did the infants reach for the Helper because they were actively drawn to prosocial altruism (positive attraction), or were they simply repulsed by the violent, obstructive actions of the Hinderer and seeking to avoid it at all costs (negative avoidance)? Alternatively, were both motivational systems operating concurrently with equal potency?

To untangle these competing hypotheses, Hamlin, Wynn, and Bloom designed a sophisticated triadic experimental architecture, detailed in their subsequent 2007 analyses and expanded in follow-up empirical papers. They introduced a third, baseline character into the Hill Climber universe: the “Neutral Agent.” Morphologically identical to the other agents—bearing the same googly eyes and distinct geometric and chromatic characteristics (e.g., an orange cylinder or a green diamond)—the Neutral Agent traversed the incline without interacting with the Climber in any way. In some trials, the Neutral Agent simply slid up the hill, passed by the Climber without contact, and exited; in others, it moved downward along the incline, ignoring the struggling protagonist entirely. It was present on the stage for an identical duration, exhibited the same kinematics of agency, but performed neither a helpful nor a harmful action.

The researchers then presented infants with two novel, distinct two-alternative forced-choice manual reaching conditions:

• Helper versus Neutral: Pitting prosocial assistance directly against inert, non-interactive baseline agency.
• Hinderer versus Neutral: Pitting antisocial obstruction directly against inert, non-interactive baseline agency.

By measuring infant choices against a standardized neutral baseline, the researchers could mathematically calibrate the precise psychological weighting of social reward versus social threat avoidance in the nascent human mind.

6.2 Empirical Superiority of the Negativity Bias

The results of this triadic choice architecture yielded a fascinating and nuanced asymmetrical pattern. When infants were presented with the Hinderer versus Neutral condition, their behavioral choice was virtually categorical: infants overwhelmingly rejected the Hinderer, selecting the Neutral character at rates exceeding 80% to 85%. The antisocial agent was subjected to systematic, widespread behavioral avoidance.

By contrast, when infants were presented with the Helper versus Neutral condition, the preference for the Helper was statistically significant, but noticeably more moderate—infants selected the Helper over the Neutral agent roughly 65% to 70% of the time. While infants were undeniably capable of positive social evaluation—they consistently chose the benefactor over the bystander—the statistical magnitude of their aversion to the antisocial disruptor was consistently and significantly more pronounced than their affinity for the prosocial collaborator.

This empirical divergence is deeply consistent with broad evolutionary psychological theories regarding the adaptive primacy of threat detection. In the brutal calculus of natural selection, failing to affiliate with a potentially helpful ally represents a missed opportunity, but failing to avoid an aggressive, unpredictable, or malicious predator can prove instantly fatal. As evolutionary theorists have long maintained, natural selection heavily penalizes complacency toward antisocial threats far more rigorously than it rewards the pursuit of benevolent partnerships. The infant cognitive architecture reflects this evolutionary asymmetry: while humans possess an innate appreciation for benevolence, our socio-moral machinery is first and foremost a sentinel system wired to detect, flag, and avoid malevolent actors.

6.3 Deconstructing the Architecture of Infant Social Valence

These findings allowed Hamlin, Wynn, and Bloom to articulate a sophisticated dual-system processing model of infant socio-moral valence. Infant social evaluation does not operate as a monolithic, one-dimensional continuum stretching smoothly from -1 (bad) to +1 (good). Rather, it comprises two distinct, semi-autonomous neuro-cognitive processing streams: an early-maturing, hyper-sensitive Aversive Detection System dedicated to encoding violations of physical and goal integrity, and a complementary Affiliative Orientation System dedicated to identifying opportunities for cooperative exchange and coalitional bonding.

This model fundamentally challenged reductionist interpretations that sought to explain the Helper/Hinderer results through simple emotional contagion. Skeptics had suggested that infants merely experienced a reflexive, visceral distress upon seeing the Climber knocked down the hill, and that this negative emotional state became associatively paired with the Hinderer via classical conditioning. However, Hamlin’s work demonstrated that infants perform a structured, teleological appraisal of the interaction. The infant does not merely register diffuse negative affect; they track the specific causal agency involved. The Climber is identified as the victim whose internal teleology was blocked, while the Hinderer is encoded as the external causal source of that disruption.

Furthermore, these developmental findings seamlessly integrated with the broader theoretical frameworks of psychological negativity dominance pioneered by Paul Rozin and Edward Royzman, as well as Roy Baumeister’s seminal thesis that “Bad is Stronger than Good” (2001). Across adult human psychology, negative events, negative feedback, and negative social interactions consistently exert a far more powerful impact on attention, memory, and physiological arousal than objectively equivalent positive events. Hamlin and colleagues demonstrated that this psychological asymmetry is not a cynical artifact of adult life, accumulated trauma, or cultural conditioning; it is an innate, foundational organizing principle of the human socio-moral mind, operational from the earliest weeks of life.

7. Intentionality, Agency, and Moral Mathematics

7.1 Mental State Attribution: Intentions Versus Outcomes

A foundational tenet of mature, civilized moral philosophy—enshrined in legal codes worldwide via the doctrine of mens rea—is that moral culpability depends fundamentally upon mental states rather than physical outcomes. A person who causes accidental harm while attempting to do good is judged far less harshly than a person whose malicious plot accidentally produces a beneficial outcome. As noted previously, Jean Piaget adamantly claimed that young children are completely blind to this distinction until around age seven or eight, remaining trapped in a state of moral realism where consequences alone dictate culpability.

To directly test whether preverbal infants could separate subjective intentions from objective physical outcomes, Kiley Hamlin published a revolutionary 2013 study in the Proceedings of the National Academy of Sciences (PNAS), titled “Failed attempts to help and hinder: Preverbal infants prefer the intentional agent.” Hamlin modified the experimental apparatus to pit intentions directly against physical consequences through four distinct experimental scripts:

• Successful Helper: An agent that intends to help and successfully facilitates the goal.
• Failed Helper: An agent that clearly tries to help the protagonist, but fails due to physical incapacity, with the protagonist sliding back down the hill.
• Successful Hinderer: An agent that intends to thwart the protagonist and successfully knocks it down.
• Failed Hinderer: An agent that clearly attempts to block and knock down the protagonist, but slips, accidentally resulting in the protagonist ascending to the summit.

When eight-month-old infants were presented with a forced choice between the Failed Helper (positive intention, negative outcome) and the Successful Hinderer (negative intention, negative outcome), the infants overwhelmingly chose the Failed Helper. Even though the protagonist ended up at the bottom of the hill in both scenarios, the infants valorized the agent’s benevolent intent. Even more remarkably, when eight-month-olds were offered a choice between the Failed Helper (positive intention, negative outcome) and the Failed Hinderer (negative intention, positive outcome), the infants systematically reached for the Failed Helper. They actively rejected the agent that had caused a positive physical outcome because its underlying subjective intention had been malicious. This empirical demonstration obliterated Piaget’s constructivist timeline: preverbal infants do not evaluate actions based on crude physical outcomes; they possess an intuitive Theory of Mind that privileges an actor’s internal mental desires over the brute material consequences of their behavior.

7.2 Context-Dependent Morality and the Roots of Retributive Justice

Having proven that infants possess an innate appreciation for benevolence and intention, Hamlin and her colleagues ventured into a vastly more complex, provocative domain: Do infants possess the seeds of retributive justice? In civilized adult jurisprudence, aggression is not universally condemned. While unprovoked violence against an innocent party is strictly prohibited, the application of retributive force against a criminal or antisocial transgressor is universally recognized as morally legitimate punishment. Is this punitive distinction an advanced, culturally scaffolded philosophical construct, or is retributive justice woven into our core biological fabric?

In a groundbreaking 2011 paper published in the Proceedings of the National Academy of Sciences, Hamlin, Wynn, Bloom, and Neha Mahajan tested whether preverbal infants’ socio-moral evaluations were context-dependent. Infants first observed a primary puppet show in which a protagonist was either systematically helped or systematically hindered by another character, establishing that secondary character’s clear status as either a “prosocial target” or an “antisocial target.” Subsequently, the infants watched a secondary drama in which a third agent stepped in and delivered either a prosocial action (giving a ball back) or an aggressive, antisocial action (stealing a ball or knocking the target down) toward this previously established character.

The findings revealed an astonishing level of socio-moral sophistication. When the target of the secondary interaction was a prosocial individual, infants maintained their default moral preference: they liked the agent who was nice to the good guy, and rejected the agent who harmed the good guy. However, when the target of the secondary interaction was the antisocial Hinderer, the infants completely inverted their default preferences: 8-month-old infants actively preferred the agent who punished, attacked, or hindered the bad actor. They rejected the agent who tried to help or reward the malefactor. This profound finding demonstrated that infants do not simply harbor a blanket, reflexive aversion to physical aggression; they intuitively understand the concept of desert. Aggression directed toward an innocent victim is coded as a crime; aggression directed toward a perpetrator is coded as justice. The preverbal human mind already harbors the intuitive scaffolding of a retributive penal code.

7.3 The Box-Opening and Ball-Dropping Paradigms

To eliminate any remaining skepticism that the Hill Climber findings might be an idiosyncratic artifact of that specific physical incline or geometric puppet design, Hamlin and Wynn formulated entirely new experimental ecologies that radically diversified the narratives of assistance and hindrance. Two of these paradigms became foundational cornerstones of modern developmental literature: the “Box-Opening Paradigm” and the “Ball-Dropping Paradigm.”

In the Box-Opening Paradigm, a central protagonist—often an anthropomorphic stuffed animal puppet, such as a dog or tiger—struggles to open a clear, heavy-lidded plastic box containing an attractive rattle. The protagonist repeatedly grasps the lid and attempts to lift it, but the lid is too heavy and continuously falls shut, communicating a transparent, thwarted goal. In the prosocial condition, a second puppet (the Helper) runs in, grasps the opposite side of the lid, and pulls collaboratively until the box opens fully, allowing the protagonist to retrieve the prize. In the antisocial condition, a third puppet (the Hinderer) runs in, leaps on top of the lid, and violently slams it shut, preventing the protagonist from accessing the toy before sprinting away.

In the Ball-Dropping Paradigm, a protagonist puppet engages in an interactive game of tossing a bright rubber ball back and forth across a small stage. Inevitably, the ball slips past the protagonist and rolls toward one of two secondary puppets sitting at the corners. In the prosocial script, the secondary puppet catches the ball, rolls it gently back to the protagonist, and resumes the cooperative game. In the antisocial script, the secondary puppet intercepts the ball, turns its body away, and runs offstage with the stolen object. When tested with manual reaching choices across both the Box-Opening and Ball-Dropping paradigms, infants at 5, 9, and 14 months overwhelmingly selected the prosocial puppets over the antisocial thieves and saboteurs. This cross-situational consistency proved decisively that infant moral evaluation is not tied to specific spatial kinematics or hill-climbing tropes, but operates across diverse, generalized social domains.

8. In-Group Favoritism, Similarity, and Parochial Altruism

8.1 The Minimal Group Paradigm in Infancy

While the early findings revealed a remarkably sophisticated moral core, subsequent investigations by the Yale team revealed that the infant moral architecture is far from an idealized, universalist paradise of unconditional human brotherhood. In social psychology, the “Minimal Group Paradigm,” pioneered by Henri Tajfel, demonstrated that adult humans will rapidly form tribal in-group alliances and display out-group bias based on the most absurd, trivial, and arbitrary criteria, such as the flip of a coin or preferences for abstract paintings. Neha Mahajan and Karen Wynn sought to determine how deeply rooted this minimal group psychology is in human ontogeny.

In a creative 2012 study, Mahajan and Wynn established an infant minimal group paradigm based on an arbitrary, early-emerging cultural marker: culinary taste preferences. Infants were presented with two bowls of food: one containing sweet graham crackers, the other containing green beans. The infant was invited to taste both and clearly indicate their preferred food (an overwhelming majority selected graham crackers, though a minority chose green beans). Following this baseline selection, the infant observed two animal puppets express intense emotional reactions to the same foods: one puppet vocally praised the food the infant liked (“Mmm, yum! I love graham crackers!”), while the second puppet praised the food the infant disliked (“Mmm, yum! I love green beans!”).

When subsequent manual choice tests were administered, infants overwhelmingly chose to play with and reach for the puppet that shared their food preferences. The infant cognitive system exhibited instant homophily—the tendency to associate with, trust, and favor similar others. Even an arbitrary dietary preference was seized upon by the infant’s social brain as a definitive coalitional marker, instantly bisecting the world into an “us” (the in-group member who shares my preferences) and a “them” (the out-group member who deviates from my preferences).

8.2 The Darker Side of Infant Social Evaluation

The evolutionary consequences of this tribal categorization were laid bare in a profoundly unsettling 2013 study published in Psychological Science by Hamlin, Mahajan, Zoe Liberman, and Karen Wynn, titled “Not like me = bad: Infants prefer those who harm dissimilar others.” The researchers combined the minimal group food-preference protocol with the retributive justice paradigm to ask a critical question: How does an infant’s moral evaluation interact with their coalitional biases when a third party interacts with an in-group or out-group member?

Infants were first exposed to puppets that established themselves as either similar (sharing the infant’s food choice) or dissimilar (opposing the infant’s food choice). Subsequently, infants watched these similar and dissimilar puppets become the targets of third-party prosocial assistance or antisocial harm in the classic Ball-Dropping or Hill Climber scenarios. The results confirmed the darker predictions of evolutionary anthropology:

• When the target was a similar individual (in-group), infants maintained their standard moral posture: they strongly preferred the agent who helped the similar individual, and vehemently rejected the agent who harmed it.
• When the target was a dissimilar individual (out-group), the infants’ moral preferences were completely inverted: infants systematically preferred the agent who harmed, thwarted, or stole from the dissimilar individual.

This finding revealed that the preverbal human infant does not possess a universalist, Kantian ethic that respects the bodily integrity and goals of all sentient beings equally. Rather, the infant mind is wired for parochial altruism—an evolutionary adaptation wherein prosocial norms, empathy, and protection are strictly reserved for members of the in-group, while hostility, aggression, and schadenfreude toward out-group members are actively celebrated and rewarded. The very same cognitive machinery that enables an infant to love justice and reward helpers simultaneously predisposes the infant to delight in the mistreatment of those who are perceived as fundamentally “other.”

8.3 Balancing Universalist Ethics and In-Group Cohesion

These findings illuminate the fundamental developmental friction that characterizes the human moral condition: the perpetual clash between core, universal prosocial intuitions and ancient, ethnocentric tribal instincts. From an evolutionary perspective, this cognitive duality is precisely what one would expect. Our hominin ancestors evolved in small, tightly knit hunter-gatherer bands locked in perpetual, high-stakes competition for scarce territory, game, and resources. In such an ecological furnace, unconditional universal benevolence toward all hominins would be an evolutionary death sentence. An individual who extended altruism indiscriminately to hostile out-groups would quickly be exploited, outcompeted, or killed.

Survival demanded a dual-core moral psychology: hyper-cooperation, empathy, and retributive fairness within the coalition, coupled with deep suspicion, hostility, and preemptive aggression between coalitions. Hamlin, Wynn, and Bloom’s empirical discoveries revealed that human culture does not invent tribalism, racism, or xenophobia out of whole cloth; rather, these dark social realities exploit ancient, preverbal psychological fault lines that split the world into “like-me” and “not-like-me” within the first year of life.

Consequently, the developmental trajectory of human morality is not a simple, linear ascent from ignorance to enlightenment. Rather, it represents a lifelong cognitive struggle wherein our expanding capacities for language, rational deliberation, and formal education must actively work to suppress, override, and expand the narrow boundaries of our innate parochialism. The challenge of modern civilization is to recruit the innate moral core—our empathy for victims and admiration for helpers—while using reason and cultural scaffolding to aggressively dismantle the tribal boundaries that artificially limit who qualifies as a legitimate member of our moral universe.

9. Methodological Debates, Replication Efforts, and Skepticism

9.1 The Perceptual Cue Critique: Scarf, Imuta, Colombo, and Hayne (2012)

As the Helper/Hinderer findings attained iconic status within the cognitive sciences, they inevitably attracted rigorous methodological scrutiny from researchers skeptical of attributing rich socio-moral mental states to preverbal infants. The most formidable empirical challenge arrived in 2012 with a high-profile publication in PLoS ONE by Damian Scarf, Koki Imuta, Michael Colombo, and Harlene Hayne, titled “Golden rule or Perkins rule? Simply attractive: An alternative interpretation of Hamlin et al.’s findings.”

Scarf and his colleagues identified what they argued was a fatal, overlooked low-level perceptual confound in the original 2007 Hill Climber visual stimuli: the “bounce.” In Hamlin et al.’s original video animations, upon successfully reaching the summit plateau with the assistance of the Helper, the Climber executed a noticeable, rhythmic bouncing motion to signify celebration. In the Hinderer condition, following the collision, the Climber tumbled straight to the base and remained completely still, executing no bounce. Scarf et al. hypothesized that this rhythmic, dynamic bouncing motion acted as an unconditioned positive reinforcer. Through basic associative learning, the infant observer might simply pair the sensory pleasure of the bouncy movement with whatever character was present on the stage at that precise moment—which happened to be the Helper.

To test this hypothesis, Scarf et al. replicated the Hill Climber apparatus and experimentally manipulated the presence and placement of the bounce across several conditions:

• They removed the bounce entirely from the Helper condition, having the Climber remain static at the summit. Under these conditions, the infants’ preference for the Helper completely evaporated.
• Most damningly, they added the celebratory bounce to the Hinderer condition—so that after the Hinderer knocked the Climber down, the Climber bounced happily at the bottom of the hill. Under this inverted condition, the infants reversed their choices and overwhelmingly selected the Hinderer.

Scarf and his team concluded that infants were not engaging in rich moral evaluation or mental state attribution at all; they were merely exhibiting low-level perceptual orienting toward whichever geometric shape happened to be paired with a visually stimulating, bouncy animation.

9.2 Hamlin’s Methodological Rebuttals and Clarifications

The publication of Scarf et al.’s critique ignited an intense, highly publicized academic dispute. In 2015, Kiley Hamlin published a comprehensive, exhaustive methodological rebuttal in PLoS ONE, titled “The case for social evaluation in preverbal infants: Gazing toward one’s goal or looking for one’s friends? A response to Scarf et al.” Hamlin demonstrated that in their effort to manipulate the bounce, Scarf and his colleagues had inadvertently broken the fundamental narrative logic of the animated events, stripping them of the very agency cues that make social evaluation possible.

Hamlin pointed out that in the original 2007 design, the Climber’s bounce at the summit was specifically animated as a directed, teleological reaction—the Climber oscillated while facing the summit, oriented toward its achieved goal. In Scarf et al.’s modified videos, characters bounced mechanically while facing away from goals or during inappropriate narrative beats, rendering the display an incomprehensible sequence of disjointed kinetic collisions rather than a coherent social drama. Furthermore, Hamlin noted that Scarf et al.’s critique completely failed to account for the numerous other paradigms—such as the Box-Opening and Ball-Dropping experiments—wherein no bouncing of any kind occurred, yet infants displayed identical, robust preferences for prosocial over antisocial actors.

To definitively settle the empirical question, Hamlin conducted a series of meticulously controlled new experiments using the Hill Climber apparatus. She completely eliminated the celebratory bounce from all conditions—the Climber arrived at the summit and remained perfectly static. The manual reaching results remained identical: even without a single bounce, infants continued to systematically reach for the Helper over the Hinderer at rates exceeding 80%. Hamlin successfully proved that while low-level sensory anomalies can disrupt infant attention if animated clumsily, the underlying phenomenon of infant socio-moral evaluation remained rock-solid when proper experimental cues of intentionality and agency were preserved.

9.3 Large-Scale Collaborative Replications and the ManyBabies Project

In the wake of the broader “Replication Crisis” that swept psychology during the 2010s, developmental psychology initiated massive, collaborative open-science initiatives to establish the true effect sizes and boundary conditions of iconic developmental phenomena. The Helper/Hinderer paradigm became a central focus of these large-scale empirical reassessments, most notably under the auspices of the ManyBabies Consortium—an international network of dozens of independent laboratories working collaboratively to replicate classic findings using massive, pre-registered sample sizes and standardized protocols.

The results of multi-laboratory replication efforts surrounding the Helper/Hinderer paradigm have revealed a nuanced and complex empirical landscape. While several independent laboratories have successfully replicated the manual reaching preference with high fidelity, others have reported smaller effect sizes or non-significant outcomes, particularly when subtle deviations in apparatus construction, character kinematics, or lighting were introduced. Meta-analyses conducted by developmental researchers have indicated that the paradigm’s reliability is extraordinarily sensitive to minute methodological details: the naturalness of the manual puppetry, the duration of pauses, the exact visual prominence of the googly eyes, and the strict blinding protocols of the testing experimenters.

The contemporary scientific consensus recognizes that while the core finding—that infants possess an unlearned capacity to parse social interactions and prefer helpful agents—is fundamentally real, it is not an indestructible, automatic reflex that can be elicited under sloppy experimental conditions. Rather, infant social evaluation represents an ecologically sensitive cognitive process that depends entirely upon the infant successfully perceiving the entities as authentic social agents pursuing genuine, unthwarted goals. When these delicate visual and intentional cues are preserved, the effect reliably emerges; when they are distorted by crude digital animations or mechanical errors, the fragile window into the infant’s socio-moral mind closes.

10. Philosophical and Evolutionary Implications: Paul Bloom’s ‘Just Babies’

10.1 The Thesis of an Innate ‘Moral Sense’

The philosophical synthesis of this empirical research program was articulated by Paul Bloom in his seminal 2013 book, Just Babies: The Origins of Good and Evil. Bloom argued that the infant experiments fundamentally vindicate the Enlightenment tradition of moral sentimentalism, championed by Scottish philosophers Adam Smith and David Hume. In The Theory of Moral Sentiments (1759), Smith posited that human morality is not grounded in abstract, detached mathematical logic, but in innate “passions” and sentiments—an intrinsic capacity for sympathy that makes the suffering of others painful to us, and the benevolence of others a source of organic pleasure. Hume famously declared that “reason is, and ought only to be the slave of the passions.”

Bloom demonstrated that modern developmental science provides the empirical bedrock for Hume and Smith’s philosophical intuition. The infant mind comes biologically pre-equipped with an evolutionary “moral sense”—a suite of intuitive emotional and cognitive adaptations that automatically categorize the social world into normative categories: good vs. bad, fair vs. unfair, kind vs. cruel. This innate core did not evolve to solve abstract ethical dilemmas; it evolved as an indispensable survival kit for early hominins navigating obligate social living. Without an immediate, unlearned mechanism to evaluate the trustworthiness, cooperativeness, and hostility of conspecifics, complex human sociality could never have gotten off the ground.

Crucially, Bloom’s framework carefully delineates between this innate “moral core” and the massive, culturally constructed edifices of mature human morality. Our biology provides the fundamental building blocks: basic empathy, an aversion to suffering, an appreciation for reciprocity, and a rudimentary desire for retributive justice. However, these biological foundations are crude, incomplete, and highly plastic. Civilization, culture, literature, philosophical discourse, and legal codes act as cognitive scaffolding that refines, broadens, and reshapes this biological clay into the expansive human rights frameworks of the modern world.

10.2 The Inherent Limitations of Preverbal Morality

A central, cautionary theme of Bloom’s thesis is that our innate moral sense is profoundly flawed and radically insufficient for navigating the complexities of modern, pluralistic societies. If we rely solely on our unvarnished, innate moral intuitions, we are doomed to ethical catastrophe. The very same infant experiments that demonstrate early benevolence simultaneously illuminate our terrifying biological vulnerabilities: visceral parochialism, immediate tribal categorization, profound out-group hostility, and a complete blindness to systemic, statistical, or long-term consequences.

Infant morality is inherently narrow, spotlighting immediate, palpable, face-to-face interactions while remaining utterly oblivious to the plight of strangers who reside beyond our immediate perceptual field. In his subsequent philosophical critiques, including Against Empathy: The Case for Rational Compassion (2016), Bloom argued that raw empathy—the very emotional engine operating in preverbal infants—is a biased, myopic, and dangerous guide for contemporary socio-political ethics. Empathy acts like a emotional spotlight: it hyper-focuses our moral concern on a single, identifiable individual who looks like us, while leaving millions of faceless, dissimilar individuals in the dark.

Therefore, human moral progress cannot be achieved by merely “listening to our inner child” or stripping away culture to reveal a pristine, noble human nature. Rather, genuine moral advancement requires the aggressive application of cold, deliberate, system-2 rational reasoning. It requires the invention of impartial institutions, the rule of law, empirical science, and universal human rights declarations—cultural artifacts designed specifically to counteract, suppress, and override the profound parochial biases hardwired into our infant brains.

10.3 Re-evaluating Moral Philosophy Through Cognitive Developmental Science

The findings of Hamlin, Wynn, and Bloom have forced a comprehensive re-evaluation of classical and contemporary moral philosophy. For decades, radical behaviorist models, social constructivist doctrines, and deconstructive postmodern theories asserted that human nature is an arbitrary, infinitely malleable blank slate, and that all ethical values are merely oppressive cultural inventions designed to maintain power hierarchies. The demonstration that three-month-old preverbal infants possess universal, cross-cultural evaluative standards of benevolence, malice, and fairness completely refutes this radical blank-slate dogma. There is an authentic, biologically grounded human nature, and it possesses an unmistakable, intrinsic moral valence.

This empirical revolution carries profound implications for meta-ethics and the famous “is-ought divide” first articulated by David Hume. Hume correctly asserted that one cannot derive a normative prescriptive statement (how humans ought to act) purely from an empirical descriptive statement (how the physical world is). The fact that infants naturally exhibit parochial altruism and delight in the punishment of dissimilar out-groups does not mean that tribalism is morally good, desirable, or ethically justified; it merely means that tribalism is natural.

However, cognitive developmental science provides the indispensable empirical parameters within which normative ethics must operate. If moral philosophers and political theorists construct idealized ethical systems that demand humans completely eradicate all cognitive distinctions between their own children and complete strangers, their theories will inevitably collapse under the weight of biological impossibility. By providing an empirical blueprint of the human mind’s innate moral architecture, developmental science shows us where the biological grain runs. It illuminates the specific innate virtues we can easily nurture, the profound innate biases we must actively guard against, and the precise cultural scaffolding required to build a just and humane society.

11. Neurodevelopmental Correlates and Social Cognitive Mechanisms

11.1 Neural Substrates of Infant Social Evaluation

In the decades following the initial behavioral breakthroughs, developmental cognitive neuroscientists sought to map the precise neural substrates underpinning preverbal social evaluation. Utilizing cutting-edge, non-invasive neuroimaging technologies adapted for pediatric populations—specifically Functional Near-Infrared Spectroscopy (fNIRS) and high-density Event-Related Potential (ERP) electroencephalography (EEG)—researchers began tracking the real-time cortical activations of infants as they observed the Helper/Hinderer dramas.

These neuroimaging studies have revealed that the observation of prosocial and antisocial interactions recruits a sophisticated, distributed neural network long before the infant can speak. When infants observe an agent obstructing, striking, or hindering another character, fNIRS recordings show immediate, pronounced hemodynamic oxygenation spikes within the Superior Temporal Sulcus (STS) and the Medial Prefrontal Cortex (mPFC). The STS is a core neural hub dedicated to biological motion processing and teleological action comprehension, while the mPFC is the primary cortical seat of mentalizing and Theory of Mind in the adult brain. This confirms that even in preverbal infants, observing the Hinderer event is not processed as simple mechanical motion, but immediately recruits the high-level cortical networks dedicated to inferring internal social intentions.

Concurrently, ERP electroencephalography has identified distinct neurophysiological signatures that differentiate moral valences within fractions of a second. When viewing antisocial characters, infants display a marked elevation in the Nc component—an endogenous ERP waveform localized over frontal and central scalp locations that indexes the automatic allocation of attentional resources and emotional arousal to salient or threatening environmental stimuli. Furthermore, when infants subsequently view the Climber approaching the Hinderer (the Violation of Expectation condition), their brains exhibit a pronounced late positive slow wave (PSW), a neural marker indicating profound cognitive surprise and the active updating of a violated mental representation. These neuroimaging correlates provide objective, biological proof that the behavioral selections observed by Hamlin and colleagues are the downstream readouts of complex, rapid cortical computations processing moral valence.

11.2 Mirror Neurons, Action Understanding, and Empathic Resonance

A crucial theoretical bridge between perceptual kinematics and socio-moral evaluation resides within the putative contributions of the mirror neuron system and sensorimotor simulation. Discovered originally in macaque monkeys by Giacomo Rizzolatti and colleagues, mirror neurons discharge both when an individual executes a specific goal-directed motor action (such as grasping a cup) and when the individual merely observes another agent executing that identical motor action. In human infants, electroencephalographic suppression of the sensory-motor Mu rhythm (typically oscillating between 6 to 9 Hz over central motor cortices) serves as an established neurophysiological proxy for mirror-system activation.

Developmental researchers hypothesize that sensorimotor simulation provides the critical foundational engine that allows an infant to interpret the Climber’s physical struggle. When the preverbal infant watches the red circular Climber strain against the 30-degree incline, slip backward, and struggle again, the infant’s own motor cortex fires in empathic resonance. The infant internally simulates the physical effort, the resistance against gravity, and the somatic frustration of a blocked motor intention. This sensorimotor resonance is what transforms an abstract visual animation of a sliding wooden circle into an emotionally charged narrative of human struggle.

This empathic simulation forms the indispensable prerequisite for subsequent moral appraisal. To evaluate the Helper as “good,” the infant must first feel that the Climber’s goal was desirable; to evaluate the Hinderer as “bad,” the infant must feel the somatic shock of the Climber’s progress being violently truncated. However, cognitive scientists emphasize that while mirror-neuron-mediated affective resonance provides the raw, visceral fuel for empathy, it does not by itself constitute moral judgment. Moral judgment requires a secondary, higher-order cognitive computation: the infant must uncouple their own somatic resonance, attribute distinct mental states to the interacting parties, evaluate the relational equity of the act, and assign a lasting normative valence to the respective agents.

11.3 The Ontogenetic Pathway Toward Theory of Mind (ToM)

The findings of the Helper/Hinderer experiments have fundamentally reorganized the developmental timeline mapping the ontogenetic emergence of Theory of Mind (ToM). Under the classical cognitive developmental paradigm established by Heinz Wimmer and Josef Perner in 1983, Theory of Mind was conceptualized as a monolithic cognitive milestone that suddenly materialized around age four, indexed by a child’s ability to pass the verbal “False-Belief Task” (such as the famous Maxi or Sally-Anne tasks). Prior to age four, children were assumed to be conceptually incapable of understanding that another mind could hold internal representations that diverge from physical reality.

The Yale infant research, alongside parallel breakthroughs by Kristine Onishi and Renée Baillargeon (2005), dismantled this late-emergence orthodoxy, establishing that Theory of Mind develops along a continuous, multi-tiered ontogenetic trajectory:

• 3 to 6 Months: Rudimentary intentionality parsing, tracking of physical and teleological goals, and basic socio-moral valence assignment (the Helper/Hinderer core).
• 6 to 10 Months: Third-party relational expectations (expecting victims to avoid aggressors) and the calculation of desert and retributive justice.
• 10 to 14 Months: Robust separation of internal intentions from physical outcomes (privileging failed helpers over failed hinderers).
• 14 to 18 Months: Implicit false-belief tracking through anticipatory looking and Violation-of-Expectation paradigms.
• 3 to 5 Years: Explicit, linguistically scaffolded, metacognitive Theory of Mind assessed through formal verbal false-belief narratives.

Far from being an isolated novelty, the preverbal moral evaluations revealed by Hamlin, Wynn, and Bloom represent the foundational ground floor of the entire Theory of Mind architecture. Furthermore, this developmental continuum holds profound clinical significance for the early diagnostic screening of atypical neurodevelopmental trajectories, including Autism Spectrum Disorder (ASD). Toddlers and infants who subsequently receive clinical diagnoses of ASD frequently exhibit specific, early-emerging impairments within this socio-evaluative network—failing to demonstrate the standard preferential gaze toward prosocial agents, showing attenuated Mu-suppression during social animations, and struggling to compute the teleological goals of others. The Helper/Hinderer paradigm has thus emerged as a potent diagnostic prototype for identifying social-cognitive vulnerabilities years before clinical verbal assessments can be administered.

12. The Enduring Legacy, Cross-Species Comparisons, and Future Horizons

12.1 Comparative Cognition: Social Evaluation in Non-Human Primates

To fully map the phylogenetic antiquity of the human moral sense, cognitive scientists and primatologists rapidly translated the Helper/Hinderer experimental architecture to comparative cognition, testing non-human primates across diverse evolutionary lineages. If the capacity to evaluate third-party interactions and valorize helpers evolved as an ancient primate adaptation for coalitional living, our closest living evolutionary cousins—chimpanzees (Pan troglodytes), bonobos (Pan paniscus), and capuchin monkeys (Cebus apella)—should display homologous behavioral choices.

The comparative findings have revealed fascinating phylogenetic continuities and striking, uniquely human divergences. In a landmark series of experiments conducted by Felix Warneken and Michael Tomasello at the Max Planck Institute for Evolutionary Anthropology, chimpanzees demonstrated a robust capacity for direct, instrumental helping. When a familiar human caregiver struggled to reach a dropped object, chimpanzees spontaneously retrieved and handed the object to the human without any expectation of a food reward. Furthermore, in adapted puppet experiments conducted by Frans de Waal and Sarah Brosnan, capuchin monkeys exhibited intense, visceral rejections of inequity, refusing to accept lower-value cucumber slices when a partner was showered with high-value grapes for the identical physical task.

However, when non-human primates were tested in the exact third-party Helper/Hinderer paradigms utilized by Hamlin and colleagues—where the subject merely observes two unfamiliar agents interact with an unrelated third party—a critical evolutionary boundary emerged. While capuchin monkeys and chimpanzees successfully avoid hostile agents who directly cheat or harm them (first-party interactions), their evaluation of purely third-party social dramas is far more fragile and context-dependent. Non-human primates rarely show the robust, spontaneous preference for third-party altruists that is so effortless for human six-month-olds, and they virtually never exhibit spontaneous third-party retributive punishment against hinderers who have harmed an unrelated stranger.

Intriguingly, domestic canines (Canis lupus familiaris) have shown remarkable success in adapted third-party Helper/Hinderer paradigms, consistently refusing food offered by an experimenter who actively hindered their human owner in opening a container. This suggests that the evolutionary capacity for third-party social evaluation can emerge either through high-order hominid encephalization or through the profound convergent selective pressures of artificial domestication and co-evolution with human cooperative niches.

12.2 Cross-Cultural Generalizability of Early Socio-Moral Judgments

A perennial, devastating vulnerability of modern behavioral science is its historical reliance upon hyper-specific subject pools colloquially termed “WEIRD”—Western, Educated, Industrialized, Rich, and Democratic. Joseph Henrich, Steven Heine, and Ara Norenzayan famously demonstrated that WEIRD subjects represent profound statistical outliers on the global human spectrum, cautioning developmental psychologists against asserting universal human traits based entirely on children tested in affluent North American university laboratories.

To confront this challenge, developmental researchers systematically exported the Helper/Hinderer experimental protocols to diverse, non-WEIRD populations across the globe. Cross-cultural research teams conducted controlled testing of the paradigm among infants and young children in traditional, rural horticulturalist villages in Vanuatu, indigenous foraging communities in the Amazon basin, small-scale pastoralist settlements in Sub-Saharan Africa, and dense, non-Western megacities across East Asia. The primary empirical finding from this vast global deployment was unequivocal: the basic manual reaching preference for the Helper over the Hinderer is a universal human trait. Across diverse geographies, subsistence strategies, child-rearing traditions, and religious systems, healthy human infants at six to ten months of age systematically reach for the cooperative agent over the obstructive disruptor.

Where cultural divergences do begin to manifest, however, is during the subsequent preschool and childhood years, particularly regarding the precise threshold and implementation of retributive punishment. In highly individualized societies with strong, centralized formal judicial institutions, children develop a heavy reliance on third-party rule enforcement and institutional sanctioning. In small-scale, interdependent traditional societies where coalitional cohesion and relational repair are paramount to daily survival, children exhibit a vastly greater emphasis on restorative justice, restitution, and rapid reconciliation, actively working to reintegrate the hinderer rather than merely inflicting retributive isolation. Culture, therefore, does not create our initial preference for goodness; culture acts as a master tuner, adjusting the volume, scope, and punitive parameters of an innate moral instrument universal to our species.

12.3 Contemporary Trajectories in Developmental Moral Psychology

As the Helper/Hinderer paradigm approaches two decades of scientific influence, the vanguard of developmental moral psychology is expanding into radical new empirical territories. The most technically sophisticated of these contemporary trajectories is the integration of Artificial Intelligence and Computational Cognitive Modeling. Researchers at institutions like MIT and Stanford are deploying Bayesian computational models and inverse reinforcement learning algorithms to reverse-engineer infant social choice architectures. By modeling the infant mind as an “intuitive utility calculus,” computational cognitive scientists can mathematically predict with astounding accuracy how infants trade off an agent’s physical costs against its social rewards, formalizing the precise algorithmic mechanics through which a baby calculates intentionality, benevolence, and malice.

Concurrently, extensive longitudinal cohort studies are tracking the predictive validity of early infant socio-evaluative metrics. Researchers are following infants who were tested in the Helper/Hinderer and Box-Opening paradigms at six and ten months old across their childhood lifespans, assessing whether the strength of an infant’s early prosocial preference or the magnitude of their negativity bias correlates with subsequent socio-emotional outcomes: childhood peer empathy, prosocial sharing behaviors, executive functioning, callous-unemotional traits, and early signs of conduct disorders or sociopathy. Preliminary longitudinal data indicate that infants who exhibit robust, early social-evaluative competence demonstrate significantly higher empathy scores, superior Theory of Mind mastery, and more collaborative peer interactions during their elementary school years.

Finally, the empirical discoveries of Hamlin, Wynn, and Bloom have permanently revolutionized the philosophy and pedagogical design of Early Childhood Education. The obsolete constructivist and behaviorist assumptions that treated toddlers as amoral, wild animals requiring rigid external behavioral conditioning are being systematically retired. In their place, modern early childhood curricula are designed around the recognition that infants and toddlers are natural moral philosophers who arrive in the classroom pre-equipped with an innate sense of fairness, empathy, and justice. Educators are moving away from authoritarian obedience models toward pedagogical frameworks that actively respect, exercise, and scaffold the child’s innate moral core, helping young humans critically navigate their hardwired tribal vulnerabilities while expanding their core prosocial capacities to embrace the full, diverse human family.

Conclusion

The landmark 2007 publication of the Helper/Hinderer infant experiment by Kiley Hamlin, Karen Wynn, and Paul Bloom stands as one of the most transformative empirical watersheds in the history of the cognitive and developmental sciences. By engineering a non-verbal kinetic methodology capable of probing the deep conceptual architecture of preverbal infants, the Yale research team successfully shattered the century-old consensus of the infant as an amoral, passive tabula rasa. Their work provided undeniable empirical proof that the human infant enters the world endowed with an evolutionarily ancient, biologically hardwired moral core: an innate capacity to discern goals, attribute intentionality, parse social valence, valorize benevolent cooperation, avoid malevolent disruptors, and execute sophisticated calculations of retributive justice long before the acquisition of language or formal cultural pedagogy.

Yet, as subsequent decades of rigorous research have demonstrated, this innate moral architecture is neither simplistic nor universally benevolent. The infant mind’s moral compass is deeply nuanced, characterized by a powerful evolutionary negativity bias that privileges threat detection over positive attraction, an intuitive Theory of Mind that values noble intentions over accidental physical outcomes, and a dark, tragic underbelly of parochial tribalism that readily weaponizes aggression against those who are categorized as different. The profound enduring legacy of Hamlin, Wynn, and Bloom’s research program is not that it painted a naive, romantic portrait of neonatal perfection, but that it revealed the authentic, unvarnished biological foundation of the human condition—a complex, dual-natured foundation upon which the fragile, ongoing triumph of human cultural morality must forever be constructed.

References

• Baillargeon, R. (1987). Object permanence in 3½- and 4½-month-old infants. Developmental Psychology, 23(5), 655–664. https://doi.org/10.1037/0012-1649.23.5.655
• Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger than good. Review of General Psychology, 5(4), 323–370. https://doi.org/10.1037/1089-2680.5.4.323
• Bloom, P. (2004). Descartes’ baby: How the science of child development explains what makes us human. Basic Books.
• Bloom, P. (2013). Just babies: The origins of good and evil. Crown Publishers.
• Bloom, P. (2016). Against empathy: The case for rational compassion. Ecco/HarperCollins.
• Chomsky, N. (1965). Aspects of the theory of syntax. MIT Press.
• Hamlin, J. K. (2013). Failed attempts to help and hinder: Preverbal infants prefer the intentional agent. Proceedings of the National Academy of Sciences, 110(18), 7523–7528. https://doi.org/10.1073/pnas.1300099110
• Hamlin, J. K. (2015). The case for social evaluation in preverbal infants: Gazing toward one’s goal or looking for one’s friends? A response to Scarf et al. PLoS ONE, 10(3), e0118827. https://doi.org/10.1371/journal.pone.0118827
• Hamlin, J. K., Mahajan, N., Liberman, Z., & Wynn, K. (2013). Not like me = bad: Infants prefer those who harm dissimilar others. Psychological Science, 24(4), 589–594. https://doi.org/10.1177/0956797612457785
• Hamlin, J. K., & Wynn, K. (2011). Young infants prefer prosocial to antisocial others. Cognitive Development, 26(1), 30–39. https://doi.org/10.1016/j.cogdev.2010.09.001
• Hamlin, J. K., Wynn, K., & Bloom, P. (2007). Social evaluation by preverbal infants. Nature, 450(7169), 557–559. https://doi.org/10.1038/nature06288
• Hamlin, J. K., Wynn, K., & Bloom, P. (2010). Three-month-olds show a negativity bias in their social evaluations. Developmental Science, 13(6), 923–929. https://doi.org/10.1111/j.1467-7687.2010.00951.x
• Hamlin, J. K., Wynn, K., Bloom, P., & Mahajan, N. (2011). How infants analyze the acts of others: The roots of retributive justice. Proceedings of the National Academy of Sciences, 108(50), 19931–19936. https://doi.org/10.1073/pnas.1112099108
• Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2-3), 61–83. https://doi.org/10.1017/S0140525X0999152X
• Hume, D. (1739). A treatise of human nature. John Noon.
• Kohlberg, L. (1969). Stage and sequence: The cognitive-developmental approach to socialization. In D. A. Goslin (Ed.), Handbook of socialization theory and research (pp. 347–480). Rand McNally.
• Mahajan, N., & Wynn, K. (2012). Origins of “us” versus “them”: Preverbal infants prefer similar others. Cognition, 124(2), 227–233. https://doi.org/10.1016/j.cognition.2012.05.003
• Onishi, K. H., & Baillargeon, R. (2005). Do 15-month-old infants understand false beliefs? Science, 308(5719), 255–258. https://doi.org/10.1126/science.1107621
• Piaget, J. (1932). The moral judgment of the child. Kegan Paul, Trench, Trubner & Co.
• Premack, D. (1990). The infant’s theory of self-propelled objects. Cognition, 36(1), 1–16. https://doi.org/10.1016/0010-0277(90)90013-X
• Rizzolatti, G., Fadiga, L., Gallese, V., & Fogassi, L. (1996). Premotor cortex and the recognition of motor actions. Cognitive Brain Research, 3(2), 131–141. https://doi.org/10.1016/0926-6410(95)00038-0
• Rozin, P., & Royzman, E. B. (2001). Negativity bias, negativity dominance, and contagion. Personality and Social Psychology Review, 5(4), 296–320. https://doi.org/10.1207/S15327957PSPR0504_2
• Scarf, D., Imuta, K., Colombo, M., & Hayne, H. (2012). Golden rule or Perkins rule? Simply attractive: An alternative interpretation of Hamlin et al.’s findings. PLoS ONE, 7(10), e46554. https://doi.org/10.1371/journal.pone.0046554
• Smith, A. (1759). The theory of moral sentiments. A. Millar.
• Spelke, E. S., & Kinzler, K. D. (2007). Core knowledge. Developmental Science, 10(1), 89–96. https://doi.org/10.1111/j.1467-7687.2007.00569.x
• Tajfel, H. (1970). Experiments in intergroup discrimination. Scientific American, 223(5), 96–102. https://doi.org/10.1038/scientificamerican1170-96
• Trivers, R. L. (1971). The evolution of reciprocal altruism. The Quarterly Review of Biology, 46(1), 35–57. https://doi.org/10.1086/406755
• Warneken, F., & Tomasello, M. (2006). Altruistic helping in human infants and young chimpanzees. Science, 311(5765), 1301–1303. https://doi.org/10.1126/science.1121448
• 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
• Wynn, K. (1992). Addition and subtraction by human infants. Nature, 358(6389), 749–750. https://doi.org/10.1038/358749a0