Humans are both, but the weight of evidence from developmental psychology, neuroscience, and evolutionary biology tilts toward cooperation as the deeper default. Infants help strangers before anyone teaches them to, people become more generous when forced to act on gut instinct, and the brain’s reward circuitry lights up during cooperative acts much the way it does for food or money. Selfishness is real and well-documented too, but it tends to emerge under specific conditions rather than as the baseline setting. The honest answer is less “one or the other” and more a story about which circumstances pull us in which direction.
Babies Help Before Anyone Teaches Them
Some of the most compelling evidence for a cooperative default comes from watching very young children. Infants as young as 14 to 18 months old spontaneously help adults reach out-of-grasp objects, open cabinets, and complete tasks they appear to be struggling with. They do this without being prompted, without being rewarded, and apparently without calculating whether they will get something back later. In fact, offering external rewards for helping actually weakens the tendency, suggesting the behavior is internally motivated rather than trained.1PubMed. The roots of human altruism A separate longitudinal study found that from around their first birthday, most infants were already helping at home in everyday situations.2PubMed Central. The Developing Social Context of Infant Helping in Two U.S. Samples
This early helping behavior also has measurable neurological precursors. Research tracking infants’ attention to emotional faces found that the way a seven-month-old processes fearful expressions predicts how much that child will help others at 14 months. Babies who engaged more deeply with fearful faces went on to show greater helping behavior, and these neural attention patterns accounted for a meaningful share of the variation in later altruism.3PLOS Biology. The neurodevelopmental precursors of altruistic behavior in infancy The takeaway is that prosocial tendencies are not purely taught. They appear to be wired into the developmental program early, though socialization powerfully shapes how those tendencies unfold.
Your First Instinct Is to Cooperate
If selfishness were the natural default, you would expect people to become more selfish when they act quickly and more cooperative only after careful deliberation. The opposite tends to happen. Experiments using time pressure to force fast decisions find that people become more cooperative when they rely on intuition and more calculating when given time to think. One series of studies confirmed this causal link: time pressure increased cooperative choices, supporting the idea that cooperative behavior reflects an internalized social heuristic rather than a grudging override of selfishness.4Nature Communications. Social heuristics shape intuitive cooperation
This pattern holds even in situations involving inequality. When participants in economic games faced asymmetric payoffs, intuitive responding still promoted giving. Deliberation, by contrast, tended to reduce it.5PubMed Central. Spontaneous Giving under Structural Inequality: Intuition Promotes Cooperation in Asymmetric Social Dilemmas None of this means people never think their way to generosity. But the consistent finding that snap judgments lean cooperative suggests that for most people, the trained habit of social life is to share, and selfishness takes effort.
Cooperation Feels Good for a Reason
The brain does not treat cooperation as a sacrifice it grudgingly makes. Neuroimaging research shows that cooperative and altruistic behavior consistently activates reward-related areas like the striatum, the same region that responds to food, money, and other pleasurable outcomes. Cooperative acts also engage prefrontal regions involved in weighing costs and benefits and processing emotions, consistent with the idea that the brain is resolving a genuine conflict between self-interest and other-regarding impulses rather than simply defaulting to one side.6Trends in Cognitive Sciences. Neural economics of cooperation and altruism
On the empathy side, seeing other people’s emotions activates some of the same brain regions involved in experiencing those emotions yourself. This is not a vague metaphor about “feeling someone’s pain.” Research on emotion simulation shows that observing someone else in distress recruits a mosaic of affective, motor, and sensory components in the observer’s brain, and these shared representations feed into empathy, social learning, and the ability to understand other minds.7PubMed Central. Evidence for mirror systems in emotions A network including the inferior frontal gyrus and inferior parietal lobule appears to be necessary for recognizing emotions and for emotional contagion to occur.8PubMed. The neural bases for empathy The architecture is there, built into the brain. Whether you act on it depends on context.
What Our Closest Relatives Tell Us
Humans are not the only primates who cooperate, but we do it on a scale that dwarfs our relatives. Comparing our two closest living kin is instructive. Bonobos, who are notably more tolerant than chimpanzees during feeding, outperformed chimpanzees on cooperative food-retrieval tasks when the reward was easy to monopolize. When the food was hard to hog, the two species performed equally well. The difference came down to temperament: bonobos’ higher social tolerance made cooperation sustainable in situations where a chimpanzee would have fought over the prize.9PubMed. Tolerance allows bonobos to outperform chimpanzees on a cooperative task
Bonobos also spontaneously shared tools with unrelated individuals even when helping did not benefit them directly. All female-female bonobo pairs in one study transferred tools without prompting, while chimpanzees almost never did so outside a mother-daughter pair.10PubMed Central. Targeted helping and cooperation in zoo-living chimpanzees and bonobos The implication is that the emotional wiring for prosocial behavior predates the human lineage. Selection on temperament, particularly on tolerance and reduced reactive aggression, can dramatically expand a species’ cooperative range.
Did Humans Domesticate Themselves Into Cooperators?
A provocative hypothesis holds that humans underwent a process analogous to domestication, not by an external breeder but through social selection pressures within our own groups. The self-domestication hypothesis proposes that over the past 300,000 years, Homo sapiens experienced unusually intense selection against reactive aggression. Individuals who were too violently impulsive may have been ostracized, killed, or otherwise excluded from reproduction by the group, gradually favoring calmer, more socially tolerant temperaments.11Frontiers in Psychology. Hypotheses for the Evolution of Reduced Reactive Aggression in the Context of Human Self-Domestication This would help explain the egalitarian male hierarchies seen in mobile hunter-gatherer groups, as well as anatomical parallels between humans and domesticated animals like reduced brow ridges and shorter faces.
The idea has gained traction because it connects behavioral and physical observations under one evolutionary mechanism. If reduced reactive aggression was selected for, it would have created the social conditions for cooperation to flourish: more tolerance during resource sharing, lower costs to group living, and expanded social networks.12Language and Linguistics Compass. Human self‐domestication and the evolution of prosody In this view, cooperativeness is not just a cultural overlay; it is partly a product of biological selection on temperament that has been running for hundreds of thousands of years.
Why Cooperative Child-Rearing May Have Been the Key
Another evolutionary account focuses on a distinctive feature of human life: we raise our children cooperatively. Unlike most great apes, where mothers bear nearly the entire burden of infant care, human societies rely heavily on allomaternal care, meaning that fathers, grandparents, siblings, and unrelated group members all pitch in. A comparative study across primate species found that the extent of allomaternal care was by far the best predictor of proactive prosociality, the willingness to help others voluntarily without being asked.13Nature Communications. The evolutionary origin of human hyper-cooperation
The logic is intuitive. When raising offspring requires a village, natural selection favors individuals who are willing to contribute to others’ welfare, who can read social cues, and who can maintain cooperative relationships across a broad network. Species that rely on cooperative breeding end up with a broader prosocial disposition than species where mothers go it alone. If our hominin ancestors adopted this childrearing strategy, it provides a parsimonious explanation for why humans cooperate on a scale that no other primate approaches.
Genetics Set the Range, Not the Outcome
Prosocial behavior has a heritable component. A twin study of seven-year-olds found that a broad prosociality factor, capturing behaviors like helping, sharing, and comforting, was roughly 69% heritable, with the remaining variance driven by individual-level environmental factors and measurement error.14Frontiers in Psychology. The prosocial personality and its facets: genetic and environmental architecture of mother-reported behavior of 7-year-old twins Empathy itself shows a split: a meta-analysis of twin studies estimated that emotional empathy, the gut feeling of sharing someone’s distress, is about 48% heritable, while cognitive empathy, the ability to understand what someone else is thinking, is about 27% heritable. For cognitive empathy, the shared family environment also mattered, explaining roughly 12% of the variation.15PubMed. The genetic and environmental origins of emotional and cognitive empathy: Review and meta-analyses of twin studies
These numbers do not mean that your generosity is “set” at birth. Heritability describes variation within a population, not the rigidity of any individual’s behavior. What the data suggest is that people genuinely differ in their baseline prosocial disposition, partly for genetic reasons, and that how those dispositions are expressed depends on the social environment. This is consistent with the broader theme: cooperation is built into our biology, but the degree and direction of it are shaped by experience and context.
The Dark Side of Bonding
The same neurochemistry that binds groups together can also divide them. Oxytocin, the hormone often called the “love molecule,” does promote trust and cooperation, but primarily toward people you already consider part of your group. Experimental research has shown that oxytocin drives a “tend and defend” pattern: it promotes in-group trust and cooperation while simultaneously increasing defensive aggression toward competing out-groups.16PubMed. The neuropeptide oxytocin regulates parochial altruism in intergroup conflict among humans A separate study found that oxytocin creates intergroup bias by motivating in-group favoritism and, to a lesser degree, out-group derogation.17PubMed Central. Oxytocin promotes human ethnocentrism
This pattern maps onto a broader evolutionary concept called parochial altruism: the idea that in-group cooperativeness and out-group hostility may have evolved together as a package deal. Modeling work suggests that under conditions resembling the late Pleistocene, neither pure altruism nor pure parochialism could have survived on its own, but in combination, they reinforced each other. Groups with altruists who preferentially helped insiders and were hostile to outsiders won conflicts with groups that lacked this combination.18PubMed. The coevolution of parochial altruism and war A review of this theory notes that the two traits, cooperativeness toward the in-group and aggressiveness toward out-groups, may have coevolved through both genetic and cultural pathways.19PubMed Central. The evolutionary interplay of intergroup conflict and altruism in humans: a review of parochial altruism theory and prospects for its extension
This is the uncomfortable part of the cooperation story. Our generosity is not unconditional. It evolved in a context of group competition, and it still carries that fingerprint. The capacity for extraordinary within-group solidarity often coexists with suspicion or outright hostility toward outsiders. Any account of human nature that emphasizes only one side of this is telling half the story.
How Culture Amplifies or Reshapes the Baseline
If cooperation were purely instinctive, you would expect it to look the same everywhere. It does not. A study across six diverse societies found that when sharing was free, meaning it cost nothing, prosocial behavior increased steadily as children matured, with little variation between cultures. But when sharing was costly, a different pattern emerged: rates of prosociality dropped during middle childhood across all six societies, and then diverged as children tracked toward adult norms specific to their own culture.20PubMed Central. Ontogeny of prosocial behavior across diverse societies The universal part is the willingness to help when it is cheap. The culturally specific part is what happens when helping actually costs you something.
Fairness norms show a similar pattern of universal trajectory plus cultural tuning. Children across seven societies developed an aversion to getting less than a peer by middle childhood, and this was consistent everywhere. But the willingness to reject getting more than a peer, a subtler form of fairness, emerged only in some populations and only later in development.21Nature. The ontogeny of fairness in seven societies A separate study of over 2,000 children in 13 countries found that whether a culture was more individualist or collectivist predicted when and how strongly children favored equity over equality, particularly in contexts involving wealth or merit.22PubMed. The development of children’s preferences for equality and equity across 13 individualistic and collectivist cultures Culture does not create the prosocial impulse from scratch, but it determines how far that impulse extends and under what conditions it applies.
Reputation and Punishment Keep Cooperation Running
Even in a species predisposed to cooperation, the system needs enforcement. Two mechanisms do most of the heavy lifting: reputation and punishment. Indirect reciprocity, the idea that helping someone can earn you a good reputation and make others more willing to help you later, is one of the most studied frameworks for explaining cooperation among strangers. When one person helps another, observers update the helper’s reputation, and that reputation influences future interactions.23PubMed Central. The benefits of being seen to help others: indirect reciprocity and reputation-based partner choice The math behind this system requires social norms that define what counts as a good or bad action, which is essentially a way of studying moral systems through the lens of evolutionary stability.24Philosophical Transactions of the Royal Society B. The complexity of human cooperation under indirect reciprocity
Punishment reinforces what reputation cannot. An experiment with over 1,100 children aged 7 to 11 found that the mere threat of third-party punishment more than doubled cooperation rates, even though children were rarely willing to actually follow through on punishing defectors. The higher cooperation was driven partly by incorrect beliefs about how much punishment others would dish out and partly by conditional cooperation: “I will share if I think my partner will share, and I think punishment will make them share.”25PubMed Central. Third-party punishment increases cooperation in children through (misaligned) expectations and conditional cooperation Cross-cultural research has found that people in larger, more complex societies engage in significantly more third-party punishment than people in small-scale societies, suggesting that as groups scale up, enforcement mechanisms grow to match.26PubMed Central. More ‘altruistic’ punishment in larger societies
What Stress Does to Cooperation
The popular assumption is that stress makes people selfish. The reality is more nuanced. Acute social stress, the kind you feel during a job interview or a public speaking task, actually increases prosocial behavior. Participants who experienced this kind of stress showed substantially more trust, trustworthiness, and sharing than unstressed controls, and the effect was specific: stress did not make people more willing to take nonsocial risks or behave antisocially.27PubMed. The social dimension of stress reactivity: acute stress increases prosocial behavior in humans This is consistent with the “tend and befriend” hypothesis: under threat, reaching out to others is itself a stress-buffering strategy.
But context matters enormously. In rats, moderate stress in a neutral setting increased social bonding and boosted oxytocin signaling. The same stressor paired with predator odor, a signal of extreme danger, eliminated those social benefits entirely.28PubMed Central. Moderate Stress-Induced Social Bonding and Oxytocin Signaling are Disrupted by Predator Odor in Male Rats In humans, psychosocial stress pushed people toward favoring their own group at the expense of outsiders: higher physiological stress responses were associated with decreased in-group generosity and increased out-group hostility.29Scientific Reports. The effects of psychosocial stress on intergroup resource allocation Manageable stress draws people together. Overwhelming or threatening stress can fracture cooperation and amplify tribal divisions.
When the Stakes Get Higher, Fairness Erodes
Economic experiments offer a controlled window into how people balance self-interest against fairness, and the results challenge simple narratives in both directions. In dictator games, where one person splits money with a stranger and the stranger has no say, the typical split is not fifty-fifty, but people consistently give something. When the framing shifts from sharing gains to sharing losses, dictators actually leave more to the other person, suggesting that the cooperative impulse strengthens when both sides face hardship.30Journal of Economic Psychology. Sharing losses in dictator and ultimatum games: A meta-analysis
But there is a limit. As the amount of real money at stake increases, generosity drops. A study examining how stakes affect offers found a significant negative relationship: the higher the stakes, the lower the share people offered. This held true for both dictator games and ultimatum games, where the receiver can reject unfair offers.31PLOS ONE. How Much Is Our Fairness Worth? The Effect of Raising Stakes on Offers by Proposers and Minimum Acceptable Offers in Dictator and Ultimatum Games People are prosocial, but they are not infinitely so. When the cost of being fair gets high enough, self-interest gains weight. This is not a flaw in an otherwise cooperative species; it is the system working as expected, recalculating the cost-benefit balance in real time.
The Hunter-Gatherer Puzzle
A common version of the “mismatch” argument suggests that humans evolved to cooperate in small, tight-knit groups of relatives, and that we struggle in large anonymous societies because our psychology has not caught up. There is a kernel of truth to this, but recent ethnographic work has complicated the picture. Studies of contemporary hunter-gatherers such as the Hadza show that relatedness within bands is actually low, and band composition is remarkably fluid. On average, an individual lives with only about one in five of the same bandmates from the previous year. These groups are not stable clusters of kin. They are constantly reshuffling.
Yet cooperation flourishes anyway. Hunter-gatherer societies commonly practice “demand sharing,” where food brought into camp is claimed and divided by group members as a social norm.32Nature Communications. High mobility explains demand sharing and enforced cooperation in egalitarian hunter-gatherers High mobility may actually promote cooperation: because people can leave groups that exploit them, group members have strong incentives to maintain fair norms. This turns the mismatch narrative partly on its head. Our ancestors were not cooperating because they lived with family. They were cooperating with near-strangers in fluid groups, which means the psychological machinery for broad cooperation may be older and more robust than the simple kin-selection story suggests.
Group Selection Is Back on the Table
For decades, evolutionary biologists dismissed group selection, the idea that natural selection can act on groups rather than just individuals, as theoretically marginal. That consensus has softened. Mathematical models have shown that when groups are numerous enough relative to their size, cooperation can evolve through a multilevel selection process where group-level traits are favored even if they impose costs on individuals.33PubMed Central. Evolution of cooperation by multilevel selection A recent study of wild animal populations found that selection for group-level social structure was just as strong as, and sometimes stronger than, selection on individual social behavior, providing empirical support for what had long been a mostly theoretical framework.34PubMed Central. Multilevel selection on individual and group social behaviour in the wild
This matters for the selfishness-versus-cooperation debate because it means the unit of selection is not exclusively the selfish gene maximizing its own replication. Groups of cooperators can outcompete groups of defectors, and this dynamic creates a selection pressure that sustains prosocial traits even when they are individually costly. The picture that emerges is layered: individual-level selection can favor selfishness within groups, while group-level selection favors cooperation between groups. Human nature sits at the intersection of both forces, which is exactly why the question in this article’s title does not have a one-word answer.
Personality Traits That Push Toward Selfishness
While most people lean cooperative, a meaningful minority display consistently exploitative tendencies. Personality research has identified a cluster of traits known as the Dark Triad: narcissism, Machiavellianism, and psychopathy. These share a common core, sometimes labeled “D,” that captures a general disposition toward self-serving behavior at others’ expense. Research has found that this common dark core is almost perfectly inversely related to the personality dimension of Honesty-Humility, with the two constructs correlating at about –0.95 at the latent level. In practical terms, the same psychological dimension that makes a person humble, fair, and averse to exploiting others is the dimension whose absence predicts manipulative and selfish behavior.
These traits are not randomly distributed. They show up on a spectrum in the general population, not just in clinical cases. And they have both genetic and environmental roots, just as prosociality does. The existence of dark personality traits does not refute the cooperative-default hypothesis; it shows that natural selection has maintained variation in this space. A population of pure cooperators is vulnerable to exploitation, so some degree of selfish strategy persists. The fact that most people cluster on the prosocial end of the continuum, while a smaller proportion occupy the exploitative end, is itself informative about where the equilibrium sits.