Altruism

Abstract

Altruism is a form of social behavior: action that benefits another individual at a cost to the actor. It is the sharpest version of social behavior's organising puzzle, because a naive account of self-interest predicts that costly helping should not exist, yet it is pervasive across insects, other primates, human infants, and anonymous one-shot giving. Cognitive psychology approaches altruism on three levels that do not compete so much as nest: an evolutionary account of why helping can pay in the currency of fitness (kin selection and reciprocity), a psychological account of the proximate motive that moves a person to help (the empathy–altruism debate), and a neural account of how the brain assigns value to another's welfare. A worked example develops the arithmetic of kin selection, reciprocity, and warm-glow giving, and three interactive demonstrations let the reader manipulate each.

Keywords: altruism, prosocial behavior, empathy, kin selection, reciprocity

Few behaviors have attracted as much theoretical attention from as many disciplines as altruism. Biologists ask how natural selection, which rewards reproductive success, could ever favor an organism that spends its own resources to raise another's. Economists ask why people give to strangers and charities when free-riding is available. Psychologists ask whether a person who helps is ever truly moved by concern for the other, or only by a well-disguised concern for themselves. Each field converges on the same observable—a costly act that benefits someone else—and each has contributed a piece of the explanation.

What makes altruism a problem for cognitive psychology in particular is that the competing explanations make different claims about the mind of the helper. An act can be altruistic in its evolutionary logic while the animal performing it feels nothing at all; it can be altruistic in its consequences while the person performing it is motivated entirely by the warm feeling of having given. Separating the behavior from the motive, and the motive from its evolutionary rationale, is where the cognitive science of altruism earns its keep.

What Altruism Is

Altruism, in the broadest usable sense, is behavior that increases another individual's welfare at a cost to one's own. The definition is deliberately behavioral: it says nothing about why the act was performed, only that it helped another and cost the actor. That neutrality matters, because the word is used in three different ways that are easy to conflate. A biologist's altruism is defined in the currency of fitness—a behavior that lowers the actor's reproductive success and raises a recipient's—and can be ascribed to an organism with no inner life at all. A psychologist's altruism is a claim about motivation—helping driven by an ultimate goal of improving another's welfare rather than one's own (Batson & Shaw, 1991). An economist's altruism is a preference—a positive weight placed on another's payoff in one's own utility function (Andreoni, 1990). The three can diverge: an act can be biologically altruistic but psychologically selfish, or psychologically altruistic but biologically self-serving.

In the Medical Subject Headings vocabulary, Altruism is a descriptor filed directly beneath Social Behavior, at the prosocial end of that family alongside cooperative and helping behaviors. The placement is a clean taxonomic one: altruism is a proper subset of social behavior, distinguished by the cost borne for another's benefit. It sits opposite the competitive and hostile behaviors—aggression, stereotyping—that occupy the other pole of the same descriptor, and shades into the neighboring categories of cooperative behavior and helping behavior with which it shares most of its mechanisms.

Altruism is best treated not as a single trait a person has or lacks but as the output of several partly independent systems, each operating on its own logic (Penner et al., 2005). The remainder of this article takes those systems in turn—the evolutionary rationale, the proximate psychological motive, and the neural valuation machinery—and shows how they fit together without reducing to one another.

Altruism and Cognition

The science of altruism is organised by a distinction the biologist Niko Tinbergen made famous: an ultimate explanation says why a behavior was favored over evolutionary time, while a proximate explanation says what mechanism produces it here and now in the individual (Preston & de Waal, 2002). The two are not rivals. Kin selection is an ultimate account; empathy is a proximate one; and a complete explanation of a single helping act needs both. The sections below work up the levels—from the evolutionary rationale, through the proximate motive, to the brain systems that implement it—and connect each to the cognition it requires.

Table 1

Mechanisms That Can Favor Altruism

The evolutionary roots of helping

The evolutionary problem is stark: a gene that disposes its bearer to pay a cost so that another may benefit should, all else equal, be driven out by a gene that lets its bearer reap the benefit without paying. The first and most powerful resolution is kin selection. A gene for altruism can spread if the help is directed at relatives who are likely to carry the same gene. Hamilton's rule states the condition precisely: an altruistic act is favored when rb > c, where r is the genetic relatedness of actor to recipient, b the benefit to the recipient, and c the cost to the actor (Hamilton, 1964). The quantity selection maximises is not the individual's own reproduction but its inclusive fitness—its reproduction plus its weighted effect on relatives'. Kin altruism is cognitively cheap, needing only a cue of relatedness, but it is bounded to kin.

Cooperation among non-relatives needs a different rationale. Reciprocal altruism explains helping between unrelated individuals who interact repeatedly: a cost borne now is repaid by the recipient later, and the arrangement pays whenever the probability of future encounters is high enough to make the expected return exceed the cost (Trivers, 1971). Unlike kin altruism, reciprocity is cognitively demanding—it requires recognising individuals, remembering past interactions, and conditioning behavior on that memory. These two mechanisms, together with reputation-based (indirect) reciprocity and the effects of population structure, make up the small set of rules by which cooperation can evolve (Nowak, 2006; Rand & Nowak, 2013). Demo 1 makes Hamilton's rule concrete, letting relatedness, benefit, and cost vary and reading off when helping is favored.

The empathy–altruism debate

The evolutionary accounts explain why a capacity for helping might be favored; they do not settle what motive moves a person in the moment. Here psychology's central question is whether genuinely other-regarding motivation exists at all. Psychological egoism holds that every apparently altruistic act is ultimately self-serving—done to gain praise, to avoid guilt, or to relieve the unpleasant arousal of seeing another suffer. Against it, the empathy–altruism hypothesis holds that empathic concern for a person in need produces motivation with that person's welfare as its ultimate goal, not merely one's own relief (Batson & Shaw, 1991).

The decisive test is experimental. If helping is really egoistic—a way to escape one's own distress—then giving people an easy exit should make them leave rather than help. If helping is genuinely altruistic, an easy exit should not matter, because leaving does not accomplish the goal of relieving the other's need. The empathy-escape experiments cross empathy (low vs high) with escape (easy vs difficult): under low empathy, helping tracks the ease of escape exactly as egoism predicts; under high empathy, helping stays high whether or not escape is easy—the signature of altruistic motivation (Batson & Shaw, 1991). The capacity rests on empathy itself, whose proximate machinery—a perception–action coupling that maps another's state onto one's own—is shared across social mammals and gives the behavior its evolutionary depth (Preston & de Waal, 2002). Demo 2 reconstructs the empathy-escape design and lets the reader read helping rates off the 2×2.

The altruistic brain

If helping has value to the helper, that value must be computed somewhere. Imaging work locates it in the brain's general-purpose valuation system. When people donate to a charity, the mesolimbic reward circuitry—ventral striatum, ventromedial prefrontal cortex—responds much as it does to receiving money for oneself, and this response appears both when giving is voluntary and when the transfer is a mandatory tax, with voluntary giving adding an extra increment of reward: a neural signature of the warm glow of choosing to give (Harbaugh et al., 2007; Moll et al., 2006). The same value machinery that prices food and money is recruited to price another's welfare, letting the brain fold social outcomes into ordinary decision-making (Ruff & Fehr, 2014). A meta-analysis of giving decisions confirms a reliable network for altruistic choice while distinguishing it from merely strategic generosity, where the giver expects a return (Cutler & Campbell-Meiklejohn, 2019).

The flip side of valuing others is enforcing fairness against them. In public-goods settings, cooperation is rescued when players can punish free-riders at a cost to themselves (Fehr & Gächter, 2002), and this altruistic punishment engages the same dorsal striatal reward circuitry, as though delivering the sanction were itself satisfying (de Quervain et al., 2004). Valuation also tracks the moral weight of harm: people represent the value of outcomes that harm others differently from outcomes that harm only themselves, and that representation can be read out of the same neural signals (Crockett et al., 2017). Synthesising the neural, cognitive, and evolutionary strands, human altruism emerges as the product of several interacting systems rather than one dedicated module (Marsh, 2016). Demo 3 illustrates the warm-glow account of giving and its best-known prediction.

Development, learning, and other species

If altruism were purely a product of socialisation, it should appear only after years of moral instruction. It does not. Human toddlers as young as 14 to 18 months help an unrelated adult achieve a goal—retrieving an out-of-reach object, opening a cabinet whose contents the adult's hands are too full to reach—spontaneously and without reward, and young chimpanzees show the same instrumental helping, implying a prosocial tendency with phylogenetic roots rather than a wholly cultural invention (Warneken & Tomasello, 2006). Across the lifespan, helping is also shaped by learning. Recent neurocomputational work shows that people learn which actions benefit others through reinforcement signals partly distinct from those for learning about rewards to themselves, and that individual differences in this prosocial learning track self-reported empathy (Lockwood et al., 2016). Whether a single social brain implements all of this, or whether social cognition reuses domain-general machinery, is an active question about the level at which altruism is organised (Lockwood et al., 2020).

Worked Example

The logic of altruism becomes concrete in three small calculations, one per level, matching the three demonstrations.

Kin selection (Hamilton's rule). An act is favored when rb > c. For full siblings r = 0.5. Suppose the act costs the actor the fitness equivalent of c = 1 offspring and benefits the sibling by b:

- Benefit b = 3: rb = 0.5 × 3 = 1.5 > 1, so altruism toward the sibling is favored. - Benefit b = 1.5: rb = 0.5 × 1.5 = 0.75 < 1, so it is not.

The same generosity that pays toward a sibling (r = 0.5) fails toward a first cousin (r = 0.125) unless the benefit is at least eight times the cost, since the break-even benefit is b\ = c/r = 1/0.125 = 8. This is why kin altruism falls off steeply with genetic distance. Demo 1 plots rb against c* as the three quantities vary.

Reciprocal altruism (the shadow of the future). Between non-kin who may meet again with probability w each round, a reciprocating strategy pays when w > c/b. With c = 1 and b = 3 the threshold is w\* = 1/3 ≈ 0.33:

- Encounters likely, w = 0.6: 0.6 > 0.33, so reciprocal helping is sustainable; the expected number of future rounds is 1/(1 − w) = 2.5, enough shadow of the future to make helping now worthwhile. - Encounters rare, w = 0.2: 0.2 < 0.33, so defection wins and cooperation unravels; the expected future is only 1/(1 − 0.2) = 1.25 rounds.

Warm-glow giving and crowding out. A pure altruist values only the total amount a charity receives; a warm-glow (impure) altruist also values the act of giving itself (Andreoni, 1990). Suppose a donor is currently giving $20 to a charity, and a $10 grant from elsewhere arrives:

- The pure altruist cares only that the charity's total is 20, so she cuts her gift to $10, leaving the total at 20 — the grant fully crowds out private giving, dollar for dollar. - The warm-glow donor still wants the private act of giving, so she cuts her gift only to $15, leaving the total at 25 — the $10 grant crowds out just $5 of private giving, a 50% crowd-out rate.

Incomplete crowding out is the warm-glow model's signature empirical prediction, and its neural counterpart is Harbaugh and colleagues' finding that voluntary giving activates reward circuitry beyond what a mandatory, equivalently sized transfer does (Harbaugh et al., 2007). Demo 3 runs both donor types against a varying external grant and shows the two crowd-out trajectories diverge.

Discussion

Read across its levels, altruism is less a single phenomenon than a convergence. The evolutionary accounts explain why organisms that help can out-reproduce ones that do not, under the specific conditions of relatedness or repeated interaction that make helping pay (Hamilton, 1964; Trivers, 1971; Nowak, 2006). The proximate accounts explain what actually moves a person in the moment—and the weight of the empathy-escape evidence is that at least some helping is genuinely other-regarding, not covert egoism (Batson & Shaw, 1991). The neural accounts explain how the trick is done: by recruiting the brain's general valuation machinery to place a positive weight on another's welfare, so that giving, and even punishing unfairness, can be experienced as rewarding (Harbaugh et al., 2007; de Quervain et al., 2004).

The levels are easy to mistake for rivals, and much of the field's confusion has come from treating them as such—as if showing that giving activates reward circuitry proved that people really give only for the warm glow, or as if an evolutionary rationale showed that empathy is a sham. It does neither. That an altruistic act feels good, or pays in fitness, is exactly what a well-designed motivational system should do; the feeling is the mechanism by which the ultimate rationale gets implemented, not evidence that the motive is a fraud (Preston & de Waal, 2002; Marsh, 2016). The honest summary is that altruism is real at every level: a behavior favored by selection, produced by genuine other-regarding motivation in at least some cases, and computed by circuitry that assigns value to others' outcomes.

Current Directions

Three fronts are active. The first is the computational dissection of prosocial motivation: rather than asking only whether people help, researchers now model the learning signals by which people discover which actions benefit others, finding that prosocial learning draws on valuation and prediction-error signals partly separable from those for self-benefit, and that these signals track trait empathy (Lockwood et al., 2016). The second is a debate over the architecture of social valuation—whether a dedicated social brain underlies altruism or whether social decisions reuse domain-general value and learning machinery, a question now framed in terms of competing implementations and algorithms rather than anatomy alone (Lockwood et al., 2020). The third is the moral dimension of value: work showing that people represent the cost of harming others distinctly from the cost of harming themselves, and that moral transgressions reshape the neural representation of value, is extending altruism research from giving toward the broader question of how the brain prices right and wrong (Crockett et al., 2017), while meta-analysis continues to sharpen which neural signatures are specific to altruistic rather than strategic generosity (Cutler & Campbell-Meiklejohn, 2019).

Common Misconceptions

If an altruistic act benefits the actor, it was not really altruistic.

This conflates levels. A behavior can be psychologically altruistic—motivated by concern for another—while also paying in reputation, fitness, or felt reward. The pay-off is how the motivational system is built, not proof that the motive was selfish (Batson & Shaw, 1991; Preston & de Waal, 2002).

Kin selection means organisms want to help their relatives.

Kin selection is an ultimate explanation about gene frequencies, not a proximate account of motive. The animal need not recognise kinship or intend anything; a simple cue—being raised together, say—can do the work. Hamilton's rule says when such behavior is favored, not how it feels (Hamilton, 1964).

Altruism is a uniquely human, purely cultural achievement.

Spontaneous, unrewarded instrumental helping appears in human toddlers before much socialisation and in young chimpanzees, and the empathic machinery that supports it is shared across social mammals. Culture elaborates altruism; it does not invent it from nothing (Warneken & Tomasello, 2006; Preston & de Waal, 2002).

Government funding of a cause simply adds to private giving.

If donors were pure altruists caring only about the total provided, public funds would crowd out private gifts dollar for dollar. Because many donors also value the act of giving—the warm glow—crowding out is only partial, so public and private giving coexist (Andreoni, 1990; Harbaugh et al., 2007).

Glossary

Altruism.

Behavior that benefits another individual at a cost to the actor; defined behaviorally, so that it says nothing in itself about the motive behind the act.

Altruistic punishment.

Paying a personal cost to sanction a norm violator with no prospect of personal gain; it stabilises cooperation in groups and engages reward circuitry when carried out.

Empathic concern.

An other-oriented emotional response—sympathy, compassion—evoked by and congruent with another's perceived need; the proposed trigger of altruistic motivation in the empathy–altruism hypothesis.

Empathy–altruism hypothesis.

Batson's claim that empathic concern produces motivation whose ultimate goal is the other's welfare, not the helper's own relief; tested by the empathy-escape experiments.

Hamilton's rule.

The condition under which kin-directed altruism is favored by selection: rb > c, where r is relatedness, b the benefit to the recipient, and c the cost to the actor.

Helping behavior.

Any act that assists another in reaching a goal; the broad behavioral category of which altruism is the costly, other-benefiting subset.

Impure altruism.

Giving motivated partly by the private satisfaction of the act itself (the warm glow) rather than purely by the outcome; contrasted with pure altruism, which values only the total benefit delivered.

Inclusive fitness.

An individual's own reproductive success plus its effect on the reproduction of relatives, each weighted by relatedness; the quantity kin selection maximises.

Kin selection.

The evolution of altruism directed at genetic relatives, favored when relatedness times benefit exceeds cost; cognitively cheap but limited to kin.

Prosocial behavior.

Voluntary behavior intended to benefit others—helping, sharing, comforting, donating; the positive pole of social behavior, of which altruism is the costly case.

Prosocial learning.

Learning which actions benefit other people, driven by reinforcement signals partly distinct from those for self-benefit and tracking individual differences in empathy.

Psychological egoism.

The thesis that every action, including apparent altruism, is ultimately aimed at some benefit to the self; the position the empathy–altruism hypothesis was formulated to refute.

Reciprocal altruism.

Trivers's account of cooperation among non-kin, in which a cost borne now is repaid by the recipient later; viable when future encounters are likely enough.

Relatedness (r).

The probability that a gene in the actor is shared by the recipient through common descent; 0.5 for full siblings and parent–offspring, 0.125 for first cousins.

Social behavior.

Action directed at or shaped by other members of one's species; the parent category under which MeSH files altruism, at its prosocial pole.

Ultimate vs proximate explanation.

An ultimate explanation says why a behavior was favored over evolutionary time; a proximate one says what mechanism produces it in the individual. Altruism needs both, at different levels.

Warm-glow giving.

Donating for the private, affective reward of the act of giving itself; Andreoni's model of impure altruism, and the leading explanation of incomplete crowding out.

Key Researchers

C. Daniel Batson

(University of Kansas, Professor Emeritus). Formulated and tested the empathy–altruism hypothesis, designing the empathy-escape experiments that remain the decisive evidence that at least some human helping is genuinely other-regarding rather than covert egoism. Faculty · Wikipedia

Ernst Fehr

(University of Zurich). Behavioral economist whose experiments on altruistic punishment showed that people pay personal costs to sanction free-riders, stabilising cooperation, and helped found the neuroeconomics of social preferences. Faculty · Google Scholar · Wikipedia

W. D. Hamilton

(1936–2000) (University of Oxford). Formalised inclusive fitness and kin selection, giving the genetical theory of altruism its founding equation, Hamilton's rule, and reframing social behavior as the product of gene-level selection. Wikipedia

Patricia L. Lockwood

(University of Birmingham). Decision neuroscientist whose work on the neurocomputational mechanisms of prosocial learning shows that people learn to benefit others through signals partly distinct from self-interested learning, and that these track empathy. ORCID · Faculty

Abigail A. Marsh

(Georgetown University). Social neuroscientist of altruism, empathy, and psychopathy whose synthesis of the neural, cognitive, and evolutionary foundations of human altruism frames it as the output of several interacting systems rather than one module. Faculty · Google Scholar · Wikipedia

Robert L. Trivers

(1943–2026) (Rutgers University, Emeritus). Developed the theory of reciprocal altruism, explaining how cooperation among non-kin evolves when partners interact repeatedly, and with it much of the modern logic of social exchange. Google Scholar · Wikipedia

Frans B. M. de Waal

(1948–2024) (Emory University). Primatologist who documented the evolutionary roots of empathy and the building blocks of morality in other primates, and argued, with Stephanie Preston, for a perception–action model of empathy linking human and animal helping. Wikipedia

Felix Warneken

(University of Michigan). Developmental and comparative psychologist whose studies of spontaneous, unrewarded helping in human toddlers and young chimpanzees established that instrumental altruism emerges early and has deep phylogenetic roots. Faculty · Google Scholar

Frequently Asked Questions

What is altruism?

Altruism is behavior that benefits another individual at a cost to the actor. The definition is behavioral: it identifies the act by what it does—helping another while spending one's own resources—without specifying the motive. That is deliberate, because the word is used differently by biologists (a fitness cost), psychologists (an other-regarding motive), and economists (a positive weight on another's payoff), and these senses can come apart.

How can altruism evolve if natural selection rewards self-interest?

Two main routes. Kin selection favors helping genetic relatives, because the helped relative is likely to carry copies of the same genes; Hamilton's rule, rb > c, states exactly when this pays. Reciprocal altruism favors helping non-relatives who are likely to return the favor in future interactions. Both make a costly act pay off in the long run, in the currency of reproduction.

What is the empathy–altruism hypothesis?

It is Daniel Batson's claim that feeling empathic concern for someone in need produces motivation whose ultimate goal is improving that person's welfare, rather than relieving one's own discomfort. It stands against psychological egoism, which holds that all helping is secretly self-serving. The empathy-escape experiments are its decisive test.

How do experiments tell genuine altruism from disguised selfishness?

By manipulating how easily a person can escape the situation. If helping is just a way to relieve one's own distress, an easy exit should let people leave instead of helping. In the experiments, people feeling low empathy do leave when escape is easy—but people feeling high empathy help whether escape is easy or hard, which is the pattern only genuine concern for the other can produce.

What is the warm glow of giving?

The warm glow is the private, affective reward a person gets from the act of giving itself, separate from the good the gift does. Because many donors value this warm glow—impure altruism—they keep giving even when others could provide the same benefit, which is why government funding of a cause only partly crowds out private donations instead of replacing them one for one.

Does altruism show up in the brain?

Yes. Charitable giving activates the brain's reward circuitry—ventral striatum and ventromedial prefrontal cortex—much as receiving money does, and voluntary giving adds an extra reward response beyond a mandatory transfer of the same size. The brain appears to use its general-purpose valuation system to place value on others' welfare, not a separate dedicated module.

Are humans the only altruistic species?

No. Spontaneous, unrewarded helping appears in human toddlers before much socialisation and in young chimpanzees, and the empathic machinery that supports it is shared across social mammals. Altruism is not a purely human cultural invention; culture elaborates a capacity with deep evolutionary roots.

Is altruism the same as cooperation?

They overlap but are not identical. Cooperation is any joint action that benefits the participants, and can be mutually profitable with no net cost to anyone. Altruism is the stricter case in which the actor bears a genuine cost so that another benefits. Much cooperation is sustained by mechanisms—reciprocity, reputation—that make the apparent sacrifice pay off, which is why altruism and cooperation are studied together.

References

Andreoni, J. (1990). Impure altruism and donations to public goods: A theory of warm-glow giving. The Economic Journal, 100(401), 464–477. https://doi.org/10.2307/2234133

Batson, C. D., & Shaw, L. L. (1991). Evidence for altruism: Toward a pluralism of prosocial motives. Psychological Inquiry, 2(2), 107–122. https://doi.org/10.1207/s15327965pli0202_1

Crockett, M. J., Siegel, J. Z., Kurth-Nelson, Z., Dayan, P., & Dolan, R. J. (2017). Moral transgressions corrupt neural representations of value. Nature Neuroscience, 20(6), 879–885. https://doi.org/10.1038/nn.4557

Cutler, J., & Campbell-Meiklejohn, D. (2019). A comparative fMRI meta-analysis of altruistic and strategic decisions to give. NeuroImage, 184, 227–241. https://doi.org/10.1016/j.neuroimage.2018.09.009

de Quervain, D. J.-F., Fischbacher, U., Treyer, V., Schellhammer, M., Schnyder, U., Buck, A., & Fehr, E. (2004). The neural basis of altruistic punishment. Science, 305(5688), 1254–1258. https://doi.org/10.1126/science.1100735

Fehr, E., & Gächter, S. (2002). Altruistic punishment in humans. Nature, 415(6868), 137–140. https://doi.org/10.1038/415137a

Hamilton, W. D. (1964). The genetical evolution of social behaviour. I. Journal of Theoretical Biology, 7(1), 1–16. https://doi.org/10.1016/0022-5193(64)90038-4

Harbaugh, W. T., Mayr, U., & Burghart, D. R. (2007). Neural responses to taxation and voluntary giving reveal motives for charitable donations. Science, 316(5831), 1622–1625. https://doi.org/10.1126/science.1140738

Lockwood, P. L., Apps, M. A. J., Valton, V., Viding, E., & Roiser, J. P. (2016). Neurocomputational mechanisms of prosocial learning and links to empathy. Proceedings of the National Academy of Sciences, 113(35), 9763–9768. https://doi.org/10.1073/pnas.1603198113

Lockwood, P. L., Apps, M. A. J., & Chang, S. W. C. (2020). Is there a ‘social’ brain? Implementations and algorithms. Trends in Cognitive Sciences, 24(10), 802–813. https://doi.org/10.1016/j.tics.2020.06.011

Marsh, A. A. (2016). Neural, cognitive, and evolutionary foundations of human altruism. Wiley Interdisciplinary Reviews: Cognitive Science, 7(1), 59–71. https://doi.org/10.1002/wcs.1377

Moll, J., Krueger, F., Zahn, R., Pardini, M., de Oliveira-Souza, R., & Grafman, J. (2006). Human fronto-mesolimbic networks guide decisions about charitable donation. Proceedings of the National Academy of Sciences, 103(42), 15623–15628. https://doi.org/10.1073/pnas.0604475103

Nowak, M. A. (2006). Five rules for the evolution of cooperation. Science, 314(5805), 1560–1563. https://doi.org/10.1126/science.1133755

Penner, L. A., Dovidio, J. F., Piliavin, J. A., & Schroeder, D. A. (2005). Prosocial behavior: Multilevel perspectives. Annual Review of Psychology, 56, 365–392. https://doi.org/10.1146/annurev.psych.56.091103.070141

Preston, S. D., & de Waal, F. B. M. (2002). Empathy: Its ultimate and proximate bases. Behavioral and Brain Sciences, 25(1), 1–20. https://doi.org/10.1017/S0140525X02000018

Rand, D. G., & Nowak, M. A. (2013). Human cooperation. Trends in Cognitive Sciences, 17(8), 413–425. https://doi.org/10.1016/j.tics.2013.06.003

Ruff, C. C., & Fehr, E. (2014). The neurobiology of rewards and values in social decision making. Nature Reviews Neuroscience, 15(8), 549–562. https://doi.org/10.1038/nrn3776

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