In evolutionary biology, an organism is said to behavealtruistically when its behaviour benefits other organisms, at a costto itself. The costs and benefits are measured in terms ofreproductive fitness, or expected number of offspring. So bybehaving altruistically, an organism reduces the number of offspring itis likely to produce itself, but boosts the number that other organismsare likely to produce. This biological notion of altruism is notidentical to the everyday concept. In everyday parlance, an actionwould only be called ‘altruistic’ if it was done with theconscious intention of helping another. But in the biological sensethere is no such requirement. Indeed, some of the most interestingexamples of biological altruism are found among creatures that are(presumably) not capable of conscious thought at all, e.g. insects. Forthe biologist, it is the consequences of an action for reproductivefitness that determine whether the action counts as altruistic, not theintentions, if any, with which the action is performed.

Altruistic behaviour is common throughout the animal kingdom,particularly in species with complex social structures. For example,vampire bats regularly regurgitate blood and donate it to other membersof their group who have failed to feed that night, ensuring they do notstarve. In numerous bird species, a breeding pair receives help inraising its young from other ‘helper’ birds, who protectthe nest from predators and help to feed the fledglings. Vervet monkeysgive alarm calls to warn fellow monkeys of the presence of predators,even though in doing so they attract attention to themselves,increasing their personal chance of being attacked. In social insectcolonies (ants, wasps, bees and termites), sterile workers devote theirwhole lives to caring for the queen, constructing and protecting thenest, foraging for food, and tending the larvae. Such behaviour ismaximally altruistic: sterile workers obviously do not leave anyoffspring of their own—so have personal fitness of zero—but theiractions greatly assist the reproductive efforts of the queen.

From a Darwinian viewpoint, the existence of altruism in nature isat first sight puzzling, as Darwin himself realized. Natural selectionleads us to expect animals to behave in ways that increase theirown chances of survival and reproduction, not those of others.But by behaving altruistically an animal reduces its own fitness, soshould be at a selective disadvantage vis-à-vis one whichbehaves selfishly. To see this, imagine that some members of a group ofVervet monkeys give alarm calls when they see predators, but others donot. Other things being equal, the latter will have an advantage. Byselfishly refusing to give an alarm call, a monkey can reduce thechance that it will itself be attacked, while at the same timebenefiting from the alarm calls of others. So we should expect naturalselection to favour those monkeys that do not give alarm calls overthose that do. But this raises an immediate puzzle. How did thealarm-calling behaviour evolve in the first place, and why has it notbeen eliminated by natural selection? How can the existence of altruismbe reconciled with basic Darwinian principles?

• 1. Altruism and the Levels of Selection
• 2. Kin Selection and Inclusive Fitness
  • 2.1 A Simple Illustration: the Prisoner's dilemma
• 3. Conceptual Issues
  • 3.1 Altruism, Co-operation, Mutualism
  • 3.2 Weak and Strong Altruism
  • 3.3 Short-term versus Long-term Fitness Consequences
• 4. Reciprocal Altruism
• 5. But is it ‘Real’ Altruism?
• Bibliography
• Academic Tools
• Other Internet Resources
• Related Entries

1. Altruism and the Levels of Selection

The problem of altruism is intimately connected with questions aboutthe level at which natural selection acts. If selection actsexclusively at the individual level, favouring some individualorganisms over others, then it seems that altruism cannot evolve, forbehaving altruistically is disadvantageous for the individual organismitself, by definition. However, it is possible that altruism may beadvantageous at thegroup level. A group containing lots ofaltruists, each ready to subordinate their own selfish interests forthe greater good of the group, may well have a survival advantage overa group composed mainly or exclusively of selfish organisms. A processof between-group selection may thus allow the altruistic behaviour toevolve.Within each group, altruists will be at a selectivedisadvantage relative to their selfish colleagues, but the fitness ofthe group as a whole will be enhanced by the presence of altruists.Groups composed only or mainly of selfish organisms go extinct,leaving behind groups containing altruists. In the example of theVervet monkeys, a group containing a high proportion of alarm-callingmonkeys will have a survival advantage over a group containing a lowerproportion. So conceivably, the alarm-calling behaviour may evolve bybetween-group selection, even though within each group, selectionfavours monkeys that do not give alarm calls.

The idea that group selection might explain the evolution ofaltruism was first broached by Darwin himself. InThe Descent ofMan (1871), Darwin discussed the origin of altruistic andself-sacrificial behaviour among humans. Such behaviour is obviouslydisadvantageous at the individual level, as Darwin realized: “hewho was ready to sacrifice his life, as many a savage has been, ratherthan betray his comrades, would often leave no offspring to inherit hisnoble nature” (p.163). Darwin then argued that self-sacrificialbehaviour, though disadvantageous for the individual‘savage’, might be beneficial at the group level: “atribe including many members who...were always ready to give aid toeach other and sacrifice themselves for the common good, would bevictorious over most other tribes; and this would be naturalselection” (p.166). Darwin's suggestion is that the altruisticbehaviour in question may have evolved by a process of between-groupselection.

The concept of group selection has a chequered and controversialhistory in evolutionary biology. The founders of modern neo-Darwinism—R.A. Fisher, J.B.S. Haldane and S. Wright—were all aware thatgroup selection could in principle permit altruistic behaviours toevolve, but they doubted the importance of this evolutionarymechanism. Nonetheless, many mid-twentieth century ecologists andsome ethologists, notably Konrad Lorenz, routinely assumed thatnatural selection would produce outcomes beneficial for the wholegroup or species, often without even realizing that individual-levelselection guarantees no such thing. This uncritical ‘good of thespecies’ tradition came to an abrupt halt in the 1960s, duelargely to the work of G.C. Williams (1966) and J. Maynard Smith(1964). These authors argued that group selection was an inherentlyweak evolutionary force, hence unlikely to promote interestingaltruistic behaviours. This conclusion was supported by a number ofmathematical models, which apparently showed that group selectionwould only have significant effects for a limited range of parametervalues. As a result, the notion of group selection fell intowidespread disrepute in orthodox evolutionary circles; see Sober andWilson 1998, Segestrale 2000, Okasha 2006, Leigh 2010 and Sober 2011 for details of thehistory of this debate.

The major weakness of group selection as an explanation of altruism,according to the consensus that emerged in the 1960s, was a problemthat Dawkins (1976) called ‘subversion from within’; seealso Maynard Smith 1964. Even if altruism is advantageous at thegroup level, within any group altruists are liable to be exploited byselfish ‘free-riders’ who refrain from behavingaltruistically. These free-riders will have an obvious fitnessadvantage: they benefit from the altruism of others, but do not incurany of the costs. So even if a group is composed exclusively ofaltruists, all behaving nicely towards each other, it only takes asingle selfish mutant to bring an end to this happy idyll. By virtue ofits relative fitness advantage within the group, the selfish mutantwill out-reproduce the altruists, hence selfishness will eventuallyswamp altruism. Since the generation time of individual organisms islikely to be much shorter than that of groups, the probability that aselfish mutant will arise and spread is very high, according to thisline of argument. ‘Subversion from within’ is generallyregarded as a major stumbling block for group-selectionist theoriesof the evolution of altruism.

If group selection is not the correct explanation for how thealtruistic behaviours found in nature evolved, then what is? In the1960s and 1970s a rival theory emerged: kin selection or‘inclusive fitness’ theory, due originally to Hamilton(1964). This theory, discussed in detail below, apparently showed howaltruistic behaviour could evolvewithout the need for group-level selection, and quicklygained prominence among biologists interested in the evolution ofsocial behaviour; the empirical success of kin selection theorycontributed to the demise of the group selection concept. However, theprecise relation between kin and group selection is a source ofongoing controversy (see for example the recent exchangeinNature between Nowak, Tarnita and Wilson 2010 and Abbotet. al. 2011). Since the 1990s, proponents of ‘multi-levelselection theory’ have resuscitated a form of group-levelselection—sometimes called ‘new’ group selection—andshown that it can permit altruism to evolve (cf. Sober and Wilson1998). But ‘new’ group selection turns out to bemathematically equivalent to kin selection in most if not all cases,as a number of authors have emphasized (Grafen 1984, Frank 1998, Westet al. 2007, Lehmannet al. 2007, Marshall 2011); this point wasalready appreciated by Hamilton (1975). Since the relation between‘old’ and ‘new’ group selection is itself apoint of controversy, this explains why disagreement about therelation between kin and group selection should persist.

2. Kin Selection and Inclusive Fitness

The basic idea of kin selection is simple. Imagine a gene whichcauses its bearer to behave altruistically towards other organisms,e.g. by sharing food with them. Organisms without the gene are selfish—they keep all their food for themselves, and sometimes get handoutsfrom the altruists. Clearly the altruists will be at a fitnessdisadvantage, so we should expect the altruistic gene to be eliminatedfrom the population. However, suppose that altruists are discriminatingin who they share food with. They do not share with just anybody, butonly with their relatives. This immediately changes things. Forrelatives are genetically similar—they share genes with one another.So when an organism carrying the altruistic gene shares his food, thereis a certain probability that the recipients of the food will alsocarry copies of that gene. (How probable depends on how closely relatedthey are.) This means that the altruistic gene can in principle spreadby natural selection. The gene causes an organism to behave in a waywhich reduces its own fitness but boosts the fitness of its relatives—who have a greater than average chance of carrying the genethemselves. So the overall effect of the behaviour may be to increasethe number of copies of the altruistic gene found in the nextgeneration, and thus the incidence of the altruistic behaviouritself.

Though this argument was hinted at by Haldane in the 1930s, and to alesser extent by Darwin in his discussion of sterile insect castesin The Origin of Species, it was first made explicit byWilliam Hamilton (1964) in a pair of seminal papers. Hamiltondemonstrated rigorously that an altruistic gene will be favoured bynatural selection when a certain condition, known asHamilton's rule, is satisfied. In its simplest version, therule states thatb >c/r, wherec isthe cost incurred by the altruist (the donor), b is the benefitreceived by the recipients of the altruism, and r istheco-efficient of relationship between donor andrecipient. The costs and benefits are measured in terms ofreproductive fitness. The co-efficient of relationship depends on thegenealogical relation between donor and recipient—it isdefined as the probability that donor and recipient share genes at agiven locus that are ‘identical by descent’. (Two genesare identical by descent if they are copies of a single gene in ashared ancestor.) In a sexually reproducing diploid species, the valueof r for full siblings is ½, for parents and offspring½, for grandparents and grandoffspring ¼, for fullcousins 1/8,  and so-on. The higher the value of r, the greaterthe probability that the recipient of the altruistic behaviour willalso possess the gene for altruism. So what Hamilton's rule tells usis that a gene for altruism can spread by natural selection, so longas the cost incurred by the altruist is offset by a sufficient amountof benefit to sufficiently closed related relatives. The proof ofHamilton's rule relies on certain non-trivial assumptions; see Frank1998, Grafen 1985, 2006, Queller 1992a, 1992b, Boyd and McIlreath 2006and Birch forthcoming for details.

Though Hamilton himself did not use the term, his idea quicklybecame known as ‘kin selection’, for obvious reasons. Kinselection theory predicts that animals are more likely to behavealtruistically towards their relatives than towards unrelated membersof their species. Moreover, it predicts that thedegree ofaltruism will be greater, the closer the relationship. In the yearssince Hamilton's theory was devised, these predictions have been amplyconfirmed by empirical work. For example, in various bird species, ithas been found that ‘helper’ birds are much more likely tohelp relatives raise their young, than they are to help unrelatedbreeding pairs. Similarly, studies of Japanese macaques have shown thataltruistic actions, such as defending others from attack, tend to bepreferentially directed towards close kin. In most social insectspecies, a peculiarity of the genetic system known as‘haplodiploidy’ means that females on average share moregenes with their sisters than with their own offspring. So a female maywell be able to get more genes into the next generation by helping thequeen reproduce, hence increasing the number of sisters she will have,rather than by having offspring of her own. Kin selection theorytherefore provides a neat explanation of how sterility in the socialinsects may have evolved by Darwinian means. (Note, however, that theprecise significance of haplodiploidy for the evolution of workersterility is a controversial question; see Maynard Smith and Szathmary1995 ch.16, Gardner, Alpedrinha and West 2012.)

Kin selection theory is often presented as a triumph of the‘gene's-eye view of evolution’, which sees organicevolution as the result of competition among genes for increasedrepresentation in the gene-pool, and individual organisms as mere‘vehicles’ that genes have constructed to aid theirpropagation (Dawkins 1976, 1982). The gene's eye-view is certainlythe easiest way of understanding kin selection, and was employed byHamilton himself in his 1964 papers. Altruism seems anomalous from theindividual organism's point of view, but from the gene's point of viewit makes good sense. A gene wants to maximize the number of copies ofitself that are found in the next generation; one way of doing that isto cause its host organism to behave altruistically towards otherbearers of the gene, so long as the costs and benefits satisfy theHamilton inequality. But interestingly, Hamilton showed that kinselection can also be understood from the organism's point of view.Though an altruistic behaviour which spreads by kin selection reducesthe organism's personal fitness (by definition), it increases whatHamilton called the organism'sinclusive fitness. Anorganism's inclusive fitness is defined as its personal fitness, plusthe sum of its weighted effects on the fitness of every other organismin the population, the weights determined by the coefficient ofrelationship r. Given this definition, natural selection will act tomaximise the inclusive fitness of individuals in the population (Grafen 2006).Instead of thinking in terms of selfish genes trying to maximize theirfuture representation in the gene-pool, we can think in terms oforganisms trying to maximize their inclusive fitness. Most people findthe ‘gene's eye’ approach to kin selection heuristicallysimpler than the inclusive fitness approach, but mathematically theyare in fact equivalent (Michod 1982, Frank 1998, Boyd andMcIlreath 2006, Grafen 2006).

Contrary to what is sometimes thought, kin selection does notrequire that animals must have the ability to discriminate relativesfrom non-relatives, less still to calculate coefficients ofrelationship. Many animals can in fact recognize their kin, often bysmell, but kin selection can operate in the absence of such an ability.Hamilton's inequality can be satisfied so long as an animal behavesaltruistically towards other animals that arein fact itsrelatives. The animalmight achieve this by having the abilityto tell relatives from non-relatives, but this is not the onlypossibility. An alternative is to use some proximal indicator ofkinship. For example, if an animal behaves altruistically towards thosein its immediate vicinity, then the recipients of the altruism arelikely to be relatives, given that relatives tend to live near eachother. No ability to recognize kin is presupposed. Cuckoos exploitprecisely this fact, free-riding on the innate tendency of birds tocare for the young in their nests.

Another popular misconception is that kin selection theory iscommitted to ‘genetic determinism’, the idea that genesrigidly determine or control behaviour. Though some sociobiologistshave made incautious remarks to this effect, evolutionary theories ofbehaviour, including kin selection, are not committed to it. So long asthe behaviours in question have a geneticalcomponent, i.e.are influenced to some extent by one or more genetic factor, then thetheories can apply. When Hamilton (1964) talks about a gene which‘causes’ altruism, this is really shorthand for a genewhich increases the probability that its bearer will behavealtruistically, to some degree. This is much weaker than saying thatthe behaviour is genetically ‘determined’, and is quitecompatible with the existence of strong environmental influences on thebehaviour's expression. Kin selection theory does not deny the truismthat all traits are affected by both genes and environment. Nor does itdeny that many interesting animal behaviours are transmitted throughnon-genetical means, such as imitation and social learning (Avital andJablonka 2000).

The importance of kinship for the evolution of altruism is very widelyaccepted today, on both theoretical and empirical grounds. However,kinship is really only a way of ensuring that altruists and recipientsboth carry copies of the altruistic gene, which is the fundamentalrequirement. If altruism is to evolve, it must be the case that therecipients of altruistic actions have a greater than averageprobability of being altruists themselves. Kin-directed altruism isthe most obvious way of satisfying this condition, but there are otherpossibilities too (Hamilton 1975, Sober and Wilson 1998, Bowles andGintis 2011, Gardner and West 2011). For example, if the gene thatcauses altruism also causes animals to favour a particular feedingground (for whatever reason), then the required correlation betweendonor and recipient may be generated. It is this correlation, howeverbrought about, that is necessary for altruism to evolve. This pointwas noted by Hamilton himself in the 1970s: he stressed that thecoefficient of relationship of his 1964 papers should really bereplaced with a more general correlation coefficient, which reflectsthe probability that altruist and recipient share genes, whetherbecause of kinship or not (Hamilton 1970, 1972, 1975). This point istheoretically important, and has not always been recognized; but inpractice, kinship remains the most important source of statisticalassociations between altruists and recipients (Maynard Smith 1998,Okasha 2002, Westet al. 2007).

2.1 A Simple Illustration: the Prisoner's dilemma

The fact that correlation between donor and recipient is the key tothe evolution of altruism can be illustrated via a simple ‘one shot’Prisoner's dilemma game. Consider a large population of organisms whoengage in a social interaction in pairs; the interaction affects theirbiological fitness. Organisms are of two types: selfish (S) andaltruistic (A). The latter engage in pro-social behaviour, thusbenefiting their partner but at a cost to themselves; the former donot. So in a mixed (S,A) pair, the selfish organism does better—hebenefits from his partner's altruism without incurring anycost. However, (A,A) pairs do better than (S,S) pairs—for the formerwork as a co-operative unit, while the latter do not. The interactionthus has the form of a one-shot Prisoner's dilemma, familiar from gametheory. Illustrative payoff values to each ‘player’, i.e., each partnerin the interaction, measured in units of biological fitness, are shownin the matrix below.

		Player 2
		Altruist	Selfish
Player 1	Altruist	11,11	0,20
	Selfish	20,0	5,5

Payoffs for (Player 1, Player 2) in units of reproductivefitness

The question we are interested in is: which type will be favoured byselection? To make the analysis tractable, we make two simplifyingassumptions: that reproduction is asexual, and that type is perfectlyinherited, i.e., selfish (altruistic) organisms give rise to selfish(altruistic) offspring. Modulo these assumptions, the evolutionarydynamics can be determined very easily, simply by seeing whethertheS or theA type has higher fitness, in theoverall population. The fitness of theStype,W(S), is the weighted average of the payoff toanS when partnered with anS and the payoff toanS when partnered with anA, where the weights aredetermined by the probability of having the partner inquestion. Therefore,

W(S) = 5 * Prob(S partner/S) + 20 * Prob(A partner/S)

(The conditional probabilities in the above expression should be readas the probability of having a selfish (altruistic) partner, giventhat one is selfish oneself.)

Similarly, the fitness of theA type is:

W(A) = 0 * Prob(S partner/A) + 11 * Prob(A partner/A)

From these expressions for the fitnesses of the two types of organism,we can immediately deduce that the altruistic type will only befavoured by selection if there is a statistical correlation betweenpartners, i.e., if altruists have greater than random chance of beingpaired with other altruists, and similarly for selfish types. Forsuppose there is no such correlation—as would be the case if thepairs were formed by random sampling from the population. Then, theprobability of having a selfish partner would be the same forbothS andA types, i.e., P(Spartner/S) = P(S partner/A). Similarly,P(A partner/S) = P(Apartner/A). From these probabilistic equalities, it followsimmediately thatW(S) is greaterthanW(A), as can be seen from the expressions forW(S) andW(A) above; so theselfish type will be favoured by natural selection, and will increasein frequency every generation until all the altruists are eliminatedfrom the population. Therefore, in the absence of correlation betweenpartners, selfishness must win out (cf. Skyrms 1996). This confirms the point noted insection 2—that altruism can only evolve if there is a statisticaltendency for the beneficiaries of altruistic actions to be altruiststhemselves.

If the correlation between partners is sufficiently strong, in thissimple model, then it is possible for theconditionW(A) >W(S) to besatisfied, and thus for altruism to evolve. The easiest way to seethis is to suppose that the correlation is perfect, i.e., selfishtypes are always paired with other selfish types, and ditto foraltruists, so P(S partner/S) = P(Apartner/A) = 1. This assumption impliesthatW(A)=11 andW(S)=5, soaltruism evolves. With intermediate degrees of correlation, it is alsopossible for the conditionW(S)>W(A) to be satisfied, given the particularchoice of payoff values in the model above.

This simple model also highlights the point made previously, thatdonor-recipient correlation, rather than genetic relatedness, is thekey to the evolution of altruism. What is needed for altruism toevolve, in the model above, is for the probability of having a partnerof the same type as oneself to be sufficiently larger than theprobability of having a partner of opposite type; this ensures thatthe recipients of altruism have a greater than random chance of beingfellow altruists, i.e., donor-recipient correlation. Whether thiscorrelation arises because partners tend to be relatives, or becausealtruists are able to seek out other altruists and choose them aspartners, or for some other reason, makes no difference to theevolutionary dynamics, at least in this simple example.

3. Conceptual Issues

Altruism is a well understood topic in evolutionary biology; thetheoretical ideas explained above have been extensively analysed,empirically confirmed, and are widely accepted. Nonetheless, there area number of conceptual ambiguities surrounding altruism and relatedconcepts in the literature; some of these are purely semantic, othersare more substantive. Three such ambiguities are briefly discussedbelow; for further discussion, see Westet al. 2007, Sachset al. 2004 or Lehmann and Keller 2006.

3.1 Altruism, Co-operation, Mutualism

According to the standard definition, a social behaviour counts asaltruistic if it reduces the fitness of the organism performing thebehaviour, but boosts the fitness of others. This was the definitionused by Hamilton (1964), and by many subsequent authors. However,there is less consensus on how to describe behaviours that boost thefitness of others but also boost the fitness of the organismperforming the behaviour. As Westet al. (2007) note, suchbehaviours are sometimes termed ‘co-operative’, but thisusage is not universal; others use ‘co-operation’ to referto behaviour that boosts the fitness of others irrespective of itseffect on self; while still others use ‘cooperation’ as asynonym for altruism. (Indeed, in the simple Prisoner's dilemma gameabove, the two strategies are usually called ‘co-operate’and ‘defect’.) To avoid this confusion, Westetal. (2007) suggest the term ‘mutual benefit’ forbehaviours that benefit both self and other, while Sachsetal. (2004) suggest ‘byproduct benefit’.

Whatever term is used, the important point is that behaviours thatbenefit both self and others can evolve much more easily thanaltruistic behaviours, and thus require no special mechanisms such askinship. The reason is clear: organisms performing such behavioursthereby increase their personal fitness, so are at a selectiveadvantage vis-a-vis those not performing the behaviour. The fact thatthe behaviour has a beneficial effect on the fitness of others is amere side-effect, or byproduct, and is not part of the explanation forwhy the behaviour evolves. For example, Sachset al. (2004)note that an action such as joining a herd or a flock may be of thissort; the individual gains directly, via his reduced risk ofpredation, while simultaneously reducing the predation risk of otherindividuals. By contrast with an altruistic action, there is nopersonal incentive to ‘cheat’, i.e., to refrain fromperforming the action, for doing so would directly reduce personalfitness.

Also indicative of the difference between altruistic behaviour andbehaviour that benefit both self and others is the fact that in thelatter case, though not the former, the beneficiary may be a member ofa different species, without altering the evolutionary dynamics of thebehaviour. Indeed, there are numerous examples where theself-interested activities of one organism produce an incidentalbenefit for a non-conspecific; such behaviours are sometimes called‘mutualistic’, though again, this is not the only way thatthe latter term has been used (Westet al. 2007). Bycontrast, in the case of altruism, it makes an enormous differencewhether the beneficiary and the donor are con-specifics or not; forif not, then kin selection can play no role, and it is quite unclear howthe altruistic behaviour can evolve. Unsurprisingly, virtually all thebona fide examples of biological altruism in the living world involvedonors and recipients that are con-specifics. (Cases of so-called‘reciprocal altruism’ are sometimes thought to beexceptions to this generalization; but see section 4 below.)

3.2 Weak and Strong Altruism

A quite different ambiguity concerns the distinction between weak andstrong altruism, in the terminology of D.S. Wilson (1977, 1980,1990). This distinction is about whether the altruistic action entailsan absolute or relative fitness reduction for the donor. To count asstrongly altruistic, a behaviour must reduce theabsolutefitness (i.e., number of offspring) of the donor. Strong altruism isthe standard notion of altruism in the literature, and was assumedabove. To count as weakly altruistic, an action need only reducetherelative fitness of the donor, i.e., its fitness relativeto that of the recipient. Thus for example, an action which causes anorganism to leave an additional 10 offspring, but causes eachorganism(s) with which it interacts to leave an additional 20offspring, is weakly but not strongly altruistic. The action booststhe absolute fitness of the ‘donor’, but boosts theabsolute fitness of other organisms by even more, thus reducing thedonor's relative fitness.

Should weakly altruistic behaviours be classified as altruistic orselfish? This question is not merely semantic; for the real issue iswhether the conditions under which weak altruism can evolve arerelevantly similar to the conditions under which strong altruism canevolve, or not. Many authors argue that the answer is‘no’, on the grounds that weakly altruistic behaviours areindividually advantageous, so can evolve with no component of kinselection or donor-recipient correlation, unlike strongly altruisticbehaviours (Grafen 1984, Nunney 1985, Westetal. 2007). To appreciate this argument, consider agame-theoretic scenario similar to the one-shot Prisoner's dilemma ofsection 4, in which organisms engage in a pair-wise interaction thataffects their fitness. Organisms are of two types, weakly altruistic(W) and non-altruistic (N).W-types performan action that boosts their own fitness by 10 units and the fitness oftheir partner by 20 units;N-types do not perform the action. Thepayoff matrix is thus:

		Player 2
		Weak Altruist	Non
Player 1	Weak Altruist	30,30	10,20
	Non	20,10	0,0

Payoffs for (Player 1, Player 2) in units of reproductivefitness

The payoff matrix highlights the fact that weak altruism isindividually advantageous, and thus the oddity of thinking of it it asaltruistic rather than selfish. To see this, assume for a moment thatthe game is being played by two rational agents, as in classical gametheory. Clearly, the rational strategy for each individualisW, forW dominatesN. Each individualgets a higher payoff from playingWthanN,irrespective of what its opponent does—30 rather than 20 if the opponent playsW, 10 ratherthan 0 if the opponent playsN. This captures a clear sensein which weak altruism is individually advantageous.

In the context of evolutionary game theory, where the game is beingplayed by pairs of organisms with hard-wired strategies, thecounterpart of the fact thatW dominatesN is thefact thatW can spread in the population even if pairs areformed at random (cf. Wilson 1980). To see this, consider theexpressions for the overall population-wide fitnesses ofWandN:

W(W) = 30 * Prob(W partner/W) + 10* Prob(N partner/W)

W(N) = 20 * Prob(W partner/N) + 0* Prob(N partner/N)

(As before, Prob(W partner/W) denotes theconditional probability of having a weakly altruistic partner giventhat one is weakly altruistic oneself, and so-on.) From theseexpressions, it is easy to see thatW(W)>W(N) even if the there is no correlation amongpartners, i.e., even if Prob(W partner/W) =P(W partner/N) and P(N partner/W)= P(N partner/N). Therefore, weak altruism can evolve in theabsence of donor-recipient correlation; as we saw, this is not true ofstrong altruism. So weak and strong altruism evolve by differentevolutionary mechanisms, hence should not be co-classified, accordingto this argument.

However, there is a counter argument due to D.S. Wilson (1977, 1980),who maintains that weak altruism cannot evolve by individual selectionalone; a component of group selection is needed. Wilson's argumentstems from the fact that in a mixed (W,N) pair, thenon-altruist is fitter than the weak altruist. More generally, withina single group of any size containing weak altruists andnon-altruists, the latter will be fitter. So weak altruism can onlyevolve, Wilson argues, in a multi-group setting—in which thewithin-group selection in favour ofN, is counteracted bybetween-group selection in favour ofW. (On Wilson's view, theevolutionary game described above is a multi-group setting, involvinga large number of groups of size two.) Thus weak altruism, like strongaltruism, in fact evolves because it is group-advantageous, Wilsonargues.

The dispute between those who regard weak altruism as individuallyadvantageous, and those like Wilson who regard it as groupadvantageous, stems ultimately from differing conceptions ofindividual and group selection. For Wilson, individual selection meanswithin-group selection, so to determine which strategy is favoured byindividual selection, one must compare the fitnesses ofWandN types within a group, or pair. For other theorists,individual selection means selection based on differences inindividual phenotype, rather than social context; so to determinewhich strategy is favoured by individual selection, one must comparethe fitnesses ofW andN types in the same socialcontext, i.e., with the same partner. These two comparisons yielddifferent answers to the question of whether weak altruism isindividually advantageous. Thus the debate over how to classify weakaltruism is intimately connected to the broader levels of selectionquestion; see Nunney 1985, Okasha 2005, 2006, Fletcher and Doebeli2006, Westet al. 2007, for further discussion.

3.3 Short-term versus Long-term Fitness Consequences

A further source of ambiguity in the definition of biological altruismconcerns the time-scale over which fitness is measured. Conceivably,an animal might engage in a social behaviour which benefits anotherand reduces its own (absolute) fitness in the short-term; however, inthe long-term, the behaviour might be to the animal's advantage. So ifwe focus on short-term fitness effects, the behaviour will seemaltruistic; but if we focus on lifetime fitness, the behaviour willseem selfish—the animal's lifetime fitness would be reduced if itdid not perform the behaviour.

Why might a social behaviour reduce an animal's short-term fitness butboost its lifetime fitness? This could arise in cases of ‘directedreciprocation’, where the beneficiary of the behaviour returns thefavour at some point in the future (cf. Sachset al. 2004). Byperforming the behaviour, and suffering the short-term cost, theanimal thus ensures (or raises the chance) that it will receive returnbenefits in the future. Similarly, in symbioses between members ofdifferent species, it may pay an organism to sacrifice resources forthe benefit of a symbiont with which it has a long-term relationship,as its long-term welfare may be heavily dependent on the symbiont'swelfare.

From a theoretical point of view, the most satisfactory resolution ofthis ambiguity is to use lifetime fitness as the relevant parameter(cf. Westet al. 2007) Thus an action only counts asaltruistic if it reduces an organism's lifetime fitness. Thisstipulation makes sense, since it preserves the key idea that theevolution of altruism requires statistical association between donorand recipient; this would not be true if short-term fitness were usedto define altruism, for behaviours which reduce short-term fitness butboost lifetime fitness can evolve with no component of kin selection,or donor-recipient correlation. However, the stipulation has twodisadvantages: (i) it makes it harder to tell whether a givenbehaviour is altruistic, since lifetime fitness is notoriouslydifficult to estimate; (ii) it has the consequence that most models of‘reciprocal altruism’ are mis-named.

4. Reciprocal Altruism

The theory of reciprocal altruism was originally developed by Trivers(1971), as an attempt to explain cases of (apparent) altruism amongunrelated organisms, including members of different species. (Clearly,kin selection cannot help explain altruism among non-relatives.)Trivers' basic idea was straightforward: it may pay an organism tohelp another, if there is an expectation of the favour being returnedin the future. (‘If you scratch my back, I'll scratchyours’.) The cost of helping is offset by the likelihood of thereturn benefit, permitting the behaviour to evolve by naturalselection. Trivers termed with evolutionary mechanism‘reciprocal altruism’.

For reciprocal altruism to work, there is no need for the twoindividuals to be relatives, nor even to be members of the samespecies. However, it is necessary that individuals should interactwith each more than once, and have the ability to recognize otherindividuals with whom they have interacted in the past.^[1] If individuals interact only once in their lifetimes and never meetagain, there is obviously no possibility of return benefit, so thereis nothing to be gained by helping another. However, if individualsencounter each other frequently, and are capable of identifying andpunishing ‘cheaters’ who have refused to help in the past,then the helping behaviour can evolve. A ‘cheat’ whorefuses to help will ultimately sabotage his own interests, foralthough he does not incur the cost of helping others, he forfeits thereturn benefits too—others will not help him in thefuture. This evolutionary mechanism is most likely to work whereanimals live in relatively small groups, increasing the likelihood ofmultiple encounters.

As Westet al. (2007) and Bowles and Gintis (2011) note, if altruism is defined by reference tolifetime fitness, then Trivers' theory is not really about theevolution of altruism at all; for behaviours that evolve viareciprocation of benefits, as described by Trivers, are ultimately ofdirect benefit to the individuals performing them, so do not reducelifetime fitness. Despite this consideration, the label ‘reciprocalaltruism’ is well-entrenched in the literature, and the evolutionarymechanism that it describes is of some importance, whatever it iscalled. Where reciprocal altruism is referred to below, it should beremembered that the behaviours in question are only altruistic in theshort-term.

The concept of reciprocal altruism is closely related to theTit-for-Tat strategy in the iterated Prisoner's Dilemma (IPD) fromgame theory. In the IPD, players interact on multiple occasions, andare able to adjust their behaviour depending on what their opponenthas done in previous rounds. There are two possible strategies,co-operate and defect; the payoff matrix (per interaction) is as insection 2.1 above. The fact that the game is iterated rather thanone-shot obviously changes the optimal course of action; defecting isno longer necessarily the best option, so long as the probability ofsubsequent encounters is sufficiently high. In their famous computertournament in which a large number of strategies were pitted againsteach other in the IPD, Axelrod and Hamilton (1981) found that theTit-for-Tat strategy yielded the highest payoff. In Tit-For-Tat, aplayer follows two basic rules: (i) on the first encounter, cooperate;(ii) on subsequent encounters, do what your opponent did on theprevious encounter. The success of Tit-for-Tat was widely taken toconfirm the idea that with multiple encounters, natural selectioncould favour social behaviours that entail a short-term fitnesscost. Subsequent work in evolutionary game theory, much of it inspiredby Axelrod and Hamilton's ideas, has confirmed that repeated gamespermit the evolution of social behaviours that cannot evolve inone-shot situations (cf. Nowak 2006); this is closely related to theso-called 'folk theorem' of repeated game theory in economics(cf. Bowles and Gintis 2011). For a useful discussion of socialbehaviour that evolves via reciprocation of benefits, see Sachset al. 2004.

Despite the attention paid to reciprocal altruism by theoreticians,clear-cut empirical examples in non-human animals are relatively few(Hammerstein 2003, Sachset al. 2004, Taborsky 2013). This isprobably because the pre-conditions for reciprocal altruism to evolve-multiple encounters and individual recognition—are not especiallycommon. However, one possible example is provided by blood-sharing invampire bats (Wilkinson 1984, 1990, Carter & Wilkinson 2013). Itis quite common for a vampire bat to fail to feed on a givennight. This is potentially fatal, for bats die if they go without foodfor more than a couple of days. On any given night, bats donate blood(by regurgitation) to other members of their group who have failed tofeed, thus saving them from starvation. Since vampire bats live insmall groups and associate with each other over long periods of time,the preconditions for reciprocal altruism are likely to be met.Wilkinson and his colleagues' studies showed that bats tended to sharefood with their close associates, and were more likely to share withothers that had recently shared with them. These findings appear toaccord with reciprocal altruism theory.

Trivers (1985) describes an apparent case of reciprocal altruismbetween non con-specifics. On tropical coral reefs, various species ofsmall fish act as ‘cleaners’ for large fish, removingparasites from their mouths and gills. The interaction is mutuallybeneficial—the large fish gets cleaned and the cleaner getsfed. However, Trivers notes that the large fish sometimes appear tobehave altruistically towards the cleaners. If a large fish isattacked by a predator while it has a cleaner in its mouth, then itwaits for the cleaner to leave before fleeing the predator, ratherthan swallowing the cleaner and fleeing immediately. Trivers explainsthe larger fish's behaviour in terms of reciprocal altruism. Since thelarge fish often returns to the same cleaner many times over, it paysto look after the cleaner's welfare, i.e., not to swallow it, even ifthis increases the chance of being wounded by a predator. So thelarger fish allows the cleaner to escape, because there is anexpectation of return benefit—getting cleaned again in thefuture. As in the case of the vampire bats, it is because the largefish and the cleaner interact more than once that the behaviour canevolve.

5. But is it ‘Real’ Altruism?

The evolutionary theories described above, in particular kinselection, go a long way towards reconciling the existence of altruismin nature with Darwinian principles. However, some people have feltthese theories in a way devalue altruism, and that the behaviours theyexplain are not ‘really’ altruistic. The grounds for thisview are easy to see. Ordinarily we think of altruistic actions asdisinterested, done with the interests of the recipient, rather thanour own interests, in mind. But kin selection theory explainsaltruistic behaviour as a clever strategy devised by selfish genes asa way of increasing their representation in the gene-pool, at theexpense of other genes. Surely this means that the behaviours inquestion are only ‘apparently’ altruistic, for they areultimately the result of genic self-interest? Reciprocal altruismtheory also seems to ‘take the altruism out ofaltruism’. Behaving nicely to someone in order to procure returnbenefits from them in the future seems in a way the antithesis of‘real’ altruism—it is just delayed self-interest.

This is a tempting line of argument. Indeed Trivers (1971) and,arguably, Dawkins (1976) were themselves tempted by it. But it shouldnot convince. The key point to remember is that biological altruismcannot be equated with altruism in the everyday vernacular sense.Biological altruism is defined in terms of fitness consequences, notmotivating intentions. If by ‘real’ altruism we meanaltruism done with the conscious intention to help, then the vastmajority of living creatures are not capable of ‘real’altruism nor therefore of ‘real’ selfishness either. Antsand termites, for example, presumably do not have conscious intentions,hence their behaviour cannot be done with the intention of promotingtheir own self-interest, nor the interests of others. Thus theassertion that the evolutionary theories reviewed above show that thealtruism in nature is only apparent makes little sense. The contrastbetween ‘real’ altruism and merely apparent altruism simplydoes not apply to most animal species.

To some extent, the idea that kin-directed altruism is not‘real’ altruism has been fostered by the use of the‘selfish gene’ terminology of Dawkins (1976). As we haveseen, the gene's-eye perspective is heuristically useful forunderstanding the evolution of altruistic behaviours, especially thosethat evolve by kin selection. But talking about ‘selfish’genes trying to increase their representation in the gene-pool is ofcourse just a metaphor (as Dawkins fully admits); there is no literalsense in which genes ‘try’ to do anything.Anyevolutionary explanation of how a phenotypic trait evolves mustultimately show that the trait leads to an increase in frequency ofthe genes that code for it (presuming the trait is transmittedgenetically.) Therefore, a ‘selfish gene’ story can bydefinition be told about any trait, including a behavioural trait,that evolves by Darwinian natural selection. To say that kin selectioninterprets altruistic behaviour as a strategy designed by‘selfish’ genes to aid their propagation is not wrong; butit is just another way of saying that a Darwinian explanation for theevolution of altruism has been found. As Sober and Wilson (1998) note,if one insists on saying that behaviours which evolve by kin selection/ donor-recipient correlation are ‘really selfish’, oneends up reserving the word ‘altruistic’ for behaviourswhich cannot evolve by natural selection at all.

Do theories of the evolution of biological altruism apply to humans?This is part of the broader question of whether ideas about theevolution of animal behaviour can be extrapolated to humans, aquestion that fuelled the sociobiology controversy of the 1980s and isstill actively debated today (cf. Boyd and Richerson 2006, Bowles andGintis 2011, Sterelny 2012). All biologists accept thatHomo sapiens is anevolved species, and thus that general evolutionary principles applyto it. However, human behaviour is obviously influenced by culture toa far greater extent than that of other animals, and is often theproduct of conscious beliefs and desires (though this does notnecessarily mean that genetics has no influence.) Nonetheless, atleast some human behaviour does seem to fit the predictions of theevolutionary theories reviewed above. In general, humans behave morealtruistically (in the biological sense) towards their close kin thantowards non-relatives, e.g. by helping relatives raise their children,just as kin selection theory would predict. It is also true that wetend to help those who have helped us out in the past, just asreciprocal altruism theory would predict. On the other hand, humansare unique in that we co-operate extensively with our non-kin; andmore generally, numerous human behaviours seem anomalous from thepoint of view of biological fitness. Think for example of adoption.Parents who adopt children instead of having their own reduce theirbiological fitness, obviously, so adoption is an altruistic behaviour.But it does not benefit kin—for parents are generallyunrelated to the infants they adopt—and nor do the parentsstand to gain much in the form of reciprocal benefits. So althoughevolutionary considerations can help us understand some humanbehaviours, they must be applied judiciously.

Where human behaviour is concerned, the distinction between biologicalaltruism, defined in terms of fitness consequences, and‘real’ altruism, defined in terms of the agent's consciousintentions to help others, does make sense. (Sometimes the label‘psychological altruism’ is used instead of‘real’ altruism.) What is the relationship between thesetwo concepts? They appear to be independent in both directions, asElliott Sober (1994) has argued; see also Vromen (2012) and Clavien and Chapuisat (2013). An actionperformed with the conscious intention of helping another human beingmay not affect their biological fitness at all, so would not count asaltruistic in the biological sense. Conversely, an action undertakenfor purely self-interested reasons, i.e., without the consciousintention of helping another, may boost their biological fitnesstremendously.

Sober argues that, even if we accept an evolutionary approach to humanbehaviour, there is no particular reason to think that evolution wouldhave made humans into egoists rather than psychological altruists (seealso Schulz 2011). On the contrary, it is quite possible that naturalselection would have favoured humans who genuinely do care abouthelping others, i.e., who are capable of ‘real’ orpsychological altruism. Suppose there is an evolutionary advantageassociated with taking good care of one's children—a quiteplausible idea. Then, parents whoreally do care about their childrens' welfare, i.e., who are‘real’ altruists, will have a higher inclusive fitness,hence spread more of their genes, than parents who only pretend tocare, or who do not care. Therefore, evolution may well lead‘real’ or psychological altruism to evolve. Contrary towhat is often thought, an evolutionary approach to human behaviour doesnot imply that humans are likely to be motivated byself-interest alone. One strategy by which ‘selfish genes’may increase their future representation is by causing humans to benon-selfish, in the psychological sense.

Bibliography

• Abbot, P. et. al., 2011, ‘Inclusive Fitness Theory andEusociality’,Nature, 471: E1–E4.
• Avital, E. and Jablonka, E., 2000,Animal Traditions:Behavioural Inheritance in Evolution, Cambridge: CambridgeUniversity Press.
• Axelrod, R. and Hamilton, W. D., 1981, ‘The Evolution ofCooperation’,Science, 211: 1390–96.
• Axelrod. R., 1984,The Evolution of Cooperation, New York:Basic Books.
• Birch, J., forthcoming, ‘Hamilton's Rule and itsDiscontents’,British Journal for the Philosophy ofScience.
• Bowles, S. and Gintis, H., 2011,A Cooperative Species: HumanReciprocity and its Evolution, Princeton: Princeton UniversityPress.
• Boyd, R. and McIlreath, R., 2007,Mathematical Models ofSocial Evolution, Chicago: Chicago University Press.
• Boyd, R. and Richerson, P., 2006, ‘Culture and the Evolutionof the Human Social Instincts’, in S. Levinson and N. Enfield(eds.),Roots of Human Sociality, Oxford: Berg.
• Bourke, A. and Franks, N., 1995,Social Evolution in Ants,Princeton: Princeton University Press.
• Carter, G. G. and Wilkinson, G. S., 2013, ‘Food Sharing inVampire Bats: reciprocal help predicts donations more than relatednessor harassment’,Proceedings of the Royal Society B,280: 20122573, doi 10.1098/rspb.2012.2573.
• Clavien, C. and Chapuisat, M., 2013, ‘Altruism across disciplines: one word, multiple meanings ’,Biology and Philosophy, 28(1):125–140.
• Cronin, H., 1991,The Ant and the Peacock, Cambridge:Cambridge University Press.
• Darwin, C., 1859,On the Origin of Species by Means of NaturalSelection, London: John Murray.
• –––, 1871,The Descent of Man and Selectionin Relation to Sex, New York: Appleton.
• Dawkins, R., 1976,The Selfish Gene, Oxford: OxfordUniversity Press.
• –––, 1979, ‘Twelve Misunderstandings ofKin Selection’,Zeitschrift fur Tierpsychologie, 51:184–200.
• –––, 1982,The Extended Phenotype,Oxford: Oxford University Press.
• Dugatkin, L. A., 1997,Cooperation among Animals: anEvolutionary Perspective, Oxford: Oxford University Press.
• –––, 2006,The Altruism Equation,Princeton: Princeton University Press.
• Fletcher, J. A. and Zwick, M., 2004, ‘Strong Altruism CanEvolve in Randomly Formed Groups’,Journal of Theoretical Biology, 228: 303–13.
• Fletcher, J. A. and Doebeli, M., 2006, ‘How AltruismEvolves: Assortment and Synergy’,Journal of Evolutionary Biology, 19: 1389–1393.
• Frank, S. A., 1998,Foundations of Social Evolution,Princeton: Princeton University Press.
• Gardner, A., Alpedrinha J. and West S. A., 2012,‘Haplodiploidy and the evolution of eusociality: split sexratios’,American Naturalist, 179: 240–56.
• Gardner, A. and West S. A., 2010,‘Greenbeards’,Evolution, 64(1): 25–38.
• Grafen, A., 1984, ‘Natural Selection, Kin Selection andGroup Selection’, inBehavioural Ecology: an EvolutionaryApproach, J. R. Krebs and N. B. Davies (eds.), Oxford: BlackwellScientific Publications.
• –––, 1985, ‘A Geometric View ofRelatedness’,Oxford Surveys in Evolutionary Biology, 2: 28–89.
• –––, 2006, ‘Optimization of InclusiveFitness’,Journal of Theoretical Biology, 238: 541–63.
• Hamilton, W. D., 1964, ‘The Genetical Evolution of SocialBehaviour I and II’,Journal of Theoretical Biology, 7:1–16, 17–32.
• –––, 1970, ‘Selfish and Spiteful Behaviourin an Evolutionary Model’,Nature, 228:1218–1220.
• –––, 1972, ‘Altruism and RelatedPhenomena, mainly in the Social Insects’,Annual Review ofEcology and Systematics, 3: 193–232.
• –––, 1975, ‘Innate Social Aptitudes inMan: an Approach from Evolutionary Genetics’, inBiosocialAnthropology, R. Fox (ed.), New York: Wiley.
• –––, 1996,Narrow Roads of Gene Land,New York: W. H. Freeman.
• Hammerstein, P., 2003, ‘Why is Reciprocity so Rare in SocialAnimals? A Protestant Appeal’, in P. Hammerstein(ed.)Genetic and Cultural Evolution of Cooperation,Cambridge MA: MIT Press.
• Kerr, B., Feldman, M. W. and Godfrey-Smith, P., 2004, ‘Whatis Altruism?’,Trends in Ecology and Evolution, 19:135–140.
• Lehmann, L. and Keller, L., 2006 ‘The Evolution ofCooperation and Altruism. A General Framework and Classification ofModels’,Journal of Evolutionary Biology, 19:1365–1725.
• Lehmann, L., Keller, L., West, S., and Roze, D., 2007 ‘GroupSelection and Kin Selection: Two Concepts but OneProcess’,Proceedings of the National Academy of theSciences, 104(16): 6736–6739.
• Leigh, E. G. jr., 2010 ‘The GroupSelection Controversy’,Journal of Evolutionary Biology, 23(1): 6–19.
• Marshall, J. A., 2011 ‘Group Selection and Kin Selection:formally equivalent approaches’,Trends in Ecology andEvolution, 26(7): 325–32.
• Maynard Smith, J., 1964, ‘Group Selection and KinSelection’,Nature, 201: 1145–1147.
• –––, 1974, ‘The Theory of Games and theEvolution of Animal Conflicts,Journal of TheoreticalBiology, 47: 209–21.
• –––, 1982,Evolution and the Theory ofGames, Cambridge: Cambridge University Press.
• –––, 1998, ‘The Origin of Altruism’,Nature, 393: 639–640.
• Maynard Smith, J., and Szathmary, E., 1995,The MajorTransitions of Life, New York: W.H. Freeman.
• Michod, R. E., 1982, ‘The Theory of Kin Selection’,Annual Review of Ecology and Systematics, 13: 23–55.
• Nowak, M., 2006,Evolutionary Dynamics: Exploring theEquations of Life, Harvard: Harvard University Press.
• Nowak, M. A., Tarnita, C. E. and Wilson, E. O., 2010,‘Evolution of Eusociality’,Nature, 466: 1057–62.
• Nunney, L., 1985, ‘Group Selection, Altruism andStructured-Deme Models’,American Naturalist, 126: 212–35.
• Okasha, S., 2002, ‘Genetic Relatedness and the Evolution ofAltruism’,Philosophy of Science, 69, 1: 138–149.
• –––, 2005, ‘Altruism, Group Selection andCorrelated Interaction’,British Journal for the Philosophyof Science, 56: 703–24.
• –––, 2006,Evolution and the Levels ofSelection, Oxford: Oxford University Press.
• Queller, D. C., 1992a, ‘A General Model for KinSelection’,Evolution, 46: 376–380.
• –––, 1992b, ‘Quantitative Genetics,Inclusive Fitness, and Group Selection’,AmericanNaturalist, 139: 540–558.
• Rosenberg, A., 1992, ‘Altruism: Theoretical Contexts’,inKeywords in Evolutionary Biology, E. F. Keller and E. A.Lloyd (eds.), Cambridge MA: Harvard University Press.
• Sachs, J. L., 2004, ‘The Evolution ofCooperation’,The Quarterly Review of Biology, 79:135–160.
• Schulz, A., 2011, ‘Sober & Wilson's EvolutionaryArguments for Psychological Altruism: AReassessment’,Biology and Philosophy, 26:251–260.
• Segerstrale, U., 2000,Defenders of the Truth: the Battle forScience in the Sociobiology Debate and Beyond, Oxford: OxfordUniversity Press.
• Skyrms, B., 1996,Evolution of the Social Contract,Cambridge: Cambridge University Press.
• Sober, E., 1988, ‘What is Evolutionary Altruism?’, inNew Essays on Philosophy and Biology (Canadian Journal ofPhilosophy Supp. Vol. 14), B. Linsky and M. Mathen (eds.),Calgary: University of Calgary Press.
• –––, 1994, ‘Did Evolution Make usPsychological Egoists?’, in hisFrom A Biological Point ofView, Cambridge: Cambridge University Press.
• –––, 2011,Did Darwin Write the OriginBackwards?, New York: Prometheus.
• Sober, E. and Wilson D.S., 1998,Unto Others: The Evolution andPsychology of Unselfish Behavior, Cambridge MA: Harvard UniversityPress.
• Sterelny, K., 2012,The Evolved Apprentice,Cambridge MA: MIT Press.
• Taborsky, M. 2013, ‘Social Evolution: Reciprocity ThereIs’,Current Biology, 23(11): R486–88.
• Trivers, R. L., 1971, ‘The Evolution of ReciprocalAltruism’,Quarterly Review of Biology, 46: 35–57.
• –––, 1985,Social Evolution, Menlo ParkCA: Benjamin/Cummings.
• Uyenoyama, M. K. and Feldman, M. W., 1992, ‘Altruism: someTheoretical Ambiguities’, inKeywords in EvolutionaryBiology, E. F. Keller and E. A. Lloyd (eds.), Cambridge MA:Harvard University Press.
• Vromen, J., 2012, ‘Human Cooperation and Reciprocity’,inEvolution and Rationality, S. Okasha and K. Binmore(eds.), Cambridge: Cambridge University Press.
• West, S. A., Griffin, A. S, and Gardner, A., 2007, ‘SocialSemantics: Altruism, Cooperation, Mutualism, Strong Reciprocity andGroup Selection’,Journal of Evolutionary Biology, 20:415–432.
• Wilkinson, G. S., 1984, ‘Reciprocal Food Sharing in theVampire Bat’,Nature, 308: 181–184.
• –––, 1990, ‘Food Sharing in VampireBats’,Scientific American, 262, 2: 64–70.
• Williams, G. C., 1966,Adaptation and Natural Selection,Princeton: Princeton University Press.
• Wilson E. O., 1975,Sociobiology: the New Synthesis,Cambridge MA: Harvard University Press.
• Wilson, D. S., 1977, ‘Structured Demes and the Evolution ofGroup-Advantageous Traits’,American Naturalist, 111:157–85.
• –––, 1980,The Natural Selection ofPopulations and Communities, Menlo Park, CA: BenjaminCummings.
• –––, 1990, ‘Weak Altruism, Strong GroupSelection’,Oikos, 59: 135–48.
• –––, 1991, ‘On the Relationship betweenEvolutionary and Psychological Definitions of Altruism andEgoism’,Biology and Philosophy, 7: 61–68.
• Wilson, D. S. and Dugatkin, L. A., 1992, ‘Altruism:Contemporary Debates’, inKeywords in EvolutionaryBiology, E. F. Keller and E. A. Lloyd (eds.), Cambridge MA:Harvard University Press.

Academic Tools

	How to cite this entry.
	Preview the PDF version of this entry at theFriends of the SEP Society.
	Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
	Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

[Please contact the author with suggestions.]

Related Entries

game theory: evolutionary |natural selection |natural selection: units and levels of |sociobiology