Transcript 03-altruism - Welcome to the EvoS Consortium

Evolutionary Psychology, Lecture 3.
Altruism and Co-operation
Learning Outcomes.
 At the end of this session you should be able to:
 1. Explain what is meant by the terms ‘kin
selection’ and ‘reciprocal altruism’.
 2. Discuss
kin selection and reciprocation
explanations for human and animal cooperative /
altruistic behaviours.
Thoughts For the Day.
 “Let us try to teach generosity and altruism because we are
born selfish” Richard Dawkins (1976).
 “…ethics, morality, human conduct, and the human psyche
are to be understood only if societies are seen as
collections of individuals seeking their own self interest”.
R.D. Alexander (1987).
 “Scratch an altruist and watch a hypocrite bleed”. Ghiselin
(1974).
Altruism.
 Altruism refers to an individual acting in a way that will
decrease its own survival chances, but improve the survival
chances of another individual.
 The Darwinian perspective emphasising ‘survival of the
fittest’ gave the impression that selfishness was the norm.
 Pioneering work involving the study of animals living in
social groups in fact revealed that co-operation and
altruism are just as ‘natural’ as selfishness.
 If co-operation and altruism have evolved, then they must
have some adaptive benefits, researchers have analysed
the conditions under which adaptations for engaging in
such behaviour can be expected to evolve.
Examples of Animal Altruism.
 Vampire bats will regurgitate and
feed blood that they have
collected from their prey to a
hungry conspecific (Wilkinson,
1990).
 Ground squirrels will warn others
of the presence of a predator,
even though making such a call
may draw the attention of the
predator to itself (Sherman,
1977).
 In many species of social insects,
workers
forgo
reproduction
entirely (they are sterile) in order
to help raise their sisters (Wilson,
1971).
Theories of Altruism.
 1. Kin Selection (Proposed by Hamilton, 1964).
 By helping relatives to reproduce (even at the cost to your
own reproductive success) then your shared genes can
spread. Assisting a close relative thereby increases one’s
‘Inclusive Fitness’.
 Using mathematical modelling, Hamilton showed that an
altruistic gene can spread through the population if it
causes an individual to help a relative, whenever the cost to
the individual is offset by the reproductive benefit gained
by the receiver.
 ‘Hamilton’s Rule’ = r B>c
 where r=coefficient of relatedness, B = benefit to the
recipient, c = cost to the giver.
Kin Selection in Action.
 Ground squirrels do not give an alarm call every time a
predator approaches. They only do so when there is a large
proportion of their relatives within earshot (Sherman,
1977).
 Vampire bats are much more likely to share their food with
relatives than with non-relatives (Wilkinson, 1990).
 This theory explained the most puzzling phenomena - that
of the sterile insects - by a genetic quirk they are more
related to their sisters than to their mothers or daughters
(Trivers & Hare, 1976).
Kin Recognition.
 It is important to be able to recognise kin, as the costs
involved in mistaking another individuals offspring for
one’s own are high, and the benefits few.
 Offspring recognition should evolve more often in colonial
species, as there is a high risk of misdirecting parental care.
 Examples.
 Bank swallows (colonial) do not accept strange chicks
whereas rough-winged swallows (solitary) do.
 Herring gulls (colonial ground-nesting) recognise offspring
and refuse strange chicks, but Kittiwakes (colonial cliffnesting) do not recognise offspring and accept substitute
offspring.
Kin Recognition in Gulls.
Data from Alcock, 1993
Kin Selection in Humans.
 Studies amongst diverse human populations consistently
support the existence of kin selection, some examples
(cited in Barrett et al., 2002) are as follows:
 Food sharing is more common amongst close relatives.
 Political alliances between kin are more stable than those
formed between distantly related, or unrelated individuals
and involve less preconditions.
 The passing on of wealth to lineal descendants (excluding
spouses) is far more common than giving to less closely
related or unrelated individuals.
 Close relatives are preferentially sought out in times of
need and such help is less likely to be reciprocal.
 Relatives typically receive more expensive presents.
How Much Pain Will You Suffer
For Your Kin?
 In an interesting experiment Fieldman et al., (cited in
Barrett et al., 2002) asked participants to maintain a
painful position. The longer they held the position the more
money they would earn.
 In different conditions participants could earn money for
individuals differing in relatedness:
 Themselves.
 Parent or sibling.
 Grandparent / niece / nephew, a cousin.
 Unrelated friend.
 The duration of maintaining the painful position varied as a
direct proportion of relatedness, with more pain being
sustained for closer relatives.
Facial Similarity and Trust?
 DeBruine (2002) argued that animals should be sensitive to
cues of genetic relatedness when making altruistic
decisions.
 In humans such decisions may be based around facial
appearance.
 Participants played a computerised game of trust in which
they had to decide whether or not to share money with an
individual.
 They were shown faces of their 'opponents' which were
either facially different to themselves, or whose faces had
been morphed to resemble their own.
 Participants showed significantly more 'trusting' behaviour
when playing against opponents that resembled
themselves.
Human Adoption.
 The adoption of unrelated children has been cited as
evidence against kin selection as helping to rear unrelated
children will not produce genetic benefits to the ‘giver’.
 However, Silk (1990) observed that among Polynesian
cultures, a substantial number of adopters cared for
children who were cousin equivalents or closer. Families
who had adopted children that were unrelated tended to be
agricultural families needing extra help.
 Similarly, in Chicago Stack (1974) reported that the
majority of foster children were adopted by kin.
 Adopting unrelated children is a recent Western
phenomenon. Alcock (1993) argues that the urge to
produce children and look after them is so beneficial in
reproductive terms that it has become deeply ingrained.
Problems for Kin Selection.
 Kin selection does not explain observed incidences of animals
helping non-relatives for example:
 Unrelated chimpanzees come to one another’s aid when
threatened (de Waal & Luttrell, 1988).
 Vampire bats will feed non-relatives (Wilkinson, 1990).
 Humans often engage in apparently altruistic acts such as:
 Giving blood.
 Donating to charity.
 Forgoing reproduction.
 Rescuing unrelated individuals (and even animals).
 Sacrificing their lives for moral or ethical principles.
 How can such behaviours be explained?
2. Reciprocal Altruism.
 Proposed by Trivers (1971).
 Natural Selection may create psychological mechanisms
designed to deliver benefits even to non-relatives, provided
that such actions lead to reciprocal beneficial actions in the
future.
 ‘you scratch my back…’.
 This is not necessarily limited to the same species e.g.
cleaner fish.
 If the benefit received is larger than the cost incurred, then
individuals who engage in such behaviour will outreproduce those who do not.
 Eg, in vampire bats, an individual will share food with a
conspecific (whether related or not) if the other has shared
food with that individual in the past (Wilkinson, 1990).
Conditions Under Which
Reciprocation Flourishes.
 1. Individuals must associate for long-enough periods of
time to develop reciprocal interactions.
 2. The likelihood of one individual performing some social
exchange with another should be predicted on the basis of
their past associations.
 3. The roles of giver and receiver should reverse at least
once.
 4. The short-term benefits to the recipient are greater than
the costs to the donor.
 5. Givers should be able to recognise and expel cheaters
from the system.
 Do such conditions apply to human social interactions?
Modelling Human
Social Exchanges.
 ‘Game Theory’ was developed by the mathematician von
Neumann and the economist Morgenstern in the 1940’s in
an attempt to model the behaviour of individuals in
economic and adversarial situations.
 Maynard-Smith (1982) adapted it to model co-operation
and competition in the social world.
 Within game theory, individuals behave rationally and
choose the action that yields the highest payoff.
 Human social interactions can be mapped into game
settings, as one individual can benefit from the actions of
others at little cost to themselves.
 There are several forms:
‘Unscrupulous Diner Scenario’.
 A group of diners agree to divide the restaurant bill equally,
most co-operate by choosing similar priced meals but an
individual can take advantage by ordering the most
expensive meal, as the cost will be absorbed by the whole
group (Glance & Huberman, 1994).
 In a one-off situation in a large social group it pays to
cheat, however in a small group who meet regularly, such
defection will be noticed and punished.
 Reciprocal social exchange has mutual costs/benefits but
one person can always benefit more than another if they
cheat - i.e. receive an act but do not reciprocate.
 This constitutes a formidable barrier to the evolution of
social exchange.
Prisoner’s Dilemma.
 Described by Axelrod & Hamilton (1981). It is a game in
which mutual co-operation benefits both players, but a
‘cheat’ can gain a higher pay-off.
 It is often described as a hypothetical situation in which
two individuals have committed a crime, and are being held
for questioning in separate cells, they are unable to
communicate.
 It is in the best interests of both to say nothing, as the
evidence is such that both may only receive a light
sentence.
 However, they are being questioned separately, and the
lawyer offers both freedom if they implicate the other in
the crime.
 When the game is played once or only a few times then the
best strategy is to inform on the other (defect).
Prisoner’s Dilemma Payoff Matrix
Player B
Co-operates
Player B
Defects
Player A
Co-operates
R=3 year sentence
each.
Reward for mutual
cooperation
S= long sentence
for A, freedom for
B.
Sucker's payoff
Player A
defects
T= freedom for A,
long sentence for B.
Temptation to
defect
P=10 year sentence
each.
Punishment for
mutual defection
Which Strategy is Best?
 Real life social exchanges may occur more than once and
individuals will probably remember those who have defected
(cheated) in the past.
 Axelrod (1984) organised a tournament in which various
computer programs played repeated prisoners dilemma
games. A program called 'tit for tat' was a clear winner. This
simple strategy always cooperated on the first move, and
then copied its opponents move on every subsequent move.
 This is an ‘Evolutionary Stable Strategy’ (ESS) - i.e. once
established it cannot be displaced by another strategy.
 In real life it pays to cheat in a one-off exchange encounter
but if there is a likelihood that you will encounter the same
person more than once, then mutual cooperation will serve
both parties the best.
Indirect Reciprocation.
 Trivers (1971) argued that an altruistic act need not
necessarily be reciprocated by the person directly assisted
but can be returned indirectly from other individuals.
 E.g if you advertise yourself as an altruist then individuals
will be more favourably inclined to deal with you in future
social exchange situations.
 This may explain blood donation, giving to beggars, and
donating to charities.
 It had been claimed that such actions indicate that human
behaviour is immune form evolutionary analysis and
demonstrates a pure form of altruism.
 Alexander (1987) suggested that giving blood is a very
good way of demonstrating your altruism at only a modest
cost.
Evidence for Indirect Reciprocation.
 We would maybe predict that individuals will not donate
blood or give to charity unless their actions are made
known – i.e. by wearing a sticker or badge indicating their
actions.
 Low & Heinen (cited in Alcock, 1993) reported that
students are significantly more likely to give to charity if
they receive a pin or tag that advertises their participation.
 Mulcahy (1999) observed who gave money to beggars and
then interviewed mixed-sex couples after the male had
donated money.
 Males at the early stage of a relationship were more likely
to give than those in a long-established relationship.
 Thus the act of giving when accompanied by a female is not
simply to do with impressing a current partner, but
demonstrating one's generosity to a potential partner.
An Alternative Theory.
 Gintis et al., (2003) argues that kin selection and
reciprocity theories do not explain why cooperation is
frequent amongst unrelated individuals in non-repeated
interactions when gains are small.
 Strong reciprocity is the predisposition to cooperate with
others, and to punish those who violate the norms of
cooperation, at some personal cost, even when such costs
may not be repaid.
 In support, Fehr & Gächter (2002) showed that when asked
to play a game for monetary reward under 'no punishment'
or 'punishment conditions‘, punishment of non-cooperators
substantially increased the amount that groups invested for
the good of the group.
Additional Evidence For
Strong Reciprocity.
 The concept of 'fairness' lies at the heart of many human
social interactions and can be modelled using the
'Ultimatum game'.
 Here a participant is given a sum of money and told they
can keep it provided that they split the sum with another
individual.
 The participant has to make a one-off offer between 0 100% of the total sum to the other person.
 If the second person agrees to the offered sum then both
keep these amounts; if they reject the offer then both
receive nothing. No haggling is allowed.
 According to one-off game theory exchanges we would
expect that the first participant would offer a sum of well
below 50% and that the receiver should accept any sum as
anything is better than nothing.
The ‘Ultimatum Game’ in Action.
 However when this game is played it is typically found that
individuals offer around 50%, and more than half of
receivers do not accept offers less than 20% (Sigmund et
al., 2002).
 Individuals do not behave completely selfishly but place a
high value on fair outcomes. Receivers are prepared to
accept smaller 'gifts' under the following conditions:
 The giver is chosen by better performance on a quiz.
 The givers offer is randomly selected by a computer.
 Several responders compete to accept a proposers offer.
 Sigmund et al., (2002) proposed that our emotional
apparatus has been shaped over millions of years of small
group living in which it is hard to cheat more than once and
where we expect conspecifics to notice our actions and
remember them.