Basic Assumptions of Evolutionary Theory

1
Basic Assumptions of Evolutionary Theory

• There are heritable variations in traits (i.e.,
  either a physical characteristic such as brain
  size, height or a psychological characteristic
  such as sociability, selfishness, generosity,
  aggresiveness and intelligence).
• In particular environments some traits contribute
  more to an individuals fitness (i.e., survival
  and reproduction) than others.
• As a result these traits are positively selected
  and increase in frequency. In a word, they
  become adaptations.

2
Basic Assumptions of Evolutionary Psychology

• Human thought, feeling and action reflect
  adaptations or traits that evolved over the past
  5,000,000 years when the human line separated
  from that of chimpanzees and bonobos, our closest
  primate relatives.
• Adaptations are modular (e.g., vision, language).
  But to what extent? How distinct an entity
  (e.g., lungs vs jealousy vs sociality)?
• Sociality (i.e., group living), is a key human
  adaptation.
• The costs and benefits associated with our
  peculiarly extensive and complex networks of
  social relations are the primary source of
  selection pressures on humans.

3
Humans Compared to Chimpanzees and Bonobos

• What adaptations or traits distinguish humans
  form their nearest primate relatives? What do
  these adaptations imply about human social
  psychology?
• Large brains
• Long periods of juvenile dependence
• Extensive parental care including the transfer of
  vast amounts of information
• Multigenerational bilateral kin networks

4
(And thats not all! Theres )

• Habitual bipedal locomotion
• Cryptic or concealed ovulation
• Menopause
• Culture, including language
• Letal competition among kin-based coaltions
• N.B. A few other species exhibit some of these
  adaptations. However, only humans possess the
  entire set of them in their most complex form.

5
The Adaptive Value of a Trait Depends on Its
Contribution to Fitness

• The ultimate and most direct measure of fitness
  is the number of healthy offspring or
  reproductive success (RS). Thus, a traits
  benefits refers to how much it increases RS and
  its costs, how much it decreases RS.
• We often use less direct or more proximal
  measures of fitness that we assume contribute to
  RS (e.g., as health, strength, wealth) for
  convenience.

6
Why Do We Think Group Living Is an Adaptation?

• Because it is universal.
• Because it is typical of species most closely
  related by common descent to humans.
• Because it has neurophysiological correlates
  (e.g., neocortex ratio and social network
  density ostracism and activation of pain area
  in brain).
• Because it has affective correlates (e.g.,
  isolation and ostracism are painful and universal
  punishments while being liked and respected are
  pleasurable and universal rewards).
• Beause it has cognitive correlates (e.g., Theory
  of Mind cheater-detection modules).
• Because bio-economic analyses of fitness (i.e.,
  benefits to RS relative to it costs) suggests
  living in groups is adaptive.

7
Do species most closely related to humans
(chimps and bonobo) live in groups? Yes. Are
chimp groups and bonobo groups similar? No. So
what?
8
Are there neurophysiological correlates of group
living? Yes. Neocortex size increases with
group size social complexity.
9
Bio-economic Analysis

• In examining sociality as an adaptive strategy
  Richard Alexander considers the recurrent
  problems faced by groups in the ancestral
  environment and compares the hypothetical fitness
  costs and benefits of increasing group size
  from living with several conspecifics (e.g., in
  separate nuclear families), through a few dozen
  (e.g., nuclear family coalitions or extended
  families), one or two hunderd (hunter-gatherer
  groups), to thousands or millions (towns, cities,
  clans, tribes and nations).

10
The cost/benefit return of increasing groups
size Minimizing home, den or nest site shortages
as an adaptive problem
11
The cost/benefit return of increasing group size
Minimizing disease as an adaptive problem
12
The cost/benefit return of increasing group size
Minimizing food shortages (when food is widely
distributed, thus, readily found) as an adaptive
problem
13
The cost/benefit return of increasing group size
Minimizing food shortages when food sites are few
and hard to find as an adaptive problem
14
The cost/benefit return of increasing group size
Minimizing food shortages when food is large,
hard to catch animals (prey) as an adaptive
problem
15
The cost/benefit return of increasing group size
Minimizing the danger of predation
16

• Protection from predation provides the largest
  benefit to fitness from living in groups during
  most of human evolution.
• However, humans have achieve ecological dominance
  so that weve not had to fear predation by other
  species for the past 15-20,000 years. Even
  before then, predation by other species was
  minimized at a relatively small group size
  compared to the size of human clans, tribes and
  nations.
• 3. So why do we live in such large groups?!

17
If humans are not the prey of other species, do
they still suffer predation?

• Yes, of course! We are our own prey! Predation
  by other human groups is likely to have been a
  big adaptive problem for our species!!
• (N.B. Our overlooking this suggest how the
  ecological dominence of modern humans biases our
  thinking about the ancestral environment?)

18

• For many thousands of years the most
  significant predator on group living humans has
  been other group living humans. To explain how
  this could cause humans live in very large
  groups, Alexander proposed the balance of power
  model

19
The Balance of Power Model

• i. In multi-group environments, the members
  felt vulnerability is inversely related to the
  relative size of their group.
• Felt vulnerability motivates smaller groups to
  form coalitions whose size counter-balances or
  excedes that of the previously largest group.
• As a result, felt vulnerability decreases among
  members of the newly formed coalition (and
  increases among members of the previously largest
  group).
• This in turn motivates the latter also to seek
  coalition partners which, if successful, motives
  the former to seek further coalition partners
  etc. Thus, group size spirals upward to some
  limit where there is a balance of power that
  minimizes feelings of vulnerability and
  additional coalitions are too costly or
  unavailable.

20
Living inKin Groups is Easier to Explain Than
Living in Non-Kin Groups

• The protection function of groups implies
  altruism Group members willingly incur large
  costs to benefit others (e.g., some will risk
  death to protect fellow members from an animal
  predator or a raiding group).
• Until the second half of the last century the
  fact of altruism, was a puzzle. How could such
  tendencies evolve if they cause harm to the actor
  and should be selected against? In 1964 Hamilton
  showed how in his analysis of inclusive fitness
  (kin selection).

21
A Heuristic for Thinking about Hamiltons Theory

• Imagine you are a gene that contributes to the
  trait of intelligence. You know that
• An individual is your vehicle carries you
  through life.
• Intelligence contributes to fitness (increases
  RS).
• The probability that copies of you exist in
  relatives of your vehicle increases with their
  degree of relatedness to your vehicle.
• Then answer the following question
• What what strategy would you want your vehicle
  to follow if your goal is insure that copies of
  you continue to exist in future generations?

22
Hamiltons Inequality Solves Two Related
Problems Why Living in Kin Groups is Adaptive
and How Altrusim Can be Positively Selected
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
Risky Decisions Involving a Large Non-Kin Group,
a Small Non-Kin Group, or My Family

• The next study indicates that when making a risky
  decision for a group, the size of the group and
  our ties to its members can cause us to behave
  seemily irrationally, i.e., compute costs and
  benefits in sub-optimal fashion.
• In what sense are such behaviors irrational?
• According to behavioral decision theory or
  inclusive fitness theory or both?
• Does this consider that adaptations are designed
  for recurrent problems, not rare events.

30
Framing of Choices in the Tversky and Kahneman
(1981) Decision Task

• The decision task
• Imagine that Lodz is preparing for the outbreak
  of an unusual disease which is expected to kill
  600 people.
• Two alternative programs to combat the disease
  have been proposed.
• Assume that the exact scientific estimates of the
  consequences of the programs are as follows

31
Positive framing of the Decision Task

• The Certain outcome. If program A is adopted 200
  people will be saved.
• The Uncertain outcome. If program B is adopted,
  there is a one-third probability that 600 people
  will be save and a two-third probability that no
  people will be saved.

32
Negative framing of the Decision Task

• Certain outcome. If program C is adopted, 400
  people will die.
• Uncertain outcome. If program D is adopted,
  there is a one-third probability that nobody will
  die and a two-third probability that 600 people
  will die.

33
Tversky and Kahnemans results

• Under positive framing of the decision people are
  risk averse
• 72 of their respondents chose the certain
  outcome.
• 28 of them chose the uncertain or probabilistic
  outcome.

34
Tversky and Kahnemans results

• Under negative framing people are risk prone
• 22 chose the certain outcome.
• 78 chose the probabilistic or uncertain outcome.

35
Peculiar Parameters of theTversky-Kahneman
Decision Task (Wang, 1996 2002)

• The Group is Large and Its Members Have No Ties
  to the Respondent.
• What Do You Think Would Happen If the Group is
  Small and the Decision Makers Ties to the Group
  Are Strong?

36
Risk Proneness Decreases with Group Size and
Kinship
37
Risk Aversion is Sensitive to Survival Rates for
Non-Kin Groups But Not for Kin Groups (Choices
are Positively Framed)
38
Decisions about Kin Groups Violate Rational
Choice Preferences Under Positive Framing for
Probabilistic Outcomes When Its Expected Value is
Less than that of the Certain Outcome
Choice Percentage
Non-kin
Kin
Proportion of Group Saved for Certain
(Note In all conditions the probalistic outcome
is 1/3rd chance of saving all group members)
39
Members of Kin Groups Are Nice to Each Other
But Not Under All Conditions

• Parental investment hypothesis (derived from
  inclusive fitness theory) argues parents should
  incur a cost to benefit a child when it
  contributes to parents inclusive fitness more
  than doing something else with their resources.
  If so, what should be predicted (think of the
  earlier heuristic)
• 1. When parents decide on investing in a male
  versus a female offspring?
• 2. When parents are rich versus when they are
  poor?
• 3. When they are step-parents?

40
Recall Hamiltons Inequality

• Note that it says under certain conditions
  altruism toward kin may actually decrease
  inclusive fitness. This happens when
• r 0 and/or C gt r B
• As relatedness (r) between donor and recipient
  decreases and the cost of altruism (C) increases
  the donor should act in an increasingly
  unaltruistic manner. The next slides summarizes
  research (Wilson Daly, 1998) comparing the
  likelihood of child abuse and child homocide,
  decidedly unaltruistic acts, in families with two
  biological parents and families with one
  stepparent (typically the father). It
  emphatically supports Hamiltons prediction.

41
(No Transcript)
42
(No Transcript)
43
(No Transcript)
44
However a Violent Father or Brother Did Have
Benefits

• When a relative was murdered, Vikings had the
  choice between a revenge murder or accepting
  blood money. Berserkers were individuals with a
  reputation of being extremely fierce and
  dangerous. If a murderer was a berserker (or his
  father or brother), the aggrieved relatives of
  the victim were significantly more likely to
  accept blood money, but to prefer a revenge
  killing if the murderer was not a berserker or a
  close relative.
• In the next slide the plotted variable is the
  ratio of observed murders relative to the number
  expected on the basis of the proportion of
  berserkers or non-berserkers in the population.
  Source 34 murders recorded in Njal's Saga.

45
(No Transcript)
46
Viking berserkers suffered significantly higher
rates of mortality at the hands of their own
community but their behavior benefitted male
members of their families. Therefore,
Berserkers were altruists, yes? (Families of the
three berserkers in the Icelandic Njals Saga
suffered significantly less mortality than the 7
families that did not contain a recognized
berserker.)
120 100 80 60 40 20 01
47
Something to Think About

• From the 8th through the 10th century, the
  Vikings were the fiercest and most feared group
  in Europe, raiding and plundering settlements on
  the northern and western coasts of the continent
  as well as the interior of eastern Europe.
• A dozen centuries later their descendents are the
  most peaceful and least feared group in Europe.

48

• In light of Hamiltons theory, how could
  cooperation among non-kin evolve?

49
Non-kin Altruism, Cooperation and Equity Some
Questions for Later

• 1. Have the benefits of cooperation
  sufficiently outweighed its costs (transaction
  costs and opportunity costs) to create selection
  pressures on human psychology?
• 2. Can large cooperative networks (e.g.,
  markets, trading networks) function without
  cognitive adaptations that allow participants to
  calculate the risks of a transaction?

50

• 3. Are there indirect benefits of non-kin
  altruism (e.g., giving money to charity to poor
  strangers)?
• 4. Are costly acts with no return benefit
  (e.g., altruistic punishment in a one-shot
  prisoners dilemma game) more a matter of
  satisfying a need for equity or fairness than
  true altruism? Or a need for vegence? If so,
  how would such motives be positively selected?

51
Skinnerians and others of their ilk say
Altruism need not assume the operation of
cognitive adaptations like cheater-detection,
empathy or Theory of Mind

• A radical behaviorist demonstrates that helping a
  stranger develops and is maintained because it
  the act of helping is reinforced by its
  consequences.
• Hence assumptions about cognitve adaptations are
  theoretically unnecessary, a violation of
  scientific parsimony.

52

• In the following experiment subjects are free to
  press a button as quickly as they want to record
  the end of a trial. In two conditions this also
  turns off a noxious noise piped into a strangers
  ears i on every trial (continuous
  reinforcment), ii on some randomly selected
  trials (partial reinforcement), or iii on none
  of the trials (control), where the noise ends
  automatically after a fixed interval. The desire
  or effort to help is indicated by how quickly the
  subject presses the button.
• N.B. The stranger is a confederate of the
  experimenter and there is no actual noise being
  piped into his ears.

53
(No Transcript)
54
A Question to Ponder

• If humans do persist in helping strangers and if
  they do so because the act is intrinsically
  reinforcing, where does that leave theories that
  assumes complex computations of costs and
  benefits plus discounting (e.g., for age, health,
  relatedness, etc.) are necessary for the
  evolution of such behavior?

55
THE STRUGGLE BETWEEN BEHAVIORISTS AND
COGNITIONISTS CONTINUES IS COOPERATION MINDLESS
OR DO YOU NEED COGNITIVE ADAPTATIONS?

• Sidowski Cooperation in essentially
  coordinating interpersonal behavior and can be
  achieved when individuals are totally unaware
  they are interacting with another person. Just
  assume the Law of Effect or Win-Stay,
  Lose-Change and forget about complex
  computations.
• Kelly Not true. You need to think, to take the
  others perspective and think about what they are
  thinking to achieve cooperation. Let me show
  you..

56
The Sidowski-Kelley Coordination Game
57
The results of applying the Win-Stay-Lose-Change
rule when both players, P and O, respond
simultaneously. Note that there are only three
combination of button-press choices possible on
the first trial and from then on the
Win-Stay-Lose-Change rule determines each
players outcomes
58
The Three Starting Combinations

• a. P chooses the right button and O the left.
  As a result both receive positive outcomes.
• b. Both P and O choose the left button. As a
  result P receives a positive outcome and O a
  negative outcome. (N.B. The case in which both
  choose the right button is isomorphic with this
  one.)
• c. P chooses the left button and O the right.
  As a result both receive negative outcomes.

59
Win-Stay, Lose-Change Wins An Ambiguous Triumph
for Radical Behavior Theory
60

• According the Win-Stay-Lose-Change rule,
  cooperation (mutual reward) is inevitable under
  conditions of simultaneous responding which turns
  out to be the case whether or not P believes O is
  another person or a computer.
• But see what happens when one simple parameter is
  changed, i.e., P and O respond in alternation
  rather than simultaneously. Suppose O responds
  first and P second. The three starting trials
  are as follows (continuing to apply the
  Win-Stay-Lose-Change rule as before)

61
Win-Stay Doesnt Win An Unambiguous Triumph for
Social Cognition
62

• Unless they start out cooperating (a mutually
  rewarding exchange), they should never achieve
  cooperation according to the Win-Stay-Lose-Change
  rule.
• But they do achieve cooperation if P knows O is
  another person (not, say, a computer). How is
  this possible? Well, not by using Skinnerian
  reinforcement theory which is inadequate to
  explain this effect. We have to go elsewhere
  for an explanation. Where? What assumption do
  we need to make beyond those of reinforcement
  theory to explain how cooperation is possible in
  a mutual-fate-control sitution when we know our
  outcome depends on another person, a stranger,
  and his or her outcome on us?

63

• New Assumptions TOM, Foresight and Planning
  Cooperative Thinking
• To achieve cooperation under these conditions we
  foresee having to adjust our strategies and
  actions with those of others so as to reconcile
  potential conflicts of interests with maximum
  benefit or least cost. To do so we represent in
  our mind, as best we can,what others intend to do
  (their plan or strategy), and what they think we
  intend to do (our plan or strategy).
• Why?
• In order to decide whether others are
  trustworthy.
• In order to anticipate and, thus, coordinate each
  othersactions, thereby achieving a mutually
  beneficial or least costly relationship (e.g.,
  reciprocity or division of labor).

64

• The most common experimental paradigms for
  observing cooperative thinking is the two-person
  prisoners dilemma game (PDG) and the n-person
  prisoners dilemma game (SDG).

65
The Classic PDG
66
The Social Dilemma Game (SDG)

• The SDG is an n-person (n gt 2) PDG. Say, as the
  next Table assumes at least 5 out of a group of 7
  members have to contribute their endowment a
  sum of 5 given them at the beginning of the
  experiment to fund a public good. The latter
  means that all members will benefit by receiving
  10 whether or not they incurred the cost of the
  public good, i.e., whether they were a
  contributor or a non-contributor. So like all
  public goods, all members, contributor and
  non-contributors benefit if the group meets the
  cost criterion. Do you see why this creates a
  conflict of interest similar to that in the PDG?

67

• As in the PDG the largest benefit or payoff goes
  to defectors (i.e., non-contributors) if there
  are enough cooperators (i.e., contributors) to
  provide the public good.
• The next largest goes to the cooperators if
  enough others cooperate to provide the public
  good.
• The next largest goes to defectors if there are
  not enough cooperators to provide the public
  good.
• And least benefit, the suckers payoff goes to
  cooperators when there are too few to provid the
  public good.

68

• Next we lay out the conditions that define the
  standard SDG
• It is a non-iterated (one-shot) game.
• Members are strangers.
• Their decisions are completely anonymous.
• There is no contact or discussion prior to,
  during or after the decisions. They arrive and
  leave the experiment never have seen any member
  of their group.

69

• Non-standard versions devised to compare with the
  standard SDG
• Money-back is a norm imposed on the group that
  guarantees cooperators will get their money back
  if there are two few of them to provide the
  public good, ergo, no one gets a suckers payoff
  and looses his endowment.
• No free-riders is a norm imposed on the group
  that guarantees defectors will not benefit more
  than cooperators if the public good is provided,
  ergo, there is no temptation to defect.

70
The standard SDG and variations in Caporeal,
Dawes, Orbell and van de Kragt (1989)
Cooperation requires a minimum number of members
to contribute.
71

• Some SDG studies also vary
• Whether or not individuals have a brief
  discussion prior to deciding anonymously.
• Whether or not the experimenter designated who
  was to contribute (but they could still defect
  because their decision is anonymous).
• Whether or not everybody had to contribute
  (called super simple because members did not
  have to decide about cooperating but again anyone
  could still defect since their decision is
  anonymous).

72
Rates of Public Goods Provision
73
Rates of Contribution when External Authority
Designates the Anonymous Contributors
74
Intergroup Cooperation Is Distrust of the Other
the Default for Outgroups (Even Minimal
Outgroups)?
75
Reciprocity Knowing and Providing What is Due
Another Uncompelled Equity and Fairness

• 1. Will a person abide by a contract when it
  is costly to do so and the person cannot be
  punished for defecting?
• 2. Will a person punish defectors when it is
  costly to do so and it cannot force them to
  cooperate?
• 3. Will a person expect to receive punishment
  as a result of defecting when punishment is
  costly to adminsiter and it cannot benefit the
  punisher (by compelling cooperation)?
• If you say yes to any or all of these
  propositions what does it imply about equity and
  fairness as an adaptation?

76
Employees Contracted and Delivered Effort in
One-Shot (non-repeated) Employer-Employee
Gain (see Gintis, et al., 2003)
Employees Average Effort
Payoff Offer to Employee by Employer
77
The Mystery of Altruistic Punishment

• Cooperation can be maintain by punishing
  free-riders.
• Humans seem designed to punish non-cooperators in
  that they do so even when it is costly and there
  is no direct return benefit (e.g., in a one-shot
  exchange).
• If punishment of free-riding is costly and cannot
  elicit return benefits from the free-rider, how
  can it be postively selected and become an
  adaptation? iveEven when it is costly to them and
  they do not directly benefit as a result (e.g.,
  in a one-shot game)?

78
Contributions in a repeated public goods game
with the same partner or a new partner (stranger)
under conditions of punishment and non-punishment
(see Gintis, et al., 2003)
Punishment option Removed
Punishment Permitted
Average Contribution
Games
79

• Milgrams study of obediance is the best known
  study with data about what people expect someone
  to do when punishing another person. At first
  glance, the findings seem to argue against
  assuming the tendency to punish free-riders is an
  adaptation.
• But does it? Is the person being punished
  free-riding? If not, does the finding imply
  anything about tendencies to punish in the
  absence of free-riding? Let look at Milgrims
  data.

80
Punishing Members Who Refuse to Punish Deviants
May Be Unnecessary to Produce Conformity
Predictions that People (including Self) Will
Refuse to Punish Deviant Learner Are Wrong.
81
(No Transcript)
82
Cooperation and the Division of Labor

• The division of labor is probably the most common
  form of cooperation in everyday life.
• It is occurs when (i) members know, prefer to or
  can do different things and (ii) coordinate their
  respective knowledge or performances to their
  mutual benefit.
• The division of labor can be not only formal,
  explicit and hierarchical (e.g., military units,
  sports teams, surgical teams, business teams,
  etc.) but also informal, implicit and egalitarian
  (e.g., families, friends, co-workers and lovers).

83
Cooperation Depends on Trust

• Contrary to Axelrod, simulation studies
  demonstrate that his best game strategy,
  TIT-FOR-TAT, really isnt. Actually, it counts
  for little in respect to inclusive fitness
  compared to the partner-selection strategy, i.e.,
  being able to distinguish between trustworthy and
  untrustworthy partners ahead of time.
• If so, then humans, being so eminently
  cooperative and cooperation being so vulneralbe
  to cheating, must be designed to detect potential
  cheaters somehow. You agree, of course? Well
  okay, what do we know about such mechanism?

84
The DOG Partner Selection Algorithm

• The secret to DOGs success
• 1 Unlike the other partner selection
  strategies, when it assesses the trustworthiness
  of a potential partner DOG ignores transaction
  costs it doesnt care whether a player
  cooperated or defected in prior transactions.
• ii Instead it focuses only on opportunity
  costs it tries to select the player offering the
  highest potential return and never selects a
  player one offering a negative return regardless
  of whether the player previously defected or
  cooperated.

85

• How does DOG work?
• 1. On the first trial DOG assigns a random
  preference rank to all other players.
• 2. From the second trial on, DOG assigns a
  preference rank to all the other players
  according to the following X-value rule
• a. For any player DOG has ever played in the
  past, the X-value is the score DOG earned in the
  most recent transaction with that player (X can
  vary from some positive value to some negative
  value, i.e., it can reflect a large, moderate or
  small positive or negative return from the
  transaction).
• b. In the case of a stranger, a player with
  whom DOG has never played, X is the average of
  the positive X-values of the players with whom
  DOG has played in the past.
• 3. On each trial DOG first selects the player
  with the largest X-value.
• 4. If that player doesnt select DOG as a
  partner within three matching rounds, DOG selects
  the player with the second largest X-value.
• 5. This process continues until all the players
  with positive X-values are exhausted at which
  point DOG returns to the player with largest
  X-value that is still available and repeats the
  whole process, ad infinitum.

86
Evidence for mechanism to assess the likelihood
of defecting, cheating, free-riding and
untrustworthiness
87
(No Transcript)
88
Abstract Rule Example

• Rule If you are in category X you have to be
  taller than 6.0 feet. Is the rule being
  violated?
• Card P Someone who is in category X.
• Card not-P Someone who is in category Y.
• Card Q Someone who is 6.5 feet.
• Card not-Q Someone who is 5.5 feet.

89
Social Norm Example

• Norm If you are drinking alcohol you have to be
  21 or older. Is it being violated?
• Card P Someone is drinking a beer.
• Card not-P Someone is drinking a coke.
• Card Q Someone is 23 years old.
• Card not-Q Someone is 17 years old.

90
What Makes Us More Trustworthy?

• H.L. Menken (Its) the little voice inside
  of you that says someone is watching. Which
  implies concern about
• 1. Being monitored.
• 2. Reputation.
• 3. Opportunity costs (i.e., other members
  reject you as a partner in transactions involving
  trust).
• 4. Other kinds of punishment (e.g., make
  him an offer he cant refuse ).

91
Concern about being monitored can implicit and
automatic in transactions involving trust
Computer Monitor (Haley Fessler, 2005)
92
Probability of Allocation Funds to an Agent to
Invest in the Trust Game
Effects of Eyes and Crowd-Noise on Allocation
N 24
N 24
N 29
N 22
N 25
(Haley Fessler, 2005)
93
Cues to Trust and Distrust

• Aside from Cosmides cheater-detection mechanism
  she demonstrated using the Wason Selection task,
  are there other situational or internal cues
  besides reputation, payoff structure (e.g.,
  temptations to defect in PDG) and transparency of
  return (e.g., rice versus rubber markets) that
  are used to compute or infer trustworthiness?
• 1. Self-resemblance Facial self-morphing
  (conscious and unconscious effects).
• 2. Facial prototypes Defector recognition
  (specific features, e.g., eye shape?).
• 3. How you feel (mood) Oxytocin inhalation.
• 4. Brain activity Anticipation of returns.

94
Whom Do You Trust? Self Resemblance Studies
Using Facial Morphs
95
Plt.08
Plt.01
Plt.001
NS
Plt.01
Experimental Control
Experiment 1
Experiment 2
Experiment 3
Experiment 4
96
Effects of Qxytocin on Investor Transfers with
Human (Trust) and Programmed (Risk)
Trustee (Kosfeld, Heinrichs, Zak, Fischbacher,
Fehr, 2005)
97
Neural Correlates of Reputation Building in
Trustee Brain (King-Casas, Tomlin, Anew, Camerer,
Quartz, Montague, 2005)
98
Human Adaptations for Cheater Detection
Information Processing and Computation Under
Focused and Unfocused Distrust

• In many situations we know who may be tempted to
  deceive us. Sometimes, however, we do not.
• Focused distrust refers to occasions when we
  suspect deception and know the source.
• Unfocused distrust refers to occasions when we
  suspect deception without being aware of its
  source something just doesnt seem right.
• Have we evolved to process information under
  these two conditions differently compared to when
  we trust others?

99
Cheater-detection makes us think more

• If we elaborate on and analyze a lot what
  suspected cheaters say, then we should confuse
  what their actual statements with inferences we
  made while encoding them.
• Examples of types of inferences
• Direct inference Her boss says Mary works
  quickly and doesnt make mistakes and we infer
  Mary is an efficient worker.
• Compound inference Her boss says Mary works
  quickly etc., and we give a bonus to our most
  efficient workers. We infer Mary won a bonus.

100
(No Transcript)
101
(No Transcript)
102
Favorable
Neutral
Favorable ? Unfavorable
Unfavorable ? Favorable
Unfavorable
103
(No Transcript)
104
Suspicion Can Influence Judgment Unconsciously

• In recent experiments a few seconds before they
  make a judgment, individuals are subliminally
  primed with a word (e.g., an adjective or noun)
  presented imbedded in a supraliminal honest or
  dishonest face.
• The word prime as well as the face is irrelevant
  to the judgments they are about to make (e.g., Is
  a second word, presented above threshold, an
  adjective or a noun?).

105

• The model tested in these experiments assumes
  they will elaborate congruent associates to the
  subliminal word prime in the honest-face context
  and incongruent associates to the subliminal word
  prime in the dishonest-face context.
• If so, the model predicts that
• when the prime and the to-be judged word are
  both nouns or both adjectives, then in the
  honest-face context individuals will elaborate
  congruent associates to the prime (e.g., the
  concept of noun or specific nouns when both the
  prime and the supraliminal word are nouns) and
  categorization of the to-be-judged word is
  facilitated (e.g., faster) whereas in the
  dishonest-face context they elaborates
  incongruent associates (e.g., the concept of
  adjective or specific adjectives when both words
  are nouns) and categorization is disrupted (e.g.,
  slower).
• by the same logic, when the to-be-judged word is
  incongruent with the prime (e.g., one is a noun,
  the other an adjective), judgments will be
  disrupted (e.g., slower) in the honest-face
  context and facilitated in the dishonest-face
  context.

106
(No Transcript)
107
(No Transcript)
108
(No Transcript)
109
(No Transcript)
110
(No Transcript)
111
Being able to efficiently process information
about our relations with others (e.g., Is he or
she a friend or foe? Is his or her status higher
or lower than mine?) is useful. Is there
evidence that humans are designed to make and
store such computations?
112
Adaptations Mechanisms for Coding Social
Relations

• If humans are designed to live in groups, then
  they are also likely to be designed to code
  (i.e., recognize, interpret, remember and
  elaborate upon) information that reduces the
  costs of group living and increase its benefits.
• Among the most adaptive pieces of information
  concern relations among group members
• Who in the group have common interests, are
  friends, who have conflicting interests, are
  enemies?
• Who is has high status (i.e., is powerful, rich,
  skillful, etc.), who has low status (i.e., is
  weak, poor, inept, etc.)?

113
(No Transcript)
114
(No Transcript)
115
(No Transcript)
116
(No Transcript)
117
(No Transcript)
118
A Procedure for Assessing Cognitive
Categorization of Individuals Are They Perceived
as Belonging Together, Forming a Coalition or
Unit?
119
Suppose You Wanted to Know If Observers Grouped
People Based on Common Interest or Opinion.

• Tell observers about some characteristic of the
  people (e.g., sex, age, race, clothes, and what
  they said that indicated whether they were pro
  or con regarding an issue).
• Next ask observers to recall if an individual
  made a particular statement (e.g., whether he or
  she said X).
• Count how often the individual is confused with
  another (e.g., observers say he or she made the
  statement when it was actually made by another).

120

• What was the reason for these confusions? Did
  they occur most often if the two individuals were
  the same sex, the same age, the same race, wore
  the same t-shirt or had the same opinion (both
  were either pro or con)?
• Intra-category confusion are most frequent.
  Therefore, if confusion occurred most often
  between those with the same opinion (both were
  pro or both con), then its evidence that
  observers were categorizing or cognitively
  grouping the individuals based on common opinions
  or common interests. Kurzban calls this
  coalitional thinking. In Heider, the
  grouping would reflect a positive unit
  relationship and, perhaps implicitly, a
  positive sentiment relationship.

121
(No Transcript)
122
(No Transcript)
123
Social Dominance Orientation Following Ethnic
Prime (Questions on Ashkenazi-Sephardi Relations)
or National Prime (Questions on
Israeli-Palestinian Relations)
124
(No Transcript)
125
Consensus and Conformity

• Intentional cooperation typically is a form of
  collective action (e.g., several friends decide
  to form a study group, a community decides to
  subsidize medical care for the poor a
  hunter-gatherer tribe decide to migrate a
  country decides to go to war) based on consensus.
• An effective consensus must prevent cheating and
  free-riding.
• This implies social psychological mechanisms to
  minimize distrust and insure conformity.
• i. Informational influence (e.g., Sherif)
• social comparison
• persuasive argumentation
• ii. Normative influence (e.g., Asch)
• extrinsic reward or punishment

126
But first, the mystery of compliance and human
nature

• What conclusion do you draw about human nature
  from the amazing (okay, amazing only to the
  naïve) relationship between (i) predicted and
  actual compliance and (ii) proximity of victim
  to punisher and the amount of pain the latter
  is willing to inflict on the former?
• Is there other evidence (e.g., studies of
  soldiers ordered to shoot someone)?
• Implications for normative influence and
  conformity and second order punishment?

127
(No Transcript)
128
(No Transcript)
129
(No Transcript)
130
(No Transcript)
131
(No Transcript)
132
The Benefits and Cost of Conformity

• You dont have to learn what to believe or how to
  do something. Just imitate others actions and
  conform to their beliefs.
• Social comparison as an adaptive mechanism for
  imitation and conformity.
• But imitation and conformity may have opportunity
  costs, i.e., you will not discover that there are
  better ways of doing something or that there are
  more valid views of the world.
• Hence, humans may be designed to discount the
  validity of others actions or beliefs to the
  extent that these actions and beliefs are
  themselves products of conformity (not
  independently arrived at) and are not objectively
  demonstrable.
• Is this why conformity in Asch-like normative
  influence settings doesnt increase when
  unanimous majority becomes relatively large?

133
Why Normative Influence Peaks at a Very Small
Unanimous Majority

• 1. Discounting mechanisms
• i. Majority has shared interests different
  from that of deviant.
• ii. Non-independence of majority members.
• 2. Futile search for independent evidence or
  objective demonstration of the majority choice.
• Search is typically done under time stress and at
  the cost of cognitive inconsistency (i.e., the
  majority seems incorrect) relative to the cost of
  social rejection.

134
Conformity to a Unanimous Blame-the-Mother-Not-th
e-Manufacturer Majority in a One Six Member
Group, Two Three Member Groups and Three Two
Member Groups
10K 9K 8K 7K 6K 5K 4K
Degree of Blame
135
Conformity to a Unanimous Blame-the-Mother-Not-th
e-Manufacturer Majority in Groups and in
Non-Groups (Aggregates of Individuals) of
Identical Sizes
Degree of Blame
136
(No Transcript)
137
(No Transcript)
138

• According to Mueller and Mazur (1997), we
  automatically judge someons status or dominence
  from his or her face (Mueller and Mazur, 1997).
• The faces used by Mueller and Mazur are from a
  yearbook published by West Point, the U.S.
  Military Academy, that trains career army
  officers. Some examples (Can you detect
  differences in facial dominance?)

139
(No Transcript)
140

• See slides based on WHO data for male-female
  differences in mortality as a function of status
  competition.

141
Mating

• What does sexual selection theory predict about
  male-female difference in
• 1. Preferred number of partners?
• 2. Probability of consenting to intercourse?
• 3. Preferred age difference in mate?
• 4. Importance of mates provisioning prospects?
• 5. Importance of mates attractiveness?

142
(No Transcript)
143
(No Transcript)
144
(No Transcript)
145
(No Transcript)
146
(No Transcript)
147
Male-Female Differences in Antecedents and
Consequences of Homicide Demographic Evidence

• How (if at all) might theories of sex differences
  in mating strategies, especially their
  implications regarding competition between and
  within the sexes, explain the differences in the
  following data sets?

148
Risky Competition Age- and sex-specific homicide
rates in Canada, 1974-1983.
Female victims
Homicides per million persons per annum
Age (years)
Female offenders
Age (years)
149
(No Transcript)
150
Age- and Sex-specific Rates of Homicide in
Detroit, 1972. (From Wilson Daly, 1985)
Male victims
Homicides per million persons per annum
0-4 10-14 20-24 30-34
40-44 50-54 60-64
70-74 80-84 85 5-9
15-19 25-29 35-39
45-49 55-59 65-69 75-79
Age (years)
Male offenders
0-4 10-14 20-24 30-34
40-44 50-54 60-64
70-74 80-84 85 5-9
15-19 25-29 35-39
45-49 55-59 65-69 75-79
Age (years)
151
Age- and Sex-specific Rates of Homicide in
Detroit, 1972. (From Wilson Daly, 1985)
Female victims
0-4 10-14 20-24 30-34
40-44 50-54 60-64
70-74 80-84 85 5-9
15-19 25-29 35-39
45-49 55-59 65-69 75-79
Homicides per million persons per annum
Age (years)
Female offenders
0-4 10-14 20-24 30-34
40-44 50-54 60-64
70-74 80-84 85 5-9
15-19 25-29 35-39
45-49 55-59 65-69 75-79
Age (years)
152
(No Transcript)
153
Unemployment Rates Among Male Homicide Offenders,
Male victims, and the Male Population-at-Large in
Detroit, 1972. (From Wilson Daly, 1985)
154
Proportions Unmarried Among Male Homicide
Offenders, Male victims, and the Male
Population-at-Large in Detroit, 1972. (From
Wilson Daly, 1985)
155
Spousal Homicde Rates as a Function of the Age
Difference Between Wife and Husband. Canada,
1974-1983.
Wife older
Wife younger
156
Motive Categories and the Number of Cases
(Victims) Within Each, for 588 Criminal Homicides
in the City of Philadelphia, 1948-1952. (From
Wolfgang, 1958)
157
Two hundred Twelve Closed Social Conflict
homicides in Detroit, 1972, in Which Victim and
Offender Were Unrelated (Friends, Acquaintances
or Strangers), Classified by Conflict Typology
and by the Sexes of the Principals. (From
Wilson and Daly, 1985)
158
Dispositions of Spousal Homicides in Various
Studies. (Data from Canada and Detroit are from
Daly Wilson, 1988 for Miami from Wilbanks,
1984 and for Houston from Lundsgaarde, 1977)
159
The Probability of Suicide After Homicide, in
Relation to the Sexes of Killer and Victim, and
Their Relationship, Canada, 1974-1983
160
Intergroup Relations

• 1. Realistic Group Conflict Theory (Sherif)
  versus Social Categorization Theory (Tajfel)
  Are they incompatible?
• 2. The minimal intergroup situation Is
  advantaging the ingroup (or disadvantaging
  outgroups) the default reaction to social
  categorization? Is strategy likely to have been
  adaptive (positively selected for) in the
  ancestral environment
• 3. What about N-group (not merely one in-group
  and one outgroup) environments and coalitions as
  in the balance-of-power model?
• 4. Group/category membership, the hierarchy of
  groups/categories, and self-evaluation Are
  there dimensions other than prestige or power for
  ordering groups?