Cognitive Adaptations for Social Exchange

1
Cognitive Adaptations for Social Exchange
Leda Cosmides and John Tooby
presented by Nat Twarog
2
Overall Statements of Paper

• The human mind does contain content-specific
  machinery adapted for a certain class of problems
  or situations
• One such group of mechanisms exists for social
  exchange
• These mechanisms apply specifically to social
  explained problems and are designed to detect
  cheaters, i.e., those who take a benefit
  without paying the cost required by the social
  contract

3
What about the Tabula Rasa?

• According to the Standard Social Science Model
  (SSSM) all innate neural mechanism are general
  purpose, non-content-specific
• Things like learning, induction,
  rationality etc.
• All content, and content specific processing, is
  learned from the environment, and derives from
  one or more of these general purpose mechanisms
• Tooby and Cosmides claim this view is held as
  sacred without justification

4
Justifications

• Why should there be any content-specific evolved
  machinery at all?

• Human culture could not exist without them

• Certain social interactions are so integral to
  human life and so universal, the mind would have
  to adapt to them

• Why shouldn't there be?

5
Why Social Exchange Theory?

• Theory is well understood, both from an
  game-theoretic and evolutionary perspective

• Social exchange is universal, happens every day,
  and has been part of everyday human life for a
  long, long time

• Specialized machinery, if it exists, is affecting
  human reasoning, which is considered one of the
  most fundamental aspects of cognition

• Basically, the idea is, if it can happen here, it
  can happen anywhere

6
What is Social Exchange?

• Also called reciprocal altruism

• You help me out, and I'll help you next time
• Ex hunter-gatherers

• But there's a problem with altruism
• CHEATERS

7
Experimental Design
Rule If P, then Q.
P
Q
not P
not Q

• Subjects are given background or context, and are
  told a rule of the form If P, then Q.
• In most experiments, subjects are then asked to
  which cards must be checked for violations of the
  rule
• Logically correct answer is P not-Q

8
First Experiment

• Give every subject several Wason selection tasks

• One task uses a rule concerning actions or
  possessions which are novel, but which is clearly
  a social contract

• Other tasks are a familiar rule which is not a
  social exchange contract, a novel rule which is
  realistic but not a social exchange contract, and
  an abstract rule

9
Results
10
Changing your Point-of-View

• Run Wason selection task with the rule If an
  employee works 10 years, then he receives a
  pension

• Half of subjects are told they are the employer,
  and half are told they are the employee

• According to social exchange theory, the
  definition of cheating depends on viewpoint, so
  answers should be different

• If subjects are simply thinking logically, they
  should answer the same

11
Results
12
Other Findings

• Mechanisms do not detect non-cheating violations
  as well as cheating, including mistakes and
  genuine altruism
• Performance only improves if situation has clear
  costs and benefits, rather than just permission
  schemas
• Performance decreases when subjects are less
  likely to see benefit as a benefit

13
Conclusions

• Content dependent, situation-specific mechanisms
  do exist, adapted for the purpose of detecting
  cheaters in a social exchange

• These mechanisms clearly demonstrate that some,
  if not many, of the mechanisms for human
  cognition are evolved and highly specialized for
  tasks

• These mechanisms, due to their universality and
  uniformity, are probably innate

14
(No Transcript)