Transcript presentation

From social psychology to
sociology - a physicist’s point
of view
Katarzyna Sznajd-Weron
Institute of Theoretical Physics
Wrocław University
Praha, November 6, 2003
Plan
Can we explain global changes by
microscopic models?
Can we treat people like particles?
Examples of social experiments.
Social psychology and sociology.
A simple model.
Almost a century ago physicists asked
the question:
Can phase transitions be explained by
microscopic theory?
M
ferromagnet
paramagnet
T*
T
Back to 1920 ...
Wilhelm Lenz proposed a very simple
microscopic model of interacting spins.
Si  1
H   J  Si S j  h  Si
i, j
J 0
i
Was it only hope?
H   J  Si Si 1  h  Si
i
i
Spontaneous magnetization
cannot be explained using this
model in its 1D version.
After two decades
hope became reality
Onsager showed that the 2D version of the
model can explain the critical phase transition.
Very simple local
interactions can lead
to qualitative changes
on the macroscopic
scale.
M
T*
T
The hope outside Physics
Rapid changes on macroscopic scale appear
in various systems.
Mostly these changes are unexpected.
Usually there is no obvious reason for them.
Can we explain them in terms of microscopic
interactions like we did for physical systems?
People sometimes behave
like particles
Conformity
The Millgram experiment:
Obedience to Authority
The Asch experiment:
peer pressure
Social validation
Conformity
- Obedience to Authority
The consequences
of nonconformity
Stanley Millgram: Obedience to Authority
(Yale, 1961-62)
Stanley Milgram (1973) Journal of
Abnormal and Social Psychology
volts
learner
experimenter
90
Ou!
Please continue
105
Ou! (lauder)
It is essential that we continue....
120
Ou! It hurts!
Continue, it’s necessary
135
Ou! It is really painful The experiment requires that you
go on until he has learned all the
word pairs correctly
150
Let me leave! My
heart ...
You have no other choice
What will happen? – ask psychiatrists
and psychologists
Predicted: most subjects would not go
beyond 150 volts, when the victim makes
his first explicit demand to be freed.
Only 4% would reach 300 volts.
Only a pathological fringe of about 0.1%
would use the highest shock on the board.
Results of the Millgram experiment
Solomon E. Asch - a pioneer
of social psychology
Solomon E. Asch, born in
Warsaw in 1907, he came to
the United States in 1920 and
received a Ph.D. from
Columbia University in 1932.
Experiment of conformity
(without Authority), 1956
Asch(1951-1955) – the experiment
on visual perception?
The experimenter asks to choose which of the
three lines on the left matches the length of the
one on the right.
Results of the Asch experiment: social
pressure
76% conformed to the majority at least once.
Conformity became more frequent as group
size increased.
“The tendency to conformity in our society is
so strong that reasonably intelligent (...)
people are willing to call white black” /Asch/
There are some limits ...
40%
Mistakes
30%
20%
10%
0%
1
2
3 4 5 6 7 8
Number of people against
9
10
The power of social validation
Milgram, Bickman
& Berkowitz, 1969
Results of experiments:
1  4%, 5  80%
Robert B. Cialdini:
Social Validation –
the fundamental way
of decision making
Social validation can cause trouble
BANK
From social psychology to sociology
Social Psychology
Fundamental unit:
A person
(micro scale)
Sociology
Fundamental unit:
A social group
(macro scale)
Social opinion
Sociologists ask if it will be better in the future?
0.5
0
-0.5
09.97
06.98
Time
04.99
02.00
Social Norm: The ratio between breast
and waist (Vogue)
2
B/W
1.8
B
W
100/60
1.6
1.4
78/60
1.2
1910
1930
1950
Year
1970
1990
From micro to macro scale ...
Person = Spin (element of the system)
Social validation = interaction between
elements
Social opinion = magnetization
A model based on social validation
YES = +1
NO = -1
What to do if you do not know what
to do ...
„United we stand divided
we fall” (original rule)
Nothing (Stauffer et al.)
Whatever (financial market)
How does it work?
A sample simulation
Number of simulation
steps
10000
1000
100
10
0
0
20
40
60
Voters
80
100
Evolution of the system: social
opinion (Yes-No)
1
m
N
N
S
i
i 1
What has
happened?
m
1
0.5
Time
We follow one person ...
number of decision changes
14
12
10
8
6
4
2
0
0
20
40
60
80
100
time interval (100 MCS)
120
Characteristic time of opinion change
does not exist!
2
10
simulations-1.5
powerfit ~ t
0
P(t)
10
-2
10
-4
10
0
10
1
10
2
t
10
3
10
What happens if we sometimes turn
off the auto-pilot?
8
Distribution
10
6
10
4
10
2
10
0
10 0
10
1
10
2
10
Waiting time
1 p
3
10
Si Si 1  1 Si 1  Si , Si 2  Si 1
There is some external field in every
social system
1
h
For h=1
-1
Advertising: with probability |h| buy product sgn(h).
Who wins ?
Probability of “conquering”
the market
1
0.8
0.6
0.4
c0=0.05
c0=0.15
c0=0.25
0.2
0
0
0.2
0.4
0.6
0.8
Power of advertisement (h)
1
Generalized model - two components
(TC)
Dynamics - the information flows outwards
Disagreement function - the change of
spins is controlled by function, which locally
is minimized:
F   J1Si 1Si  J 2 Si 1Si 1
Possible transitions
 J1  J 2
J1  J 2
 J1  J 2
 J1  J 2  J1  J 2  J1  J 2  0
J1  J 2   J1  J 2  J 2  0
J1  J 2
USDF
Phase diagram for 1D TC model
1
J2
0
-1
-1
0
J1
1
TC model in 2D
Phase diagram for 2D TC model
1
Phase 2 (1)
0
2
3
4
5
6
Phase I
(1,2,3,4)
Phase 4 (2)
J2
1
Phase 3
(5,6)
-1
-1
0
1 J1
You never know which state you will
reach …
J1=1, J2=2
How to predict the future?
 J1  J 2
1/8
 J2
1/2
 J2
1/4
J1  J 2
1/8
Local agreement – global
disagreement
-1.5
disagreement function
-2
-2.5
-3
-3.5
-4
-4.5
-5
-5.5
-6
0
500
1000
1500
2000
time (MCS)
2500
3000
Social norm as an external field?
Social opinion
Social norm
Time to say „goodbye”