Transcript Ginges ARO - School of Computer Science

Sacred Values as cultural
factors in collaboration and
negotiation
Scott Atran (CUNY, University of Michigan)
Jeremy Ginges (New School for Social Research,
NYC)
Baruch Fischhoff (Carnegie Mellon University)
Consultant: Robert Axelrod (University of
Michigan)
Team
• Social Psychology (Ginges),
Anthropology (Atran), Judgement and
decsion-making (Fischhoff), Political
science (Axelrod)
• Ethnically and linguistically diverse team
of graduate students
• Languages: Arabic, Urdu, Hindi,
French, German, Polish, Hungarian
•
Sacred Values
Research broadly examines the influence of sacred
values (SVs) on judgement and decision-making
•
•
Sacred values are things (issues, resources, beliefs,
practices) that are infused, in the minds of specific
groups of people, with moral, spiritual and often
religious components
•
•
•
•
“I would not ___ no matter how great the benefit or cost”
Land becomes “sacred”, buildings become “holy”
Product of “moral communities” (Tetlock, 2003) and so
culturally bound
Moral intuitions (Haidt);Strong ties to the affective
system
We suppose strong ties to individual identities
• EarlySacred
research by Phil Tetlock and Jon
Values
Baron
•
Tetlock (2003) argued that SVs are merely
“pseudo-sacred “and in a real world of scarce
resources they are remarkably easily dodged to
produce flexibility - we dispute this
• Judgments and decisions about “sacred
values” differ from judgment and
decisions about mundane (marketplace)
objects and behaviors
•
Taboo against measuring commitments along
an instrumental scale/metric
• Significantly impact collaborative versus
What we know
• Judgments and decisions over sacred
values tend not to be instrumentally
rational (lack cost-benefit characteristics)
and tend to be influenced by
instrumentally irrelevant but emotionally
laden symbolism
• This basic finding published in leading
peer review journals PNAS, Science,
Psychological Science
• No framing effects (Tanner & Medin,
2004; Ginges 2007), insensitivity or
hypersensitivity to scope (Bartels &
Two other findings
i. When people have a strong
preference for _____ (nonabsolutists), offering material
sweeteners leads to less
anger/violence
ii. When people hold _____ to be a SV
(moral-absolutists), material
sweeteners backfire
★
However symbolic gestures work
well
Better deals can produce worse
results
•Sample: 720
Palestinians in West
Bank and Gaza.
•Recruited across 14
campuses, individual
interviews
•Half members of
Hamas or PIJ
•Two experiments
compared reactions of
moral absolutists versus non
absolutists
Measuring SVs
•
•
•
•
•
“Do you agree that there are some extreme
circumstances where it would be permissible for the
Palestinian people to compromise over ______ ….”
Yes
Don’t Know/Unsure
No indicates SV - moral absolutist
Don’t know/unsure
- non absolutist
No
Recognizing Israel
•
•
Taboo deal: “Suppose the United Nations
organized a peace treaty between Israel
and the Palestinians. Under this treaty
Palestinians would recognize the sacred
and historic right of the Jewish people to
Israel. There would be two states - a Jewish
state of Israel and a Palestinian state in
99% of the West Bank and Gaza.”
Taboo+: “On their part, Israel will pay
Palestine 1 billion dollars a year for 10
years.’
Violence
Within subjects design
•
•
•
102 Indonesian
students at 4
madrassahs
One of the four
schools, al-Islam,
associated with
Jemmah Islamiyah
Deals involved US
recognition of
Islamic party rule,
compromise over
sharia
•
trade
agreement
Indonesia
Replicable finding
•
•
•
Other samples: Jewish settlers, Palestinian
refugees, Indonesian madrassah students
Other topics: right of return (refugees),
giving up land (settlers), sharia
Other material incentives: more money, life
free of violence
Value of symbolic
concessions
•
•
Surveyed rep
sample of settlers
in August ‘05,
days before Gaza
disengagement
Approx 50% of
those surveyed
were moral absolutists
regarding the
“Land of Israel” to
be a SV
Violence: Settlers
•
•
Taboo: “Israel would
give up 99% of Judea
and Samaria. Israel will
not absorb ANY
refugees. There would
be two states - a Jewish
state of Israel and a
Palestinian state.”
+ Symbolic: “On
their part, Palestinians
will give up any claims
to the “right of return” which is sacred to
them”
Characteristics of
research
• Not just students
but participants in
actual conflicts
• Not made-up “toy” hypotheticals but real
problems important to our participants
• Product of development of research
partnerships over long term
•
Challenges of this type of research (trust,
developing of real collaborations with trained
or trainable local partners; dealing with
suspicious populations
•
Needs
We know that judgments and decisions over SVs
are different to those over material values
•
•
•
Yet we have little empirical knowledge over
purported social identity functions, or whether the
category “SV” hides a complex taxonomy of SVs
While we have done some cross cultural
replication -Indonesia, India - this is still in relative
infancy
Need to develop more precise predictions and
policy recommendations concerning cross-cultural
collaboration/negotiation when SVs are implicated
•
need to investigate SVs in real time
experiments (online, simulations) in addition to
hypotheticals
Key goals
1.Investigate identity functions & taxonomy of
SVs across cultures (1st Wave)
2.Develop more precise descriptive theory for
methods of achieving cooperation in contexts
involving clashing or threatened SVs (2nd
Wave)
•Research sites: Palestine, Turkey, Morocco*
• * Possibility of extending to Lebanon, India, Indonesia,
Egypt
Data Collection
Strategy
• Staggered phases of data collection
• Begin each wave in Palestine, followed by other
sites
• Wave 1 of data collection will focus on defining
SVs of interest within different communities &
developing taxonomy
• Subsequent waves include studies on
judgement/decision-making in cross cultural
negotiations
Key goals
1.Investigate identity functions of SVs
• Group identity part of self identity (Turner, Tajfel), to the extent
that things are important to group identity they tend to be
valued more (Ledgerwood, Liviatan, Carnevale, 2007)
• SVs may be essential to group identity thus moral
commitments and non-instrumental behavior
• Hypothesis: SVs may play key roles in specific proximate
identity mechanisms that lead to non-instrumental behavior
in defense of the collective (essentialism, entitativity and
sharp group boundaries).
• We will probe whether different types of SVs are associated
with these three mechanisms and thus with different aspects
of social identity
Investigating identity functions of
• We will proceed inSVs
4 phases
1. Semi-structured interviews. Participants will be asked
to nominate values, ideas and practices that are
important to them as ______, and that distinguish their
group from other relevant groups. We will then probe
content domain, consequences (cognitively available
information)
2. Moral Categorization Use of spontaneous sorting of
scenarios into kinds to infer how SVs are
conceptualized.
3. Identity function probes. Based on the interview and
sorting tasks we will generate a set of SVs and directly
probe their identity functions using scenarios (e.g.,
adoption task)
4. Population surveys
Key goals
2.Develop more precise descriptive theory for
methods of achieving cooperation in contexts
involving clashing or threatened SVs
•
•
We know that reasoning over SVs differs from
reasoning over material values; clear implications
for collaborations/negotiations across cultures
We need to know more about the precise
circumstances that influence outcomes in crosscultural interactions that implicate SVs
Achieving cooperative outcomes
in interactions implicating SVs
•
During second wave of data collection
•
Specific questions to be investigated:
1. Specific emotional mediators (anger/humiliation)
2. Domain importance
3. Apologies
4. Order effects
5. Psychological distance
6. Asymmetries in mutual mis/perceptions in moral
universes
?
a. Modes of symbolic/moral compromise. Symbolic
compromises may work best when they are in the same
domain as the SV trade-off our participants are being
asked to make.
Domain importance
•e.g., if participants are being asked to trade-off an-in
group distinctive SV, the adversary must compromise the
same type of SV for their symbolic compromise to be
most effective.
b. Violation - reaction. Outrage at violation of SV may
depend on domain (e.g., what type of identity is being
threatened? to what extent). While reaction to violation is
likely dictated (somewhat) by specific cultures, but
perhaps also by type of violation.
Apologies
•Apologies. Pilot research shows that people tend to judge their
own morality through introspection (“what were my intentions”), but
the morality of others through observable behaviors (“what did they
do?”) (See Pronin, 2008).
•Discontinuities like these might prove barriers in the use of
apologies to facilitate political compromises. People may apologize
to others by claiming a lack of intention, but expect the apologies of
others to focus on effects of wrongdoings.
•Studies will present scenarios dealing with wrong-doings (in one
set of studies) and apologies (in another set). Scenarios will vary
between subjects: (a) whether the subject is in- or out-group; and
(b) whether focus is on intentions of behaviors. Rate
morality/immorality of subject & efficacy of apologies. Will probe for
emotional reactions.
Order effects
• In Ginges et al (2007) symbolic adversarial compromises
decreased opposition to compromise over SVs. In these studies,
compromise was simultaneous.
• In the real-world compromises are often sequential. Ginges &
Malhotra (2003): in between group interactions initial compromise
often backfires
• We will examine the effect of sequential compromises, such as a
symbolic (e.g., Nasser visiting Jerusalem) or material compromise
(e.g., offering to increase economic opportunities) on willingness to
compromise over SVs.
•
Psychological distance
Unclear how psychological distance (time/space) will influence
cross-cultural interactions that implicate SVs
•
On the one hand, greater psychological distance leads to
more abstract construals (e.g., desirability) than concrete
construals (e.g., means ends) (Trope, Liberman - multiple
studies); thus greater distance may lead to greater activation
of SVs
•
On the other, riming perceptions of psychological distance
(e.g., spatial difference) tends to decrease subsequent
affective reactions (e.g., feelings of connectedness with family
and hometown; Williams & Bargh, in press).
•
In one set of studies we plan to prime spatial distance prior to
measuring aversive reactions over SVs.
•
In a second set of studies we will directly manipulate spatial and
temporal distance by asking people to imagine they are negotiating in
the distant versus near future or in a spatially distant or spatially close
location.
•
•
Another Q: are SVs a product (or cause) of long term thinking?
If these experiments are effective when run with the normal
population, we will replicate them in interviews with leaders and
militants