|
Running head: The emotional dog does learn
new tricks
The emotional dog does learn new tricks
(A reply to Pizarro and Bloom, 2003)
Jonathan Haidt University of Virginia
March 6, 2002
This is the final word processing file of
the document published as: Haidt, J. (2003).
The emotional dog does learn new tricks:
A reply to Pizarro and Bloom (2003). Psychological
Review, 110, 197-198.
Abstract Pizarro and Bloom (this issue) argue
against the social intuitionist model of
moral judgment (Haidt, 2001) and for a modified
rationalist model. This reply responds to
their two main arguments by demonstrating
that an intuitionist model allows for malleability
and flexibility in judgment, and it allows
for cases of moral deliberation.
The emotional dog does learn new tricks (A
reply to Pizarro and Bloom, 2003)
A critique and reply are most productive
when the authors on both sides share some
areas of agreement and a sense of mutual
respect. I am very fortunate that the present
exchange seems likely to be productive. Pizarro,
Bloom and I all agree on the big picture
of moral psychology: the field badly needed
updating. A field that was a major part of
the cognitive revolution needed to be brought
forward through the affect revolution of
the 1980's, and into the age of automaticity
and dual process models that began in the
1990's (e. g., Bargh,
1994).
We also agree with the insight of Nisbett
and Wilson (1977) that people often don't
know or can't explain why they feel what
they feel, but they frequently make up explanations
anyway. The Kohlbergian method of taking
verbal justifications of moral judgments
at face value is therefore suspect. Pizarro
and Bloom, however, believe that I went too
far in my updating (Haidt, 2001). They defend
a modified rationalist theory in which people
have intuitions, but these intuitions merely
serve as a "starting point for deliberative
reasoning." In other words, it is "the
rational dog that wags the emotional tail,
not vice-versa." Pizarro and Bloom make
two major claims in support of this position:
1) Fast and automatic moral intuitions are
shaped and informed by prior reasoning, and
2) people actively engage in reasoning when
faced with real-world moral dilemmas. To
respond to these claims I would like to correct
two common misunderstandings of the social
intuitionist model. The first is that intuitionism
implies that we are prisoners of our initial
intuitions, unable to change our minds once
we have taken a position.
But as I will show, intuitionism allows for
a great deal of malleability and responsiveness
to new information and circumstances. In
other words, the emotional dog does learn
new tricks. The second misunderstanding is
that intuitionism means that we do not have
times of deep moral reflection. Educating
the moral intuitions Pizarro and Bloom point
out that emotions and intuitions depend upon
cognitive appraisals - on the facts that
one believes about the case at hand. They
point out that changing the facts of a situation
(e. g., discovering that a student's absence
was due to a death in the family) changes
the emotional response and the moral judgment
that one makes. The social intuitionist model
is built on this insight, and the model is
quite explicit that moral judgments change
when a situation is suddenly viewed in a
new light and new intuitions are triggered.
But how does it come about that people experience
new cognitive appraisals?
Pizarro and Bloom suggest that this is a
common occurrence in private deliberative
reasoning. I can only refer again to the
empirical research on reasoning, which shows
that people rarely search on their own for
evidence on both sides of an issue (Perkins,
Farady, & Bushey, 1991; Kuhn, 1991).
Rather, the social intuitionist model stresses
the importance of social interaction as the
best way to trigger new appraisals: "Yet
ever since Plato wrote his Dialogues, philosophers
have recognized that moral reasoning naturally
occurs in a social setting, between people
who can challenge each other's arguments
and trigger new intuitions (links 3 and 4)"
(Haidt, 2001, p. 820).
It is noteworthy that the empirical studies
of appraisal change that Pizarro and Bloom
cite all involve participants being told
by an external agent to take another person's
perspective. A second way in which Pizarro
and Bloom suggest that intuitions get altered
by reasoning is by second-order control strategies.
Even if we can't fully control our emotions
in the moment of judgment, we can choose
(rationally) to give ourselves experiences
that, over the course of months or years
will shape our intuitions and change our
judgments. Again I completely agree. The
social intuitionist model says that deliberately
socializing with people with particular values
should over time change one's own values
via both the reasoned persuasion link and
the social persuasion link. However once
again I believe that such deliberate efforts
to go against our moral intuitions are rare
in practice. The authors make the analogy
to cases of conflict between first-order
and second-order desires, e. g., wanting
to not want ice cream or a cigarette.
But how often do such conflicts arise in
our moral lives? Does it ever happen that
a person has the gut feelings of a liberal
but a second order desire to become a conservative?
If so, the person could set out on a several-year
program of befriending nice, articulate,
and attractive conservatives. But are such
cases common enough to support Pizarro and
Bloom's modified rationalist approach? I
believe such people are rare, and most of
us would treat them with suspicion. Pizarro
and Bloom discuss a study by Rudman, Ashmore,
and Gary (2001) in which Yale students who
enrolled in a class on racism showed less
implicit stereotyping at the end of the course.
But how many of these people enrolled in
order to exercise second order control over
their implicit stereotypes? I believe a greater
attraction was the joy of morally meshing
with other people who are outraged by racism.
We rarely seek out challenges to our moral
world-views, and a course on racism seems
like an excellent place to avoid the threat
of moral diversity (Haidt, Rosenberg, and
Hom, in press). Secondary control strategies
may be important in dieting and in smoking
cessation, but I do not think they play a
significant role in moral development.
The scope of moral reasoning Finally, Pizarro
and Bloom argue that when we look outside
the lab at people facing real-world moral
dilemmas, we see clearer evidence of people
using deliberative reasoning. It is certainly
true that interviews with people facing gripping
dilemmas, such as those conducted by Gilligan
(1982) and Coles (1986), reveal people agonizing
over a decision and fully aware of arguments
on both sides. But what exactly can we infer
about psychological processes from such studies?
Do we know that the protagonists actively
searched for arguments on both sides? Do
we know that they made their decision by
weighing the strength of the arguments and
acting on their logical entailments?
The social intuitionist model easily handles
such cases moral agonizing, but it handles
it in a different way from rationalist models.
Real-world dilemmas involve real people with
real claims on one's loyalties. These people
cannot be ignored, and they make it difficult
to simply endorse one solution and then close
the case (as often happens with hypothetical
dilemmas). A pregnant woman whose first concern
is for her own future may want an abortion,
but each time she sees or thinks about other
relevant people (her partner, her parents,
or babies that make her think of the fetus
in her own womb) she may feel strong and
conflicting emotions.
Such a process is a kind of emotional deliberation,
moving back and forth through the claims
and perspectives of the parties involved.
The social intuitionist model captures this
process as repeated cycles through the main
loop of the model, with different people
triggering different and conflicting intuitions.
The model also allows that such cycling can
in theory happen in the privacy of one's
own head (link #6, the private reflection
link) as one spontaneously takes the perspective
of others. But once again we must ask: how
common is this process? I believe we do not
at present have evidence that would allow
us to answer this question. Until somebody
does the relevant beeper study I can only
ask the reader to make a mental list of how
many times he or she has agonized over a
moral issue in the past year, and has gone
back and forth in his or her judgment. Now
compare that to an estimate of the total
number of moral judgments the reader has
made in the last year, while reading the
newspaper, participating in gossip, or driving
on roads surrounded by drivers less competent
than oneself.
My prediction is that for most people, the
first number is less than one hundredth of
the second number. And even in cases that
felt like deliberative reasoning, Nisbett
and Wilson (1977) warn us that we may have
just been making up reasons post hoc for
the strong and conflicting intuitions that
arose within us on both sides. Conclusion
"The Emotional Dog" was written
before the events of September 11, 2001.
The attacks on the United States and the
subsequent war in Afghanistan gave us an
opportunity to examine reasoning in the presence
of strong gut feelings. Americans were faced
with an urgent need to understand the motives
and methods of terrorists, and to reach decisions
on a range of moral issues such as ethnic
profiling, restrictions on civil rights,
and the justifiability of civilian deaths
in Afghanistan. How did Americans resolve
these questions? Did they seek out evidence
on both sides? Did they engage in role taking
(Kohlberg, 1969), viewing the attacks from
the attackers' perspective? No.
For several months afterwards, powerful emotions
of horror and outrage made many Americans
actively hostile towards anyone who engaged
in role taking (e. g., the comedian Bill
Mahar, who said that it was incorrect to
call the terrorists cowards). In those difficult
months the social part of the social intuitionist
model was very much in evidence as Americans
sought to unite with each other in condemning
evil, not to understand the real motives
and causes of terrorism. Pizarro and Bloom
are correct that moral intuitions may in
theory be a starting point in moral deliberations,
and that conscious moral reasoning can be
used to build upon those initial intuitions.
We disagree, however, on how often this happens,
and therefore about whether moral reasoning
is best modeled as the dog wagging an emotional
tail, or as the tail wagged by the emotional
dog. I hope that we will work together to
update moral psychology, and in the process,
figure out how the dog works.
|
Evans Experientialism 
