Emotions and the Interpersonal Self 341relationships. Idealization was associated with greater satis-
faction and happiness about the relationship. A follow-up
study (Murray, Holmes, & Griffin, 1996b) found that favor-
able views of one’s partner predicted greater stability and
durability of the relationship. This research suggests that per-
haps love should be blind (or at least nearsighted enough to
wear rose-colored glasses when looking at the loved one).
The authors argue that idealized love is not blind, but instead
farsighted; partners who idealized each other created the re-
lationships they wanted. Idealization and positive illusions
about one’s partner seem to strengthen the relationship, mak-
ing it more pleasant and more likely to last. Seeing the real
person beneath the facade is not always the beginning of real
intimacy: Sometimes it is the beginning of the end.
These somewhat discrepant results do at least agree that it
is quite important for people to believe that their friends and
lovers appreciate their good points. It is less clear whether
people want their partners to also see their faults and flaws.
One possible explanation for the discrepant results of the two
authors is that most of Swann’s self-consistency work has
emphasized traits that the person is highly certain of and
committed to having. On the other hand, Murray’s favorabil-
ity effects tend to emphasize a broader spectrum of less cer-
tain traits. People might want their close relationship partners
to recognize one or two favorite faults but otherwise maintain
a highly favorable view of them.
There is also intriguing but preliminary evidence that rela-
tionship partners can help sustain consistency. Swann and
Predmore (1985) gave people feedback that was discrepant
from their self-views and watched how they and their roman-
tic partners responded. When the subject and his or her part-
ner agreed that the feedback was wrong, the pair then joined
forces to reject it: They discussed its flaws and decided how
best to refute or dismiss it. In contrast, when the partner’s
view of the subject differed from the subject’s self-view,
the discrepant feedback led to further disagreements between
the subject and partner. It may be that one vital function of
close relationship partners is to help maintain and defend
one’s self-concept against the attacks of the outer world (see
also De La Ronde & Swann, 1998).
Sedikides, Campbell, Reeder, and Elliott (1998) explored
another important link between self-deception and the inter-
personal self. They examined theself-serving bias,a classic
pattern of self-deception that occurs when people take credit
for success but deny blame for failure. When people work in
groups, the self-serving bias produces the tendency to claim
all the credit for success at joint tasks but to dump the blame
for failure on the other group members. However, the authors
found that self-serving bias is mitigated when the group
members feel a close interpersonal bond with each other.
Thus, people will flatter themselves at their partner’s
expense—but only when they do not care much about the part-
ner. The interpersonal context dictates whether people will
display the self-serving bias.Self-HandicappingWhen someone self-handicaps, he or she tries to explain
away failure (or even possible failure) by attributing it to
external causes (often external causes of his or her own
making). Self-handicapping is usually studied within the
context of individual performance, but it has a strong inter-
personal aspect as well. One study manipulated whether
several crucial aspects of the situation were public (known
to others) or private (known only to the subject; Kolditz &
Arkin, 1982). Self-handicapping emerged mainly in the
public conditions, when the subject’s handicap and subse-
quent performance would be seen by others. In contrast,
subjects did not self-handicap when the experimenter was
unaware of the handicap. Apparently, self-handicapping is
primarily a self-presentational strategy used to control the
impression one makes on other people. Self-handicapping
rarely occurs when people are concerned only with their
private self-views.EMOTIONS AND THE INTERPERSONAL SELFEmotions often reflect value judgments relevant to the self.
Recent work has increasingly emphasized interpersonal
determinants and processes of emotion (Tangney & Fischer,
1995).Shame and GuiltBoth shame and guilt have strong interpersonal components.
The two terms were used in varying ways for decades, but in
recent decades a consensus has emerged about how to define
the two. The distinction was first proposed in the theoretical
work by Lewis (1971), based on clinical observations, and it
received considerable elaboration and further support from
factor analytic and scale-development studies by researchers
such as Tangney (1992, 1995). The difference between the
two lies in how much of the self is affected: Guiltdenounces
a specific action by the self, whereas shamecondemns the
entire self.
Shame is usually the more destructive of the two emo-
tions. Because shame signifies that the entire self is bad, sim-
ple reparations or constructive responses seem pointless. This
absence of constructive solutions probably leads to many of