468 Altruism and Prosocial Behavior
person should help another in need when the latter is depen-
dent on the former—that is, when others are not available to
help and thus the second person is counting specifically on
the first. Although this norm does seem to exist, its effect on
helping has been surprisingly difficult to demonstrate. After
more than a decade of research attempts to do so, Berkowitz
(1972) concluded, “The findings do not provide any clear-cut
support for the normative analysis of help-giving.... The
potency of the conjectured ‘social responsibility norm’ was
greatly exaggerated” (pp. 68, 77).
Why has evidence that the norm of social responsibility
leads to prosocial behavior been so elusive? Darley and
Latané (1970) suggested that this norm may be at once too
general and too specific. The norm may be too general in that
everyone in our society adheres to it. If this is true, it cannot
account for why one person helps and another does not. On the
other hand, the norm may be too specific in that it comes with
a complex pattern of exceptions, situations in which an indi-
vidual may feel exempt from acting in accordance with the
norm. The norm may be characterized not simply by a rule that
says, “If someone is dependent on you for help, then help,” but
by a more complex rule that says, “If someone is dependent on
you for help, then help,except when.. .” There may be indi-
vidual differences in readiness to accept exceptions—that is,
to deny responsibility (Schwartz, 1977). Moreover, excep-
tions may vary for individuals in different roles and in differ-
ent social situations. One advantage of remembering the
dramaturgical roots of the concept of norms is that it makes
explicit their role specificity.
Darley and Latané (1970) also pointed out that in addition
to norms for helping, there are norms for not helping. A per-
son may be taught, “Help those in need,” and at the same time,
“Mind your own business.” Which norm is the one to follow?
If the former, one may help; if the latter, probably not.
Effects of Race and Sex
Exceptions to and conflicts among norms may account for the
highly inconsistent effects on prosocial behavior of demo-
graphic variables such as race and sex. It has sometimes been
found that same-race helping is more frequent (e.g., Gaertner &
Bickman, 1971), sometimes that cross-race helping is more
frequent (Katz, Cohen, & Glass, 1975), and sometimes that the
race of the victim or helper makes no difference (Wispé &
Freshley, 1971). Similarly, sometimes men help more than
women (West, Whitney, & Schnedler, 1975), sometimes
women help more than men (Wegner & Crano, 1975), and
sometimes the sex of the helper makes no difference (J. A.
Piliavin & Piliavin, 1972). It does appear, however, that women
are generally more likely to be helped than are men (Gruder &
Cook, 1971).
How can we account for these seemingly contradictory
findings? One possibility is that given their different social
roles in different situations, Blacks and Whites—and men
and women—may feel more or less obligated to help a
dependent other. For example, Black students on a predomi-
nantly White campus, acutely aware of their minority status,
may feel strong responsibility for helping a fellow Black stu-
dent but very little responsibility for helping a White student;
White students may be more likely to help a Black student
when failure to do so clearly violates norms proscribing
racial prejudice. Helping may be more normative for men
than for women in one situation—for example, intervening in
a potentially dangerous emergency. Helping may be more
normative for women than for men in another situation—for
example, providing sympathy and support after a friend’s
breakup with her fiancé (Eagly & Crowley, 1986). A role-
sensitive normative analysis renders the apparent inconsis-
tencies comprehensible.Norm SalienceSome researchers have suggested that the problem with
social norms lies in norm salience and focus of attention.
Only when attention is focused on the norm as a standard for
behavior is concern about violating it likely to affect behav-
ior (Cialdini, Kallgren, & Reno, 1991). Consistent with this
suggestion, Gibbons and Wicklund (1982) found that if nor-
mative standards of helpfulness were salient and thus a focus
of attention, then focusing on oneself increased helping.
Presumably, being self-focused when the norm was salient
highlighted the threat of sanctions for failing to act in line
with personal standards. In the absence of salient standards
for helpfulness, however, self-focus led to less helping;
it seemed to inhibit attention to others’ needs (see also
Karylowski, 1984).Personal NormsBecause broad social norms like social responsibility have
limited ability to predict whether a person will help,
Schwartz (1977) proposed a change of focus in thinking
about norms. Rather than thinking about social norms,
Schwartz suggested that we should think of more specific,
personal norms. By personal norms he meant internalized
rules of conduct that are socially learned, that vary among in-
dividuals within the same society, and that direct behavior in
particular situations.
Applied to helping, a personal norm involves a sense of
obligation to perform a specific helping act. For example,
people may say (either publicly or to themselves), “I ought to
give a pint of blood in the blood drive.” Such statements