Different Research Programs on Emotion 289without medical problems. They found a similar split between
reports of genuine and as-if emotions. Cantril and Hunt
pointed not to medical conditions, but to the unique situational
circumstances of each subject. In cases of genuine emotion,
the subject’s current situation bore a “logical relationship”
to the emotion reported; in cases of as-if-emotion, no such
situation was present. Landis and Hunt (1934) also replicated
Marañon’s experiment, this time with psychiatric patients,
and obtained similar results. Landis and Hunt therefore con-
cluded that emotion was influenced by “environmental” fac-
tors and “higher intellectual and perceptual functions.”
Cantril (1934) placed subjects in four successive negative
situations (e.g., watching photographs of mutilated war
victims or hearing sudden loud noises as the lights were unex-
pectedly turned off). Subjects went through the four situations
in different orders. Each was injected four times, getting a
placebo for the first three trials and epinephrine for the fourth.
In comparison with the placebo, epinephrine increased the
subjects’ ratings of their emotional reactions in fear situations
but decreased their rated emotional reaction in disgust situa-
tions. Cantril suggested that “the awareness of some object or
situation around which the emotion is intellectually organized
is the immediate cause for the emotional experiences”
(p. 578), and that “the quality of an emotion is primarily de-
pendent upon the attitude aroused in the [subject] by the stim-
ulus” (p. 579). In this way, Cantril, Hunt, and Landis moved
away from a view in which emotion is an entity triggered by a
stimulus and defined by bodily changes, as assumed in early
ontological theories. They moved toward a view in which
emotion depends not just causally but logically on a complex
situation intellectually organized in the context of bodily
arousal.
An interesting development of this view was Nina Bull’s
(1951) attitude theory of emotion. Bull begins with a simple
framework: Situations elicit actions. Action consists of two
successive stages: (a) a preparatory phase and (b) a consum-
matory movement (e.g., fight or flight). The first stage is a
motor attitude or action readiness and includes involuntary
changes in posture and in various organs. This phase has both
a direct and an indirect consequence: The direct consequence
is the particular action (the second stage) for which the first
phase prepares. The indirect consequence is a feeling. It is
this feeling that is usually known as emotion. The feeling of
an emotion is thus an epiphenomenon of the sequence in
which a motor attitude becomes action. Although Bull shares
with ontological thinkers the attempt to identify the one
event that is the emotion, she also moved in the direction of
thinking of emotion as something that can be understood only
in terms of a process that necessarily includes both situation
and action. Indeed, unless the feeling of the emotion is
equated with the emotion itself, there is no emotion per se
within Bull’s theory of emotion.
Schachter (1964) further articulated and developed
this perspective in his two-factor theory of emotion.
Schachter’s specific ideas were an application of Lewinian
principles (Ross & Nisbett, 1991): Emotion is the result of a
tension between environmental constraints and cognitive
construals. The environmental constraints were both situa-
tional (mainly others’ behaviors) and internal (nonspecific
arousal). The cognitive construals were originally cognitive
labels but shortly afterward became attributions (Nisbett &
Schachter, 1966). Schachter’s theory of emotion dominated
the study of emotion in social psychology for several decades.
The combination of nonspecific arousal and (mis)attribution
inspired important theoretical models of aggression (e.g.,
Zillmann & Bryant, 1974), helping (e.g., Piliavin, Piliavin, &
Rodin, 1975), interpersonal attraction (e.g., Dutton & Aron,
1974), environmental behavior (e.g., Anderson & Anderson,
1984), and attitude change (e.g., Zanna & Cooper, 1974).
Mandler (1984) developed a related theory of emotion that
dominated the study of emotion in cognitive psychology
around the same time.
Ginsburg and Harrington (1996) recently proposed an ac-
count of emotion along more purely conceptualist lines. The
concept of emotion refers to an action in a context. The con-
text has two structural features. The first is hierarchical: a
broad system of events and social relationships that are nec-
essary to give meaning to the action. The second feature of the
context is linear; that is, it includes a sequence of actions un-
folding over time. Prior actions lead to (or, in Lewinian terms,
create a channel for) subsequent actions. Actions also alter
bodily state and felt experience. The entire sequence of ac-
tions in context with its accompanying bodily and mental
state is construed (conceptualized) as emotion. In this way,
there is no emotion in addition to the action in context. No sin-
gle event within the sequence can be equated with emotion.
Ginsburg and Harrington described the proper study of
emotion as descriptive. They suggested creating natural
histories of specific emotional episodes. In turn, these emo-
tional episodes are to be understood as a subsystem of larger
and more complex systems. The search for universal entities
is abandoned in favor of an exhaustive description of such
systems relevant for a particular culture.Formal Definitions for Emotion TermsMuch philosophical work on emotion has been aimed at a
formal analysis of emotion terms. Solomon’s (1976) inspired
analysis was a precursor to appraisal theories. Wierzbicka’s
(1992, 1999) linguistic analysis provides a formal framework