You are on page 1of 24

Review of General Psychology Copyright 2007 by the American Psychological Association

2007, Vol. 11, No. 2, 155–178 1089-2680/07/$12.00 DOI: 10.1037/1089-2680.11.2.155

Emotion and Rationality: A Critical Review and Interpretation


of Empirical Evidence
Michel Tuan Pham
Columbia University

The relation between emotion and rationality is assessed by reviewing empirical


findings from multiple disciplines. Two types of emotional phenomena are examined—
incidental emotional states and integral emotional responses—and three conceptions of
rationality are considered—logical, material, and ecological. Emotional states influence
reasoning processes, are often misattributed to focal objects, distort beliefs in an
assimilative fashion, disrupt self-control when intensely negative, but do not necessar-
ily increase risk-taking. Integral emotional responses are often used as proxies for
values, and valuations based on these responses exhibit distinct properties: efficiency,
consistency, polarization, myopia, scale- insensitivity, and reference-dependence. Emo-
tions seem to promote social and moral behavior. Conjectures about the design features
of the affective system that give rise to seeming sources of rationality or irrationality
are proposed. It is concluded that any categorical statement about the overall rationality
or irrationality of emotion would be misleading.

Keywords: emotion, cognition, rationality, decision, affect

The relation between emotion and rationality, needed, therefore, is a comprehensive review of
affect and reason, is an ageless question. This the wide range of empirical findings that have
question has preoccupied philosophers, com- emerged across various literatures about the re-
moners, and classical writers for many centu- lation between emotion and reason. This is the
ries. It is only recently, however, that it has object of this article.
become the subject of scientific inquiry and The article is structured in five sections. The
empirical investigations. In the past 20 years, first section introduces distinctions between two
investigations related to this question have been types of emotional phenomena—incidental
conducted across a wide range of scientific dis- emotional states and integral affective respons-
ciplines, including cognitive and social psychol- es—and three conceptions of rationality—
ogy, economics, decision research, consumer
logical, material, and ecological. The next sec-
research, and neuroscience. Unfortunately, be-
tion focuses on incidental emotional states and
cause empirical studies are necessarily
grounded in a certain theoretical, substantive, reviews their effects on reasoning, belief accu-
and methodological context, any one study can racy, self-control, and risk-taking. The third
provide, at best, only a very partial answer to section focuses on the role of integral affective
the extremely complex question of emotion and responses in judgment and decision making.
rationality. The empirical literature on emotion This section identifies distinct properties of af-
and rationality is thus very fragmented and fective responses as proxies for value and eval-
sometimes seemingly inconsistent. What is uates the “somatic marker hypothesis.” The
fourth section examines the role of emotions in
social and economic interactions. The conclud-
ing section discusses identified empirical regu-
Preparation of this article was supported by a grant from larities and advances theoretical conjectures
the Graduate School of Business of Columbia University.
The research assistance of Hannah Chang and Kaiya Liu about the principles of an affective system of
was appreciated. judgment and behavioral regulation that gives
Correspondence concerning this article should be ad- rise to seeming sources of rationality and irra-
dressed to Michel Tuan Pham, Graduate School of Busi-
ness, Columbia University, 3022 Broadway, Uris Hall 515,
tionality. It is concluded that any categorical
New York, NY 10027. E-mail: tdp4@columbia.edu statement about the overall rationality or irra-
155
156 PHAM

tionality of emotion may be simplistic and mis- Dictionary (Neufeldt, 1991, p. 1115), the word
leading. “rational” implies “the ability to reason logi-
cally, as by drawing conclusions from infer-
Types of Emotional Phenomena and ences.” People are rational (irrational) if their
Types of Rationality beliefs, judgments, choices, and actions respect
(violate) certain standards of logic. For exam-
Emotions refer to complex states of the or- ple, in the standard economic theory of choice,
ganism characterized by changes in autonomic rationality requires that preferences be transi-
nervous system arousal accompanied by distinct tive: if a person prefers A over B and prefers B
physiological expressions, specific action ten- over C, then this person must also prefer A over
dencies, and subjective feeling experiences of a C. Similarly, according to normative (Bayesian)
certain valence (see Strongman, 1987). Emo- rules of inference, if a person has to guess
tions generally, though not always, arise from a which of two types of taxis was more likely
cognitive appraisal of the emotional object or involved in a hit-and-run accident, it would be
situation in terms of its meaning for one’s well- rational to take into account the relative propor-
being (Lazarus, 1991). In this review, the term tion of each type of taxi in the area. This first
“emotion” will be used somewhat broadly to conception of rationality has been referred to as
refer to the presence of affect in general. It will logical (Kahneman, 1994).
be used not only in reference to emotions prop- A second conception of rationality empha-
er—that is, intense affective experiences such sizes the consistency between a person’s deci-
as anger, fear, joy, and love that have clear sions and actions and this person’s objectives
emotional referents— but also in relation to and self-interests. According to renowned econ-
milder affective responses, feelings, and states, omist Amartya Sen (1990, p. 210), “rationality
including moods that do not have clear refer- . . .demands cogent relations between aims and
ents. This wide-ranging use of the term “emo- objectives actually entertained by the person
tion” is intentional. If one is to have a full and the choices that the person makes.” This
appreciation of the rationality or irrationality of conception is central to standard economic the-
emotional phenomena, it is important not to orizing where it is posited that rational individ-
restrict one’s analysis to the most intense emo- uals choose courses of actions in a way that
tional experiences.1 maximizes these individuals’ own utility.
When studying the effects of emotion on Choices of inferior alternatives are irrational so
judgment, decision, and behavior, two types of are behaviors that are not in the person’s self-
emotional phenomena should be distinguished: interest (e.g., compulsive gambling, excessive
incidental emotional states and integral emo- smoking, and unprotected sex with strangers).
tional responses (Bodenhausen, 1993). Inciden- This second conception of rationality may be
tal emotional states are those whose source is referred to as material.
unrelated to the object of judgment or decision. The study of emotion raises a third type of
These states include current emotions not rationality. Certain types of behaviors and ac-
caused by the target object, preexisting mood tions are “rational” not because of they are
states, and enduring emotional dispositions such logically consistent or serve the person’s self-
as chronic anxiety. Integral emotional responses interest but because they fulfill broader societal
are those experienced in relation to the object of
judgment or decision. More specifically, inte-
gral affective responses are emotions and feel- 1
For instance, in his treaty on emotion and rationality,
ings that are elicited by features of the target Elster (1999) concentrates his analysis on intense emotional
experiences of the kind discussed by classical writers. This
object, whether these features are real, per- analytical strategy introduces two major sampling prob-
ceived, or only imagined (Cohen, Pham, & lems. First, it is unlikely that classical writings, however
Andrade, 2007).2 insightful, are statistically representative of human reality.
Three conceptions of rationality also need to Second, even if they were, an exclusive focus on intense
be distinguished in discussing the relation be- emotions is bound to overstate their general consequences.
2
Unfortunately, space constraints prevent a discussion of
tween emotion and rationality. The first concep- the extensive work on memory for affective experiences and
tion emphasizes reasoning, consistency, and affective forecast of future experiences (see, e.g., Kahneman,
logic. According to the Webster’s New World 1994).
SPECIAL ISSUE: EMOTION AND RATIONALITY 157

goals, meet higher moral standards, or serve arousal, which is known to impair working mem-
greater evolutionary purposes. Some of these ory capacity (Darke, 1988a; Humphreys &
behaviors and actions, in fact, may be against Revelle, 1984). This decrement in processing
the person’s material self-interest. For example, capacity has a variety of consequences that
it would not be in a bystander’s self-interest to seem detrimental to sound reasoning. For exam-
take on an armed mugger and attempt to rescue ple, compared to nonanxious participants, anx-
the mugger’s victim. However, if the bystander ious participants tend to (a) have lower ability to
elects to do so, one could hardly call this act recall information and organize this information
irrational. Such benevolent, altruistic acts are in memory (Mueller, 1977, 1978), (b) take
quite reasonable, even desirable, from a societal longer to verify the validity of logical inferences
or moral standpoint, even if they seem irrational (Darke, 1988b), (c) scan alternatives in a more
from a strictly material standpoint. Similarly, haphazard fashion (Keinan, 1987), (d) select an
people’s almost universal attraction to certain option without considering every alternative
ideals of beauty may seem irrational from a (Keinan, 1987), (e) commit more errors in geo-
logical standpoint and could also be materially metric and semantic analogical problems
irrational if it leads to unfortunate outcomes (Keinan, 1987; Leon & Revelle, 1985), and (f)
(e.g., heartbreak). However, there is evidence
process persuasion arguments less thoroughly
that the attraction to certain standards of beauty
(Sanbonmatsu & Kardes, 1988, but see Pham,
is sensible from an evolutionary standpoint (see
1996). Intense emotional states, such as anxiety,
Etcoff, 1999). Certain behaviors and attitudes
may therefore be “rational,” not in the logical or therefore appear to produce deficits in people’s
material sense, but in terms of their consistency reasoning abilities. However, this conclusion
with societal goals, moral standards, or evolu- needs to be qualified in several respects.
tionary purposes. This third form of rationality Because most of these findings pertain to the
can be termed “ecological” in that it reflects effects of high anxiety, it is not clear whether
humans’ ability to relate to their environment, they generalize to other intense emotions (e.g.,
whether social, cultural, or natural.3 A primary joy, anger, intense pride). For example, unlike
function of emotions may in fact be to support other intense emotions, anxiety involves a cog-
this ecological form of rationality. nitive element of worry that could be driving
some of the above-described deficits. In addi-
tion, the effects of intense arousal on cognitive
Rationality/Irrationality of Incidental performance are not always negative (Humphreys
Emotional States & Revelle, 1984). Finally, states of intense
emotional arousal appear to benefit reasoning in
Emotional states are incidental if their source at least one respect. In task settings where mul-
is unrelated to the object of judgment or deci-
tiple cues are available, emotionally aroused
sion. Incidental emotional states have a variety
individuals seem to adjust to their reduced pro-
of rational and irrational influences on judg-
cessing capacity by narrowing down their cue-
ments, decisions, and behaviors. They influence
people’s reasoning processes, the accuracy of utilization to the more diagnostic cues at the
their beliefs, their ability to exert self-control, expense of the less diagnostic cues (e.g., Bacon,
and their tendency to take risks. They are also 1974; Hockey, 1973). As a result, states of high
often misattributed to target objects. emotional arousal tend to increase the relative

3
Effects of Emotional States on Reasoning Clore (2005) recently proposed that in the study of
emotion and rationality, the notion of “value” should be
defined more broadly than in typical economic discourse. In
Emotional states influence people’s reason- addition to (a) the desirability of outcomes (captured by the
ing processes, and therefore their logical ratio- notion of material rationality), Clore suggests including (b)
nality. The desirability of these influences the consistency of actions with standards and (c) the con-
seems to be a function of the intensity of the sistency of attributes of objects with tastes and attitudes.
The ecological rationality proposed here subsumes Clore’s
states, their valence, and their appraisal content. second category (the consistency with standards) and part of
Most intense emotional states, except sadness, his third category, namely the consistency with tastes and
are accompanied by high levels of autonomic attitudes rooted in evolutionary logic.
158 PHAM

reliance on diagnostic versus nondiagnostic in- Negative moods, especially those of the sad-
formation in judgment (Pham, 1996). ness kind, have generally been found to have
Milder emotional states also influence rea- effects that mirror those described above. Com-
soning processes. Compared to neutral moods, pared to neutral and pleasant moods, sad moods
good moods have been found to lead individu- have been found to increase the care with which
als to (a) categorize objects more broadly (Isen people process substantive information in per-
& Daubman, 1984; Isen, Niedenthal, & Cantor, suasion (Bless, Bohner, Schwarz, & Strack,
1992), (b) generate more creative answers in 1990; Sinclair, Mark, & Clore, 1994), decrease
response-generation tasks (Greene & Noice, the reliance on general knowledge structures
1988; Hirt, Melton, McDonald, & Harackiewicz, such as scripts and stereotypes (Bless, Schwarz,
1996), (c) perform better in problem-solving Clore, Golisano, & Rabe, 1996; Bodenhausen,
tasks that require ingenuity (Greene & Noice, Kramer, & Suesser, 1994), increase the ability
1988; Isen, Daubman, & Nowicki, 1987), and to estimate covariation from scatter plot data
(d) solve multiattribute decision problems more (Sinclair & Mark, 1995), reduce the suscepti-
efficiently (Isen & Means, 1983). According to bility to halo effects (Sinclair, 1988), reduce
Isen (2001), these and other findings show that fundamental attribution errors (Forgas, 1998),
positive moods have generally beneficial effects and increase the transitivity of preferences
on reasoning, problem solving, judgment, and (Fiedler, 1988). Overall, sad moods seem to
decision making. This conclusion needs to be trigger a more systematic, data-driven, and an-
tempered because other studies suggest that alytical form of reasoning consistent with logi-
positive moods lead to poorer reasoning perfor- cal rationality. One possible explanation, based
mance in a variety of respects. Positive mood on the “affect-as-information” hypothesis
individuals are more likely to overestimate the (Schwarz, 1990), is that sad moods signal to the
degree to which others’ actions are driven by individual that the situation is problematic and
therefore requires a more vigilant form of pro-
personals disposition as opposed to situational
cessing (Schwarz, 2002). Not all negative
factors, a bias known as the “fundamental attri-
moods trigger this vigilant form of processing.
bution error” (Forgas, 1998). Positive mood has
States of anger and disgust seem to decrease the
also been found to decrease performance in
depth of processing and increase the reliance on
deductive reasoning tasks (Oaksford, Morris, stereotyping and other heuristic cues, appar-
Grainger, & Williams, 1996) and result in more ently because these states trigger a sense of
intransitive preferences (Fiedler, 1988). Numer- certainty (Bodenhausen, Sheppard, & Kramer,
ous attitude and persuasion studies also indicate 1994; Tiedens & Linton, 2001). Note also that
that positive moods decrease the depth with only mild states of sadness (sad moods) seem to
which people process substantive information produce these beneficial effects. More intense
(Batra & Stayman, 1990; Bless, Bohner, states of sadness such as chronic depression
Schwarz, & Strack, 1990; Bless, Mackie, & seem to interfere with reasoning and effortful
Schwarz, 1992; Mackie & Worth, 1989; Worth processing (Conway & Giannopoulos, 1993;
& Mackie, 1987). Positive mood individuals Hartlage, Alloy, Vazquez, & Dykman, 1993;
seem to rely instead on global knowledge struc- Silberman, Weingartner, & Post, 1983).
tures and internal cues including scripts (Bless,
Schwarz, Clore, Golisano, & Rabe, 1996), ste- Misattribution of Incidental Affective
reotypes (Bodenhausen, Kramer, & Suesser, States
1994), and judgmental heuristics, such as ease
of retrieval (Ruder & Bless, 2003). Overall, A robust result about the psychology of emo-
positive moods seem to have mixed effects on tion is that people tend to attribute their affec-
people’s reasoning. On the one hand, they seem tive states to whatever object is the current
to promote greater flexibility and creativity in focus of their attention (Schwarz & Clore,
problem solving, which appears logically desir- 1996). When the object of attention is indeed
able; on the other hand, they seem to promote a the source of feelings, this is logically rational.
more top-down, less data-driven, and less thor- However, people tend to make this attribution
ough mode of processing, which appears logi- even when the actual source of the feelings is
cally less desirable. totally unrelated to the object of attention. In a
SPECIAL ISSUE: EMOTION AND RATIONALITY 159

classic study, Schwarz and Clore (1983) found related to the source of disgust. This is presum-
that respondents who were in a good mood as a ably because disgust typically signals an inter-
result of being surveyed on a sunny day re- nal source of discomfort, which encourages the
ported higher levels of life satisfaction than rejection of possessions.
respondents who were in a bad mood as a result The degree to which people misattribute their
of being surveyed on a rainy day. Respondents affective states is not absolute, however. Misat-
mistakenly inferred that their weather-induced tributions usually disappear when people are
moods reflected how they felt about their per- made aware of the true source of their affective
sonal lives. Similar logically irrational misattri- states (e.g., Gorn, Goldberg, & Basu, 1993;
butions have been found in numerous studies Schwarz & Clore, 1983; Siemer & Reisenzein,
showing that incidental mood states generally 1998). This explains why intense emotions,
have assimilative (mood-congruent) effects on whose source tends to be salient, are less likely
object evaluation (Albarracin & Kumkale, to be misattributed. Attributions seem to depend
2003; Ottati & Isbell, 1996; Siemer & Reisenzein, on the perceived representativeness of the af-
1998). fective state with respect to the target (Pham,
The widespread misattribution of incidental 1998; Strack, 1992). For example, incidental
mood states may explain the puzzling effects of affective states are more likely to be misattrib-
weather on the stock market, the presumed cit- uted when there is a surface resemblance be-
adel of rationality. In a challenge to the hypoth- tween the source of the affective state and the
esis that financial markets are efficient, a num- domain of the decision (Raghunathan, Pham, &
ber of studies have recorded above-average Corfman, 2006).
stock market performance on sunny days and
below-average performance on rainy and winter Effects of Emotional States on Belief
days (Hirshleifer & Shumway, 2003; Kamstra, Accuracy
Kramer, & Levi, 2003; Saunders, 1993). A
plausible explanation is that a sunny weather A basic requirement of logical rationality is
puts investors in a good mood that they misin- an accuracy of perceptions and beliefs. A large
terpret as optimism about the stock market, body of evidence indicates that incidental affec-
therefore taking more risk; rainy or winter tive states tend to distort people’s perceptions
weather puts investors in a depressed mood that and beliefs about objects in an assimilative fash-
they misinterpret as pessimism about the stock ion (Isen, Shalker, Clark, & Karp, 1978; Mayer,
market, therefore taking less risk. Gaschke, Braverman, & Evans, 1992), espe-
People misattribute not only the valence of cially if the target is ambiguous (Gorn, Pham, &
their incidental affective states, but also their Sin, 2001; Isen & Shalker, 1982; Miniard,
arousal and cognitive appraisal components. Bhatla, & Sirdeshmukh, 1992). For example,
Numerous studies have shown that incidental participants who watched a commercial appear-
emotional arousal is often misinterpreted as an ing in a happy TV program perceived it to be
integral affective response to a target, resulting more effective than participants who watched
in more polarized evaluations of this target the same commercial embedded in a sad pro-
(Dutton & Aron, 1974; Foster, Witcher, Campbell, gram (Goldberg & Gorn, 1987), and consumers
& Green, 1998; Gorn, Pham, & Sin, 2001; who tasted a soft drink after watching a pleasant
Mattes & Cantor, 1982; White, Fishbein, & movie rated the beverage’s attributes more fa-
Rutsein, 1981; Zillmann, 1971). Incidental vorably than participants who tasted it after
arousal is misconstrued as an integral response watching an unpleasant movie (Dommermuth
to the target, resulting in polarization because & Millard, 1967). These assimilative influences
people “feel strongly about” the target (Gorn, extend beyond strict evaluative judgments. For
Pham, & Sin, 2001). People also seem to inter- example, risks are perceived to be higher under
pret the appraisal content of their incidental negative moods than under positive moods
emotional states as if the states were related (Johnson & Tversky, 1983; Wright & Bower,
to the target. For example, Lerner, Small, and 1992) and under incidental states of fear than
Loewenstein (2004) found that incidental states under incidental states of anger (Lerner &
of disgust decreased the perceived value at- Keltner, 2001). Sad events are also perceived to
tached to possessions that were objectively un- be more likely under incidental states of sadness
160 PHAM

and angering events more likely under inciden- have been a methodological and statistical arti-
tal states of anger (DeSteno, Petty, Wegener, & fact (Allan, Siegel, & Hannah, 2007); Msetfi,
Rucker, 2000). These assimilative influences Murphy, Simpson, & Kornbrot, 2005). Empiri-
can be explained in terms of misattribution: cal support for the depressive realism hypothe-
incidental affective states may be misinter- sis is therefore very mixed. One possible expla-
preted as integral affective responses to the tar- nation for this inconsistency may be that depres-
get (Schwarz, 1990). However, other processes sion only enhances realism in tasks that involve
may also be at work (Forgas, 1995), including some reassessment of the current situation. This
selective encoding or retrieval of affect- is because, as explained in the concluding sec-
consistent information under incidental emo- tion, sadness-related states may serve as a signal
tional states (Bower, 1981; Isen, Shalker, Clark, for situational reassessment.
& Karp, 1978). In summary, it seems that incidental affective
Although momentary emotional states often states undermine logical rationality by distort-
distort people’s perceptions, it has been hypoth- ing beliefs in an assimilative fashion, especially
esized that chronically depressed individuals if the target is ambiguous and if the state is
have more accurate perceptions of reality than improperly attributed to the target. One debated
nondepressed individuals, whose perceptions exception could be chronic states of depression,
tend to be self-enhancing—a hypothesis known which may enhance belief accuracy under cer-
as depressive realism (Alloy & Abramson, tain conditions.
1988). In a seminal study, Alloy and Abramson
(1979) found that depressed individuals had Effects of Emotional States on Self-
more accurate perceptions of the contingency Control and Risk-Taking
between their behavior and some environmental
outcome than did nondepressed individuals who Improper self-control and excessive risk-
tended to overestimate this contingency when taking (or avoidance) are primarily matters of
the outcome was desirable and underestimate it material rationality (although they also raise
when the outcome was undesirable. Seemingly issues of ecological rationality). It is well estab-
more objective perceptions among depressed lished that intense drive states such as hunger,
individuals have been observed in several other pain, sexual arousal, drug cravings, and sleep
studies (e.g., Alloy & Ahrens, 1987; Gotlib, deprivation produce breakdowns in self-control
McLachlan, & Katz, 1988; Keller, Lipkus, & and increase people’s willingness to take risks
Rimer, 2002; Lewinsohn, Mischel, Chaplin, & in order to alleviate the drive state (Loewenstein,
Barton, 1980; Martin, Abramson, & Alloy, 1996). For example, states of high sexual arousal
1984; see Dobson & Franche, 1989, for a re- increases people’s willingness to use unethical
view). For example, compared to nondepressive means of getting sex and decrease their willing-
individuals, depressive individuals have been ness to practice safe sex (Ariely & Loewenstein,
found to attend more evenly to positive, neutral, 2005; Bouffard, 2002).
and negative words (Gotlib, McLachlan, & Negative emotional states, especially intense
Katz, 1988) and revise their estimates of health ones, produce similar breakdowns in self-
risks more accurately after receiving medical control. Negative affective states have been
feedback (Keller, Lipkus, & Rimer, 2002). found to (a) reduce the ability to resist tempta-
However, other studies have failed to support tion and delay gratification among children
the depressive realism hypothesis (Benassi & (Fry, 1975; Schwarz & Pollack, 1977; Seeman
Mahler, 1985; Dunning & Story, 1991) and & Schwarz, 1974), (b) increase the tendency to
suggest that the phenomenon might not gener- overeat among dieters (Ruderman, 1986), (c)
alize to more meaningful and consequential increase the chance of relapse among people
tasks (Pacini, Muir, & Epstein, 1998). For ex- trying to quit smoking (Shiffman & Waters,
ample, Dunning and Story (1991) found that 2004), (d) encourage shopping among compul-
depressed individuals were less accurate and sive buyers (Faber & Christenson, 1996), (e)
more overconfident in predicting the probability increase procrastination (Tice, Bratslavsky, &
of future personal events than nondepressed in- Baumeister, 2001), and (f) produce overcon-
dividuals. Some analyses suggest that the orig- sumption of limited collective resources (Knapp
inal Alloy and Abramson (1979) findings may & Clark, 1991). Recent studies suggest that this
SPECIAL ISSUE: EMOTION AND RATIONALITY 161

phenomenon is not because of a decrease in the level of arousal is held constant (Mano,
motivation or ability to self-regulate under neg- 1992, 1994). Lerner and Keltner (2001) simi-
ative affective states, but to a shift in priority larly observed that fear tends to trigger risk-
among distressed individuals who seem to place aversion, whereas anger tends to trigger risk-
the immediate goal of feeling better ahead of seeking even though both are high-arousal neg-
other goals (Tice, Bratslavsky, & Baumeister, ative emotions. This is because fear, like
2001). anxiety, is typically associated with situations
Given the disruptive effects that intense neg- of uncertainty and low control, whereas anger is
ative affective states have on self-control, one typically associated with situations of certainty
would intuitively predict that these states should and high control. Disgust, another high arousal
also make people more risk-risking. The empir- emotion, has also been found to decrease risk-
ical evidence is in fact very mixed. Several seeking in gambling among women (Fessler,
studies indicate that negative emotional states Pillsworth, & Flamson, 2004).
with strong arousal increase risk-seeking In summary, unlike drive states, intense neg-
(Fessler, Pillsworth, & Flamson, 2004; Leith & ative emotions do not have a uniformly positive
Baumeister, 1996; Mano, 1992, 1994). For ex- effect on risk-seeking (see Hockey, Maule,
ample, Leith and Baumeister (1996) found that Clough, & Bdzola, 2000). High emotional
participants who were angry or anticipated an arousal seems neither necessary, nor sufficient
impending embarrassment were more likely to to explain risk-seeking under negative emo-
choose economically inferior “long-shot” gam- tions. Rather, the effects of negative emotions
bles over superior “safe-bet” gambles, whereas on risk-seeking seem to depend on complex
sad participants did not exhibit this bias. interactions between the goals activated by the
Fessler, Pillsworth, and Flamson (2004) also emotional state and the nature of the risks to be
found that anger triggered more risk-seeking in taken. This may explain why a meta-analysis of
gambling, especially among men. Similarly, published studies relating chronic states of an-
Mano (1994) found that intense emotional ger, sadness, and anxiety to risky sexual behav-
arousal increased the willingness to pay for ior found virtually no correlation (Crepaz &
lotteries and decreased the willingness to pay Marks, 2001).4
for insurance, in other words increased risk-
taking for both potential gains and potential
losses. Rationality/Irrationality of Integral
However, other findings indicate that peo- Emotional Responses as Proxies for Value
ple’s attitude toward risk under negative emo-
tions is not just a function of the level of arousal Unlike incidental affective states, integral af-
associated with the emotion, but also a function fective responses are those elicited by perceived
of the appraisal content of the emotion (Lerner or imagined features of the target object. Inte-
& Keltner, 2001; Raghunathan & Pham, 1999). gral affective responses play a major role in
Raghunathan and Pham (1999) observed that in people’s evaluations of, decisions about, and
risk-reward-tradeoff situations, anxious individ- behavior toward objects, even if the attributes of
uals tend to prefer low-risk/low-reward options, the objects are held constant (e.g., Abelson,
whereas sad individuals tend to prefer high-risk/ Kinder, Peters, & Fiske, 1982; Hsee & Kunreuther,
high-reward options—a seeming reversal of the 2000; Pham, 1998; Pham, Cohen, Pracejus, &
high-arousal/high-risk pattern observed else- Hughes, 2001). The primary reason seems to be
where (see also Raghunathan et al., 2006). Ac- that integral affective responses are often used
cording to Raghunathan and Pham (1999), this as proxies for value: things that feel good must
is because anxiety, which is typically associated be desirable, and things that feel bad must be
with situations of low control and high uncer- undesirable (see, e.g., Damasio, 1994; Pham,
tainty, activates a goal of risk and uncertainty 2004; Schwarz & Clore, 1996; Slovic, Finucane,
minimization, whereas sadness, which is typi- Peters, & MacGregor, 2002), a pervasive infer-
cally experienced in response to the loss of a
source of reward, activates a goal of reward 4
Another explanation could be that chronic emotional
maximization. Other studies have also uncov- states have lesser effects on risk-taking than momentary
ered decreased risk-seeking under anxiety when states.
162 PHAM

ence that accounts for the often-observed assim- conveyed in a nonemotional manner (e.g.,
ilative effect of incidental affective states on Hendrickx, Vlek, & Oppewal, 1989;
target evaluations. Loewenstein, Weber, Hsee, & Welch, 2001). In
a different setting, Ratner and Herbst (2005)
Properties of Integral Emotional observed that after investment with a broker
Responses as Proxies for Value with a strong track record produced a disap-
pointing outcome, individuals who focused on
Speed and processing efficiency. Judgments their affective responses to the outcome tended
and decisions based on integral emotional re- to “overreact” and switch their investments to a
sponses are generally reached more rapidly than broker with an inferior track record. Emotional
those based on descriptive inputs, both in stim- responses to a single outcome appeared to over-
ulus-based tasks (Pham et al., 2001; Zajonc, ride the presumably more reliable track-record
1980) and in memory-based tasks (Verplanken, of the brokers.
Hofstee, & Janssen, 1998). This seems to be Several factors can explain why evaluative
because integral affective responses can arise and behavioral responses based on integral af-
very rapidly (LeDoux, 1996; Zajonc, 1980) and fect tend to be more polarized. First, affective
enter evaluations through simple associations responses to everyday objects tend to be more
(De Houwer, Thomas, & Baeyens, 2001) or extreme than reason-based assessments of the
straightforward interpretation (Pham et al., same objects, even when the information about
2001; Strack, 1992). Judgments and decisions the object is held constant (Pham et al., 2001).
based on integral affective responses also Second, initial affective responses to an object
generally require less processing resources seem to trigger a confirmatory search for infor-
(Epstein, 1990). This is evident from the ro- mation that supports the initial feelings (Pham
bust finding that constraints on processing et al., 2001; Yeung & Wyer, 2004). This con-
resources such as time pressure, distraction, firmatory search increases the subjective coher-
or cognitive load generally increase the reli- ence of judgments based on affect (Pham,
ance on integral affective responses in judg- 2004), increasing polarization. In addition, as
ment and choice (e.g., Avnet & Pham, 2004; discussed further, affective responses seem to
Nowlis & Shiv, 2005; Pham et al., 2001; Shiv be relatively insensitive to probability and
& Fedorikhin, 1999). Overall, it seems that quantity, which would otherwise mitigate in-
integral affective responses provide fast and terpretations of these responses (Hsee &
resource-efficient assessments of value. Rottenstreich, 2004; Rottenstreich & Hsee,
Whether these assessments are logically and 2001). Finally, integral affective responses have
materially rational depends on a variety of inherently strong drive properties (Frijda, 1988;
considerations—some of which are evaluated Lazarus, 1991).
below. However, from an ecological stand- The greater polarization of affect-based eval-
point, judgmental speed and resource- uative and behavioral responses may lead to
efficiency seem to be desirable properties. logical irrationalities. For instance, Johnson,
Extremity and polarization. Everything else Hershey, Meszaros, and Kunreuther (1993) ob-
equal, judgments, decisions, and behaviors served that people were willing to pay more for
based on integral affective responses tend to be a flight insurance policy covering “death due to
more extreme and polarized than those based on any act of terrorism”—a concrete, emotion-
more descriptive inputs—a phenomenon related producing threat—than for an insurance policy
to the so-called “vividness effect” in judgment covering “death due to any reason”—a logically
(Nisbett & Ross, 1980). For example, in France, higher, but less concrete threat. Similarly, the
newspaper articles using the emotional label “overreaction” of emotion-focused investors
“Mad Cow disease” resulted in more dramatic observed by Ratner and Herbst (2005) seems to
decreases in beef consumption than comparable be a case of logically irrational overweighing of
articles using the scientific label “Creutzfeldt- single instances relative to more reliable base-
Jakob disease” (Sinaceur, Heath, & Cole, rates. Nevertheless, it would seem premature to
2005). In general, responses to risks seem to be draw firm conclusions about the general ratio-
stronger when the risks are conveyed in an nality or irrationality of affect-based response
emotion-provoking manner than when they are polarizations. For example, in the Ratner and
SPECIAL ISSUE: EMOTION AND RATIONALITY 163

Herbst (2005) studies, it is not clear that it was ments and decisions based on integral affective
irrational for the emotional investors to place responses tend to be sensitive to the presence or
more weight on the broker’s recent performance absence of affect-producing stimuli but rela-
than on the broker’s stated track record, espe- tively insensitive to variations in the magnitude
cially in a domain where past records are noto- of these stimuli. In one study, respondents were
riously unpredictive of the future performance. asked how much they would be willing to do-
It is also not clear that the decrease in beef nate to save either one or four pandas. When the
consumption following articles emotional titled number of pandas saved was represented in an
“Mad Cow Disease” was materially, ecologi- abstract fashion, donations were much higher in
cally, or even logically irrational. Rather, it the four-panda condition than in the one-panda
seems that response polarization is a natural condition, as would logically be expected.
by-product of an important function of affective However, when the number of pandas saved
and emotional responses, which is to motivate was represented in an affectively rich fashion,
behavior and redirect action if necessary. That donations were not different in the four- and
this function occasionally produces “overreac- one-panda conditions. These results echo other
tions” does not necessarily undermine the ap- findings showing that when assessing the value
parent ecological rationality of this function. of programs designed to save a large number of
Myopia. Evaluations and decisions based human lives—an emotionally charged judg-
on integral affect tend to be more myopic in that ment—people exhibit substantial insensitivity
immediate affective rewards and punishments to the absolute number of lives saved (Fether-
are weighted much more heavily than delayed stonhaugh, Slovic, Johnson, & Friedrich, 1997).
affective consequences (Loewenstein, 1996). According to Hsee and Rottenstreich (2004),
This property is very salient in self-control sit- this phenomenon arises because integral affect-
uations where people have to trade-off the im- based evaluations are often based on mental
mediate hedonic consequences of an option images (see also Pham, 1998). These images
against its long-term consequences. According tend to involve discrete prototypical representa-
to Loewenstein (1996), the myopia of affect- tions of the target but not continuous quantita-
based judgments and decisions is because of the tive information (see also Kahneman, Ritov, &
differential accessibility of current and delayed Schkade, 1999).
affective responses. Whereas the experience of Similarly, evaluations and decisions based on
immediate integral affect has strong drive prop- integral affective responses tend to be insensitive
erties, it is much more difficult to picture a to probabilities, except for the presence or absence
future affective experience. Consistent with this of uncertainty (Loewenstein, Weber, Hsee, &
proposition, recent brain imaging studies indi- Welch, 2001; Monat, Averill, & Lazarus, 1972;
cate that preferences for immediate rewards are Rottenstreich & Hsee, 2001; Sunstein, 2003). For
associated with greater activation in parts of the example, Rottenstreich and Hsee (2001) observed
limbic system associated with affect (McClure, that, consistent with economic theory, people
Laibson, Loewenstein, & Cohen, 2004). Affec- were willing to pay much more to avoid a high
tive rules of valuation thus seem to be geared to probability of losing $20 than to avoid a low
the present (Pham, 2004), which will lead to probability of losing $20. However, people were
material irrationality in domains where present not willing to pay much more to avoid a high
and long-term hedonic consequences are nega- probability of receiving an electric shock—a pros-
tively correlated. pect rich in negative affect—than to avoid a low
Concreteness and scale insensitivity. As probability of receiving the same shock (see also
epitomized by the notion of expected-monetary Sunstein, 2003). According to Loewenstein and
value, a logically rational measure of value colleagues (2001), this phenomenon again arises
would take into account the magnitude of the because affective decisions under uncertainty rely
value-producing stimulus and the uncertainty on discrete images of the options that do not
that surrounds it. It appears, however, that when incorporate probabilities. This is consistent with
integral affective responses are used as proxies the finding that awareness of the timing of immi-
for value, these responses are not scaled prop- nent threat produces the same level of stress and
erly for either magnitude or probability. Hsee physiological arousal whether the threat has a 5%,
and Rottenstreich (2004) observed that judg- 50%, or 100% probability of occurrence (Monat,
164 PHAM

Averill, & Lazarus, 1972). This is also consistent trical shocks is important, regardless of the num-
with the finding that when given a chance to draw ber of lives saved or the actual probability of
a winning red bean either from a small bowl shock.
containing a single red bean and 9 white beans or Reference-dependence. Compared to those
from a larger bowl containing between 5 and 9 red based on more descriptive inputs, assessments
beans and 91 to 95 white beans, many people’s of value that are based on integral affective
“gut feeling” is to draw from the larger bowl, even responses tend to be more relativistic or refer-
though the probability of winning is greater when ence-dependent. That is, affective valuations
drawing from the smaller one (Denes-Raj & are often not based on the focal object or out-
Epstein, 1994; Kirkpatrick & Epstein, 1992). This come in isolation, but in relation to other objects
finding again suggests that affective valuations of or outcomes. For example, emotional responses
chances are driven by concrete representations of to the outcome of a gamble are driven not only
exemplars (seeing multiple red beans) rather than by the monetary value of the actual (realized)
more abstract notions of probability (the distribu- outcome, but also by how this outcome com-
tion of beans in a random draw process). pares relative to unrealized outcomes (Mellers,
Note, however, that affect-based decisions and Schwartz, Ho, & Ritov, 1997). Winning $10 in
evaluations are very sensitive to one range of a gamble will elicit greater pleasure if the other
probability: deviations from absolute certainty, possible outcome is losing $5 than if it is losing
from impossibility to small probability and vise only $1. Similarly, losing $5 in a gamble will
versa (Brandstatter, Kuhberger, & Schneider, elicit greater displeasure if the other possible
2002). For example, people grossly overpay to outcome is winning $10 than if it is winning
turn zero probabilities of winning in big lotteries, only $2. This finding is consistent with a large
a prospect rich in affect, into probabilities that are body of research showing that emotional re-
infinitesimal. Similarly, most people would be sponses to outcomes are very sensitive to sponta-
willing to pay large insurance or security premi- neous comparisons with outcome counterfactuals
ums to convert minute probabilities of cata- (e.g., Kahneman & Miller, 1986; Landman, 1987;
strophic events, prospects also rich in affect, into Medvec, Madey, & Gilovich, 1995). More direct
zero probabilities. In affective valuations people evidence comes from a study by Tversky and
thus appear to be sensitive to possibility—that is, Griffin (1991) in which participants were asked to
deviations from certainty—rather than actual evaluate two hypothetical jobs: one company of-
probability (see also Slovic et al., 2002). fered a higher salary, but offered other colleagues
Overall, the above findings suggest that valua- even more money; the other offered a lower sal-
tions based on integral affect tend to be insensitive ary, but offered other colleagues even less money.
to scale, whether scale refers to the quantitative When asked to predict which of the two jobs they
magnitude of the stimulus or to the probability that would by happier at—that is, when asked to make
surrounds it. This scale insensitivity violates log- an affective evaluation—most participants se-
ical rationality. One interpretation is that integral lected the lower-paying job. When asked to make
affect is a mostly categorical means of assessing a choice between the two jobs—a presumably less
value. Integral affective responses arise from ob- affective evaluation—most participants chose the
jects being categorized in terms of their signifi- higher-paying job. Therefore, participants asked to
cance for well-being (Lazarus, 1991). This cate- make a “cold” decision seem to focus on the
gorization appears to obey a principle of concrete- objective personal payoff, whereas those asked to
ness. For objects to be categorized as emotionally make a more affective assessment seem to also
significant—that is, for them to elicit integral af- take into account social comparisons with their
fective responses—they need to be represented colleagues. In a similar study, participants were
concretely. For example, people will be happier if asked to compare two hypothetical jobs that were
they know for sure whether they have won a identical in terms of compensation and workload:
dinner or a CD than if they are uncertain about company A offered a small office and gave an-
which of these two prizes they have won (Vandijk other comparable employee an equally small of-
& Zeelenberg, 2006). In this representation, the fice, whereas company B offered a larger office
identity of the object is more critical than its and gave another comparable employee an even
distribution. In the resultant affective accounting larger office (Hsee, Zhang, Yu, & Xi, 2003).
of value, saving people’s lives or avoiding elec- When asked to make a cold choice, participants
SPECIAL ISSUE: EMOTION AND RATIONALITY 165

tended to choose the job with the objectively tures (Langlois et al., 2000). Similarly, emo-
larger office. However, when asked to assess tional responses to music appear largely shared
which job would make them feel happier, partic- (Peretz, Gagnon, & Bouchard, 1998). People
ipants tended to select the one with the smaller also exhibit high consensus in how outraged
office. they are by various types of wrongdoings
The greater relativism or reference-dependence (Kahneman, Schkade, & Sunstein, 1998) and
of affective assessments of value is logically irra- how upset they feel about various environmen-
tional. Logically, utility should depend only on tal problems (Kahneman, Ritov, Jacowitz, &
realized and personal outcomes, not on counter- Grant, 1993). People seem to agree more on
factual comparisons with unrealized outcomes or how they feel toward everyday stimuli such as
social comparisons with others’ welfare. Whether magazine pictures and TV commercials than
this relativism is also materially irrational depends they do on their reason-based assessments of the
on assumptions about people’s utility function. If same stimuli (Pham et al., 2001). Therefore,
we assume that this function incorporates only contrary to popular beliefs that feelings are
objective arguments (e.g., the magnitude of the highly subjective, a variety of findings suggest
person’s lottery gain, the absolute size of the per- that judgments based on integral affect are quite
son’s office), this relativism would be materially consensual, sometimes even more so than judg-
irrational. However, if we assume that people’s ments based on descriptive inputs.5
long-term well-being also depends on the broader There is also growing evidence that judgments
context in which personal outcomes are realized and decisions based on integral affect also tend to
and experienced, this relativism may in fact be be more consistent intrapersonally. Indirect evi-
materially rational. After all, a person working for dence comes from the finding that individuals
a more modest salary in a smaller office at a who verbalized their reasons for liking or disliking
company that treats its employees with equity may various posters before making a choice, and pre-
be happier than a comparable employee working sumably relied on these reasons, were subse-
for a larger salary in a larger office at a company quently less satisfied with their choice than indi-
that treats other comparable employees better. viduals who were not asked to verbalize their
This relativism may be also beneficial from an reasons and presumably relied on their spontane-
ecological standpoint, as discussed further in the ous affective responses to the posters (Wilson et
discussion of the role of emotion in social and al., 1993). More direct evidence comes from the
economic interactions. In addition, the counterfac- finding that for high-involvement products (e.g.,
tual outcome comparisons that underlie emotional camcorders, cell phones), integral affective re-
experiences of regret, rejoicing, disappointment, sponses to the products are more predictive of
or elation may help people learn from their fail-
long-term satisfaction with the products than util-
ures and successes (Roese, 1997), whereas social
itarian beliefs about the product’s benefits (Darke,
comparisons with other’s welfare may support
Chattopadhyay, & Ashworth, 2006). Other re-
norms of justice and equity that seem broadly
search suggests that reliance on integral affective
desirable at a societal level. One possible interpre-
responses increases not only intrapersonal consis-
tation of this relativism is that the affective system
tency over time, but also intrapersonal consistency
of valuation is mostly ordinal, as opposed to car-
dinal. That is, the affective system may be more over choices. Lee and Ariely (2006) recently ob-
concerned with the relative desirability ordering of
alternative states of the world or alternative 5
According to Pham and colleagues (2001), affect-based
courses of actions than with their absolute desir- judgments will tend to be very consensual whenever the
ability. underlying integral affective responses are triggered
through hardwired programs involved in bioregulation or
Interpersonal and intrapersonal consistency. through emotional schemata acquired through socialization.
A growing body of evidence suggests that judg- Such affect-based judgments will generally be more con-
ments based on integral affective responses sensual than reason-based judgments that are constructed in
have high interpersonal consistency. Contrary a piecemeal fashion. However, affect-based judgments will
to the notion that “Beauty is in the eye of the be less consensual when based on integral affective re-
sponses arising through controlled appraisal processes.
beholder,” judgments of physical attractiveness, Such judgments will generally be less consensual than rea-
which are largely affective, exhibit a high de- son-based judgments based on shared stereotypes and wide-
gree of consensus both within and across cul- ly-accepted normative criteria.
166 PHAM

served that conditions that are known to increase did not exhibit such anticipatory activation
the reliance on affect in decision making (e.g., (Bechara, Damasio, Tranel, & Damasio, 1997).
pictures, time pressure, memory load, expression According to the somatic marker hypothesis
of own preference) also increase the transitivity of (Damasio, 1994), in normal individuals, emo-
choices between products. tional responses evoked by objects are stored
In summary, compared to judgments and deci- with memory representations of these objects as
sions based on descriptive inputs, judgments and somatic markers of these objects’ value. Subse-
decisions based on integral affective responses quent encounters with these objects will trigger
tend to exhibit higher consistency, both within and anticipatory feelings that will steer the decision
across people. Intrapersonal consistency is logi- maker either toward or away from these objects
cally desirable. It is also materially desirable if it depending on the valence of the stored markers.
reduces the chance of postdecisional regret. In Among normal participants in the above stud-
domains where a common criterion of value can ies, emotional responses to early penalties of the
be assumed, interpersonal consistency is also log- risky decks were registered and subsequently
ically desirable because, even if interpersonal steered participants away from these decks, ap-
agreement does not guarantee accuracy, lack of parently even unconsciously. Among emotion-
agreement implies that at least one party is inac- ally deficient VMPC patients, no such learning
curate (Kruglanski, 1989). Consensus is also de- took place; as a result, they continued to draw
sirable on ecological grounds. For example, if a from the more tempting but less advantageous
society is to set standards of punishment for var- decks. According to Damasio and his col-
ious types of crimes, it would be desirable to have leagues (see Bechara, 2004; Damasio, 1994),
a common metric to assess each crime’s undesir- emotional deficits associated with prefrontal
ability. The emotional outrage that people have damages impair performance not just in the IGT
been found to share could provide such a metric. but in decision making in general. For example,
Bar-On, Tranel, Denburg, and Bechara (2003)
Integral Affective Responses as Somatic found that compared to control patients with
Markers brain lesions outside those associated with emo-
tional responding, patients with VMPC dam-
In an influential series of studies, patients ages score lower on various measures of emo-
with emotional deficits related to damages in the tional and social intelligence and various di-
ventromedial prefrontal cortex area (VMPC) mensions of social functioning such as
were found to perform more poorly than normal postlesion employment status, social status, and
and presumably emotionally functional partici- interpersonal relationships.
pants on a task known as the Iowa Gambling Although these findings have been widely in-
Task (IGT; Bechara, Damasio, Damasio, & terpreted as demonstrating the importance of emo-
Anderson, 1994; Bechara, Damasio, Tranel, & tions for (materially) rational decision making,
Damasio, 1997; Damasio, 1994). The IGT in- more recent studies have challenged their original
volves repeated drawings from four decks of interpretation. Findings by Maia and McClelland
cards. Two decks have higher nominal card (2004) suggest that in the IGT task normal partic-
value but lower expected value because of se- ipants are more conscious of decks’ structure than
vere occasional penalties, and two decks have previously thought, challenging the notion that
lower nominal card value but higher expected somatic markers could guide choice uncon-
value because of lesser penalties. Compared to sciously. Findings by Fellows and Farah (2005)
normal participants, VMPC patients were found suggest that the poor performance of VMPC pa-
to draw more from the riskier and less advan- tients in the original IGT studies may be because
tageous decks, resulting in lower monetary per- of an inability to reverse learned associations ac-
formance (Bechara, Damasio, Damasio, & quired in the early rounds of the game, not the
Anderson, 1994). Over time, normal partici- inability to encode somatic markers. In tasks that
pants also exhibited heightened levels of gal- do not involve outcome feedback— one-shot
vanic skin response whenever they were about choices among risky lotteries, intertemporal pref-
to choose from the risky decks, and this appar- erences, and behavior in wealth-sharing games—
ently even before they could consciously recog- VMPC patients and normal subjects exhibit com-
nize the structure of the desks; VMPC patients parable levels of risk-seeking and impulsivity
SPECIAL ISSUE: EMOTION AND RATIONALITY 167

(Leland & Grafman, 2005). This suggests that simply that integral-affect-motivated approach
presumably emotionally impaired VMPC patients and avoidance—that is, affective behavioral reg-
are not inherently more risk-seeking and impul- ulation—is very sensitive to emotion-producing
sive; rather they differ in how they respond to and outcome feedback. Moreover, the contrast be-
learn from outcome feedback. Moreover, studies tween the original Damasio findings and more
that have used psychopathy as an alternative op- recent Shiv and colleagues findings shows that it
erationalization of emotional deficit have uncov- is not possible to draw firm conclusions from
ered inconsistent effects of psychopathy on per- these studies about the general material rationality
formance on the IGT (e.g., Blair & Cipolotti, or irrationality of integral affect as a proxy for
2000; Losel & Schmucker, 2004; Mitchell, value: it depends on the correlation between the
Colledge, Leonard, & Blair, 2002; Schmitt, emotional responses to the target and its criterion
Brinkley, & Newman, 1999). The lack of paral- value— correlation that is under the researcher’s
lelism between results obtained with VMPC pa- control.
tients and with psychopaths raises the possibility
that an emotional deficit may not be the only
factor at work in the original Damasio findings. Integral Affective Responses in Social
Finally, even if the original VMPC/IGT findings and Economic Interactions
were correct, the superior economic performance
Although most of the empirical evidence on
of emotionally functional participants does not, by
emotion and rationality pertains to individuals
itself, establish the superiority of emotion-based
considered in isolation, it is important to keep in
decision-making. One should note that in the IGT
mind that humans are also part of social groups
there is a negative correlation between the riski-
such as families, organizations, communities,
ness of the decks and their long-term expected
markets, and societies. As highlighted by the
monetary value. Therefore, negative emotional re-
notion of ecological rationality, the desirability
sponses to the risky decks’ penalties are good
of emotions should also be assessed in terms of
predictors of the decks’ actual undesirability. In
how they influence an individual’s ability to
other domains, however, the correlation between
function as a member of a group and how they
risks and expected returns is in fact positive. In
affect the group’s overall welfare (see also
such situations, emotional apprehension toward
Clore, 2005 and Loewenstein & Small, 2007).
taking risks may be materially detrimental. Con-
Emotional responses appear to play an impor-
sistent with this reasoning, emotionally deficient
tant and mostly positive role in the regulation of
VMPC patients (who do poorly in the original
social and moral behavior (Eisenberg, 2000).
Damasio task) were found to perform better than
Emotional responses are a necessary component
emotionally functional control subjects in a re-
of empathic responses, an established determinant
peated investment task where investment risk was
of prosocial behavior (Eisenberg et al., 1994;
associated with higher expected returns. Emotion-
Eisenberg & Miller, 1987). More importantly,
ally functional subjects tend to be overly appre-
emotions are very sensitive to the fulfillment or
hensive about taking risks, especially after incur-
violation of social and moral norms such as hon-
ring a loss (Shiv, Loewenstein, Bechara, Damasio,
esty, reciprocity, and loyalty. Guilt and shame, for
& Damasio, 2005).
example, typically arise from the perception of
The overall evidence about the somatic marker
having transgressed such norms while pride often
hypothesis seems to warrant only a weaker (and
arises from the perception of having fulfilled or
relatively mundane) version of this hypothesis. It
exceeded them. Anger and indignation often arise
is well established that integral affective responses
from observing others transgressing social and
to a target that are positive generally trigger ap-
moral norms, whereas gratitude and admiration
proach tendencies, whereas those that are negative
are generally reserved for those who fulfill or
generally trigger avoidant tendencies,6 even if de-
scriptions of the targets and their cognitive assess- exceed them (Ortony, Clore, & Collins, 1988).
ments are held constant (e.g., Abelson, Kinder, There is growing evidence that emotional re-
Peters, & Fiske, 1982). In other words, integral sponses are not just consequences of social and
affective responses often serve as distinct proxies
for value. What the Damasio studies, along with 6
One exception is anger, which generally triggers a drive
other studies (Ratner & Herbst, 2005), suggest is to confront the anger-producing object.
168 PHAM

moral appraisals but also inform these appraisals. The ecological rationality of these responses
For example, people tend to make more severe transpires as well in experimental studies of eco-
moral judgments and become less tolerant of nomic interactions in which participants play
moral violations if the level of repulsion that they “games” where they interact with one another
experience when making moral judgments is arti- along a specific set of rules to maximize some
ficially increased through an incidental affect economic payoff (see Goette & Huffman, 2007,
manipulation (Schnall, Haidt, & Clore, 2006; for a review). For example, in the ultimatum
Trafimow, Bromgard, Finlay, & Ketelaar, 2005) game, two players have to split a given amount of
or through hypnotic conditioning (Wheatley & money (e.g., $20). One player, the proposer,
Haidt, 2005). Conversely, they become more tol- makes an offer (e.g., $5 for you/$15 for me),
erant of moral violations if they are led to misat- which the other player, the receiver, either accepts
tribute part of their integral affective responses to or rejects. If the offer is accepted, the money is
these violations to external sources (Trafimow et split accordingly; if it is rejected, neither player
al., 2005). These findings suggest that people infer receives anything. According to economic theory,
the severity of moral violations in part from their a materially rational receiver would accept any
spontaneous affective responses to these viola- offer greater than zero because this offer would
tions (see also Kahneman, Schkade, & Sunstein, always be more attractive than rejecting and re-
1998). Spontaneous emotional responses may in ceiving nothing. Anticipating this, a materially
fact be the primary means by which many moral (and logically) rational proposer would make the
dilemmas are evaluated, including some that are smallest offer possible (e.g., $0.05 for you/$19.95
very difficult to solve using pure logic (Haidt, for me), knowing that it would be accepted. The
2001). Neuroimaging studies indicate that parts of logically and materially rational “equilibrium” of
the brain typically associated with emotions are this game is thus a split in which the proposer
uniquely engaged by certain moral dilemmas and keeps most of the money and the receiver receives
that it is this emotional engagement that drives the a positive residual. Numerous studies have shown
resolution of the dilemma (Greene & Haidt, 2002; that this prediction is rarely fulfilled (Camerer,
Greene, Sommerville, Nystrom, Darley, & Cohen, 2003; Roth, 1995). Offers of less than 20% of the
2001; Moll et al., 2002, 2005). total amount typically have a 50% chance of being
The importance of emotions in social and moral rejected by the receiver. Many receivers would
regulation is also evident from studies of psycho- therefore sacrifice their material self-interest rather
pathic populations. Clinical psychopaths typically than accept an offer they perceive unfair. These
exhibit antisocial personality disorders such as materially irrational rejections appear to be driven
consistent disregard for social norms, pervasive primarily by emotional responses of anger follow-
violations of the rights of others, and a tendency ing unfair offers (Pillutla & Murnighan, 1996;
for aggression and violence (Hare, 1985). Numer- Sanfey, Rilling, Aronson, Nystrom, & Cohen,
ous studies suggest that psychopaths tend to have 2003). Interestingly, however, many proposers
lower baseline levels of emotional activity and seem to anticipate these emotional responses⶿ and
weaker physiological responses to emotional stim- make offers that are more equitable and therefore
ulation, especially with respect to negative stimuli more likely to be accepted. The modal empirical
(Blair & Cipolotti, 2000; Lorber, 2004; Patrick, offer in ultimatum games is 40 –50%, which is
1994; Pham, Philippot, & Rime, 2000). This generally accepted (Camerer, 2003). Therefore, a
lower general emotionality may explain psycho- materially irrational emotional response—anger at
paths’ characteristic lack of guilt, remorse, and an unfair, tough positive offer—and its anticipa-
empathy (Blair, 1995), and therefore their com- tion lead to an ecologically desirable solution, one
mon pattern of antisocial and immoral behavior that leaves both players better off than the dead-
(Blair, 1997; Blair & Cipolotti, 2000). Therefore, end usually produced with materially rational
while emotional responses such as guilt, shame, strategies. This equilibrium is similar to the type
indignation, or empathy entail psychic costs that of social and moral equilibrium discussed above.
may seem materially irrational, these responses Transgressions of social norms (here, fairness)
appear to fulfill an important ecological function: usually elicit angry responses among others and
that of promoting socially and morally desirable activate unpleasant moral emotions such as guilt
behavior, both by the person experiencing the or shame within the self. Anticipation of these
emotion and by others eliciting the emotion. emotions acts a strong deterrent against norm vi-
SPECIAL ISSUE: EMOTION AND RATIONALITY 169

olations, leading to more socially sustainable equi- their material self-interest that emotional re-
libriums. sponses and their anticipations produce these so-
Similar findings emerge with other games. For cially desirable outcomes (Frank, 1988; Howard,
example, in public-good games, players have to 1993).
decide how much to contribute to a collective
good. The more people contribute, the better off The Affective System of Behavioral
the entire community is. However, noncontribut- Regulation and Judgment: Empirical
ing members cannot be easily excluded from the Generalizations and Theoretical
benefits of the public good. In these games, the Conjectures
materially rational course of action is to “free-
ride”: not contribute anything and enjoy the com- This review suggests several empirical gen-
mon good supported by others. Of course, if every eralizations and associated theoretical conjec-
player behaves this way, the system collapses and tures about the affective system of behavioral
the entire community suffers. Research suggests regulation and judgment.
that a potent mechanism for ensuring cooperation Negative emotional states of high intensity
in public good games is to allow players to punish generally interfere with people’s reasoning abil-
free-riders (Fehr & Fischbacher, 2002). Public- ities. This is true of both high activation states
good experiments show that a majority of players such as anxiety and low activation states such as
are willing to incur personal financial costs to intense depression. The only reliable reasoning
punish free-riders even in settings where players benefit of intense emotional states seems to be a
meet only once and where, therefore, there is no greater selective reliance on diagnostic informa-
material incentive to punish—a phenomenon tion under states of high activation. These find-
called altruistic punishment (Fehr & Gachter, ings seem to reflect an affective system of be-
2002). Again, it appears that the main driver of havioral regulation, which under high activation
these altruistic—and materially irrational— mobilizes responses based on quick assess-
punishments is the anger that punishers feel to- ments of diagnostic features of the situation
ward the free-riders (de Quervain et al., 2004; rather than more careful considerations of po-
Fehr & Gachter, 2002). Interestingly, when such tential consequences of alternative courses of
punishment opportunities are present, cooperation action. This conjecture is consistent with other
increases over time, whereas when they are ab- findings showing that judgments and decisions
sent, cooperation decreases over time. Therefore, based on integral affective responses are less
as with the ultimatum games, in public-good resource-demanding and reached faster than
games, emotional responses to norm-violating be- those based on reason-based assessments. In
havior again produce norm-enforcement behav- contrast, the effects of intensely pleasant emo-
iors that may appear materially irrational in the tional states (e.g., intense joy or pride) are less
short term but lead to more ecologically rational well understood. If the proposed resource-
outcomes in the long run. mobilization conjecture is correct, intense pos-
Emotional responses and their anticipation ap- itive emotional states might have less influence
pear to have similar effects in various other games on reasoning processes because situations con-
such as the dictator game (a variant of the ultima- ducive of such states typically require less be-
tum game), the classic prisoner’s dilemma game, havioral adjustment.
and various “trust” games that involve costly in- Milder incidental states of sadness generally
vestments in a partnership that may or may not be promote a more systematic, data-driven, and
honored by the other party (Camerer, 2003; analytical form of reasoning, whereas positive
Dawes, McTavish, & Shaklee, 1977; Fehr & mood states generally promote a less system-
Fischbacher, 2002; Lehmann, 2001). In many eco- atic, more top-down, but more flexible and cre-
nomic interactions, emotional responses to the vi- ative form of reasoning. Whereas the effects of
olation or fulfillment of norms such as fairness, intense emotional states on reasoning may re-
equity, and trust, and anticipations of these re- flect the requirements of fast response mobili-
sponses by the various players seem to promote zation, the effects of milder sadness and posi-
ecologically desirable cooperation and norm- tive moods seem to arise from a signaling func-
consistent behavior. Some theorists suggest that it tion of these milder states (Schwarz, 2002).
is precisely because they make players override From an evolutionary standpoint, positive mood
170 PHAM

states may have served as a signal to engage in typical drive states. Although a person would
more contemplative thoughts and explorative likely forego food or sex to alleviate an intense
behaviors. This conjecture is consistent with the sleep deprivation, an emotionally upset person
finding that positive mood encourages variety- would likely consider a wider range of options
seeking (Kahn & Isen, 1993) and the positive- in order to feel better. This suggests that the
psychology hypothesis that a primary function affective system of behavioral regulation oper-
of positive affect is to broaden the thought- ates at a more abstract level than the various
action repertoire in order to build future re- physiological drive systems.
sources (Fredrickson, 1998). In contrast, be- Although the evidence does not support a
cause sadness typically highlights a discrepancy radical form of the somatic marker hypothesis,
between a desired state and a current state there is considerable support across disciplines
(Higgins, 1987), states of sadness may signal a for a more benign version of this hypothesis,
need to reassess the situation and analyze envi- which is that (a) integral emotional responses to
ronmental inputs more carefully, which may objects are often interpreted as signals of these
explain occasional findings of depressive real- objects’ value and (b) these integral responses
ism among individuals whose depression is not are very sensitive to recent experiences with the
too severe on tasks that are interpretable in target object. Besides being typically faster and
terms of situational reassessment. Tasks that resource-efficient, affective assessments of
cannot be interpreted in those terms may not be value differ from typical reason-based assess-
amenable to depressive realism. ments in important respects. Affective assess-
Incidental emotional states are often misat- ments of value tend to be more extreme and
tributed to attentional objects and tend to distort polarized and more sensitive to recent concrete
beliefs about these objects in an assimilative outcomes. This phenomenon may again reflect
fashion, especially if the emotional states are the response-mobilization function of emotions,
perceived to be representative of the objects. which are meant to motivate and redirect be-
Although logically irrational, this phenomenon havior if necessary. Although occasionally log-
may be rooted in an ecologically rational prop- ically irrational, the greater polarization of af-
erty of emotions. If an original function of emo- fect-based valuations may thus have been evo-
tion was to promote fast responses to the envi- lutionary adaptive.
ronment, it would be ecologically efficient to Valuations based on integral affect tend to be
assume by default that one’s emotional experi- myopic, emphasizing immediate hedonic con-
ences are genuine responses to the focal objects, sequences (positive or negative) over future
especially if these feeling experiences appear consequences. The affective system of valua-
representative of these objects. The assimilation tion and behavioral regulation seems to be a
of beliefs toward the content of emotions may system of the present. This anchoring in the
have had the purpose of promoting faster be- present may have been ecologically rational in a
havioral responses by promoting intrapsychic world where the current value of objects was
consistency (Pham, 2004). This is consistent generally positively correlated with their long-
with findings showing that integral affective term value. However, in today’s world, this
responses to target objects also steer thoughts in anchoring in the present would produce mate-
the direction of initial affective responses. rially irrational valuations in domains where the
Negative emotional states that are intense correlation between short-term hedonic value
generally disrupt self-control, but do not neces- and long-term criterion value is negative. One
sarily increase risk seeking. This phenomenon of these domains is the investment domain,
again seems to reflect a response mobilization where the fear of losses may inhibit investments
function of intense negative emotional states. in risky prospects such as stocks, even though
Like drive states such as hunger, tiredness, and stocks generally have higher expected re-
sexual arousal, intense negative emotional turns—a phenomenon known as myopic loss
states seem to direct behavior toward goal ob- aversion (Benartzi & Thaler, 1995). Still, one
jects that appear capable of alleviating these should be careful not to overgeneralize this case
states. However, the goal objects being pursued of material irrationality, for it is restricted to
under intense negative emotional states appear those domains where the correlation between
to be less specific than those pursued under short-term hedonic value and long-term crite-
SPECIAL ISSUE: EMOTION AND RATIONALITY 171

rion value is negative. One should bear in mind value of objects is thus determined by their
that in many domains the correlation is likely identity, not by their distribution. This conjec-
positive. ture would explain why affective valuations are
The above caveat applies more broadly to particularly sensitive to concrete representa-
alleged demonstrations of the material rational- tions of exemplars.
ity or irrationality of integral affective responses Affective valuations also tend to be more
as proxies for value. For example, some studies consistent, both interpersonally and intraperson-
suggest that VMPC-related emotional deficits ally. This consistency may reflect a general re-
decrease performance (and economic welfare) liability and stability of integral feeling re-
in the Iowa Gambling Task, and other findings sponses as a basis for judgment. If the affective
suggest that the same deficits enhance perfor- system assigns value primarily through object
mance in other investment tasks. In any partic- categorization, evolutionary pressures would
ular study, appearances of rationality or irratio- have demanded that these categorizations be
nality will be driven by the correlation between consistent both within and across individuals.
the emotional responses to the target (conveyed For example, an ecological benefit of consistent
for instance through outcome feedback) and its of affective valuation across individuals would
criterion value. One should not forget that this be improved social coordination. A society
correlation is a parameter under the researcher’s would not be stable if object-emotion mappings
control. A more pertinent question, therefore, is (e.g., what is considered shameful) varied
whether integral affective responses are good widely among its members. However, the eco-
predictors of value in a broader ecological logically rational consistency of affective valu-
sense. As discussed by Pham (2004) using ations comes at a cost. As is evident in patho-
Brunswick’s (1952) terminology, this would de- logical phobias, affect-producing categoriza-
pend on the relative magnitude of two correla- tions are difficult to unlearn, which will produce
tion coefficients in the representative universe instances of logical or material irrationality.
of objects to be evaluated: (a) the correlation Valuations based on integral affect are also
between the integral feelings elicited by this more relativistic and reference-dependent. They
universe of objects and these objects’ true cri- are very sensitive, for instance, to spontaneous
terion value (the ecological validity of the feel- comparisons with outcome counterfactuals and
ings) and (b) the correlation between other to social comparisons. One conjecture is that the
available proxies of value and the objects’ cri- affective system of valuation is mostly ordinal
terion value (the ecological validity of alterna- as opposed to cardinal (i.e., interval scaled).
tive bases of evaluation). The ecological merits This would be sensible from an evolutionary
of integral feelings as proxies of value would standpoint, if affective valuations were primar-
additionally depend on the relative reliabilities ily meant to support behavioral choices (do A,
of integral feelings and alternative inputs be- B, or C), which require only ordinal judgments.
cause, everything else equal, more reliable pre- The affective system may be more concerned
dictors make better proxies. Unfortunately, em- with the relative desirability ordering of alter-
pirical estimates of these correlations and reli- native states of the world or alternative courses
abilities across broadly representative sets of of actions than with their absolute desirability.
targets have yet to be documented. An ordinal affective system of valuation may
Valuations based on integral affect also tend also explain why people do not place much
to be insensitive to scale, whether scale refers to weight on the duration of hedonic experiences
the quantitative magnitude of the stimulus or in retrospective valuations these experiences
the probability that surrounds it. Although it (Fredrickson & Kahneman, 1993).
seems to violate logical rationality, this scale Finally, a major function of emotion appears
insensitivity may indicate that the affective sys- to be the promotion of socially and morally
tem assigns value primarily through processes desirable behavior and the deterrence of unde-
of categorization: target objects or events are sirable behavior. Emotions are a necessary com-
mapped onto existing categories or schemas ponent of empathic responses, which are impor-
and, depending on which category or schema is tant drivers of prosocial behavior. Emotions are
activated, a particular value-laden emotional re- also very sensitive to the fulfillment or violation
sponse is triggered. In the affective system, the of social and moral norms, and therefore an
172 PHAM

essential component of social and moral self- Batra, R., & Stayman, D. M. (1990). The role of
regulation. Finally, in social and economic in- mood in advertising effectiveness. Journal of Con-
teractions, emotions promote the fulfillment and sumer Research, 17, 203–214.
Bechara, A. (2004). The role of emotion in decision-
enforcement of social and moral obligations by
making: Evidence from neurological patients with
overriding the players’ material self-interests. In orbitofrontal damage. Brain and Cognition, 55,
this sense, the emotional person is not irrational, 30 – 40.
but ecologically rational. Therefore, although Bechara, A., Damasio, A. R., Damasio, H., & Ander-
the physiological drive systems (fatigue, pain, son, S. W. (1994). Insensitivity to future conse-
hunger, sexual arousal, etc.) may be inherently quences following damage to human prefrontal
selfish (Loewenstein, 1996), the affective sys- cortex. Cognition, 50, 7–15.
tem (love, pride, anger, guilt, shame, etc.) may Bechara, A., Damasio, H., Tranel, D., & Damasio,
A. R. (1997). Deciding advantageously before
be distinctively social and moral. knowing the advantageous strategy. Science, 275,
1293–1295.
Benartzi, S., & Thaler, R. H. (1995). Myopic loss
References aversion and the equity premium puzzle. Quarterly
Journal of Economics, 110, 73–92.
Abelson, R. P., Kinder, D. R., Peters, M. D., & Fiske, Benassi, V. A., & Mahler, H. I. M. (1985). Contin-
S. T. (1982). Affective and semantic components gency judgments by depressed college-students—
in political person perception. Journal of Person- Sadder but not always wiser. Journal of Personal-
ality and Social Psychology, 42, 619 – 630. ity and Social Psychology, 49, 1323–1329.
Albarracin, D., & Kumkale, G. T. (2003). Affect as Blair, R. J. R. (1995). A cognitive developmental-
information in persuasion: A model of affect iden- approach to morality—Investigating the psycho-
tification and discounting. Journal of Personality path. Cognition, 57, 1–29.
and Social Psychology, 84, 453– 469. Blair, R. J. R. (1997). Moral reasoning and the child
Allan, L. G., Siegel, S., & Hannah, S. (2007). The sad with psychopathic tendencies. Personality and In-
truth about depressive realism. Quarterly Journal dividual Differences, 22, 731–739.
of Experimental Psychology, 60, 482– 495. Blair, R. J. R., & Cipolotti, L. (2000). Impaired social
Alloy, L. B., & Abramson, L. Y. (1979). Judgment of response reversal—A case of “acquired socio-
contingency in depressed and nondepressed stu- path.” Brain, 123, 1122–1141.
dents—Sadder but wiser. Journal of Experimental Bless, H., Bohner, G., Schwarz, N., & Strack, F.
Psychology-General, 108, 441– 485. (1990). Mood and persuasion—A cognitive re-
Alloy, L. B., & Abramson, L. Y. (1988). Depressive sponse analysis. Personality and Social Psychol-
realism: Four theoretical perspectives. In L. B. ogy Bulletin, 16, 331–345.
Alloy (Ed.), Cognitive processes in depression. Bless, H., Mackie, D. M., & Schwarz, N. (1992).
New York: The Guilford Press. Mood effects on attitude judgments—Independent
effects of mood before and after message elabora-
Alloy, L. B., & Ahrens, A. H. (1987). Depression and
tion. Journal of Personality and Social Psychol-
pessimism for the future—Biased use of statisti-
ogy, 63, 585–595.
cally relevant information in predictions for self
Bless, H., Schwarz, N., Clore, G. L., Golisano, V., &
versus others. Journal of Personality and Social
Rabe, C. (1996). Mood and the use of scripts: Does
Psychology, 52, 366 –378. a happy mood really lead to mindlessness? Journal
Ariely, D., & Loewenstein, G. (2005). The heat of the of Personality and Social Psychology, 71, 665–
moment: The effect of sexual arousal on sexual 679.
decision making. Journal of Behavioral Decision Bodenhausen, G. V. (1993). Emotions, arousal, and
Making, 19, 87–98. stereotype-based discrimination: A heuristic model
Avnet, T., & Pham, M. T. (2004). Should I trust my of affect and stereotyping. In D. M. Mackie & D.
feelings or not? The metacognition of affect-as- L. Hamilton (Eds.), Affect, cognition, and stereo-
information. Unpublished manuscript, Columbia typing (pp. 13–37). San Diego, CA: Academic
University. Press.
Bacon, S. J. (1974). Arousal and range of cue utili- Bodenhausen, G. V., Kramer, G. P., & Suesser, K.
zation. Journal of Experimental Psychology, 102, (1994). Happiness and stereotypic thinking in so-
81– 87. cial judgment. Journal of Personality and Social
Bar-On, R., Tranel, D., Denburg, N. L., & Bechara, Psychology, 66, 621– 632.
A. (2003). Exploring the neurological substrate of Bodenhausen, G. V., Sheppard, L. A., & Kramer,
emotional and social intelligence. Brain, 126, G. P. (1994). Negative affect and social judg-
1790 –1800. ment—The differential impact of anger and sad-
SPECIAL ISSUE: EMOTION AND RATIONALITY 173

ness. European Journal of Social Psychology, 24, nal of Personality and Social Psychology, 66,
45– 62. 819 – 829.
Bouffard, J. A. (2002). The influence of emotion on deq Uervain, D. J. F., Fischbacher, U., Treyer, V.,
rational decision making in sexual aggression. Schelthammer, M., Schnyder, U., Buck, A., et al.
Journal of Criminal Justice, 30, 121–134. (2004). The neural basis of altruistic punishment.
Bower, G. H. (1981). Mood and memory. American Science, 305, 1254 –1258.
Psychologist, 36, 129 –148. DeSteno, D., Petty, R. E., Wegener, D. T., & Rucker,
Brandstatter, E., Kuhberger, A., & Schneider, F. D. D. (2000). Beyond valence in the perception of
(2002). A cognitive-emotional account of the likelihood: The role of emotion specificity. Jour-
shape of the probability weighting function. Jour- nal of Personality and Social Psychology, 78, 397–
nal of Behavioral Decision Making, 15, 79 –100. 416.
Brunswick, E. (1952). The conceptual framework of Dobson, K., & Franche, R.-L. (1989). A conceptual
psychology. Chicago: University of Chicago Press. and empirical review of the depressive realism
Camerer, C. F. (2003). Behavioral game theory: Ex- hypothesis. Canadian Journal of Behavioural Sci-
periments on strategic interaction. Princeton: ence-Revue Canadienne Des Sciences Du Com-
Princeton University Press. portement, 21, 418 – 433.
Clore, G. L. (2005). For love or money: Some emo- Dommermuth, W. P., & Millard, W. J. (1967). Con-
tional foundations of rationality. Chicago Kent sumption coincidence in product evaluation. Jour-
Law Review, 80, 1151–1165. nal of Marketing Research, 4, 388 –390.
Cohen, J. B., Pham, M. T., & Andrade, E. B. (2007). Dunning, D., & Story, A. L. (1991). Depression,
The nature and role of affect in consumer judg- realism, and the overconfidence effect - Are the
ment and decision making. In C. P. Haugtvedt, sadder wiser when predicting future actions and
P. M. Herr, & F. R. Kardes (Eds.), Handbook of events. Journal of Personality and Social Psychol-
consumer psychology. Mahwah, NJ: Erlbaum. ogy, 61, 521–532.
Conway, M., & Giannopoulos, C. (1993). Dysphoria Dutton, D. G., & Aron, A. P. (1974). Some evidence
for heightened sexual attraction under conditions
and decision-making—Limited information use
of high anxiety. Journal of Personality and Social
for evaluations of multiattribute targets. Journal of
Psychology, 30, 510 –517.
Personality and Social Psychology, 64, 613– 623.
Eisenberg, N. (2000). Emotion, regulation, and moral
Crepaz, N., & Marks, G. (2001). Are negative affec-
development. Annual Review of Psychology, 51,
tive states associated with HIV sexual risk behav-
665– 697.
iors? A meta-analytic review. Health Psychol-
Eisenberg, N., Fabes, R. A., Murphy, B., Karbon, M.,
ogy, 20, 291–299.
Maszk, P., Smith, M., et al. (1994). The relations
Damasio, A. R. (1994). Descartes’ error: Emotion, of emotionality and regulation to dispositional and
reason, and the human brain. New York: Putnam. situational empathy-related responding. Journal of
Darke, P. R., Chattopadhyay, A., & Ashworth, L. Personality and Social Psychology, 66, 776 –797.
(2006). Going with your “gut feeling”: The impor- Eisenberg, N., & Miller, P. A. (1987). The relation of
tance and functional significance of discrete affect empathy to prosocial and related behaviors. Psy-
in consumer choice. Journal of Consumer Re- chological Bulletin, 101, 91–119.
search, 33, 322–328. Elster, J. (1999). Alchemies of the mind: Rationality
Darke, S. (1988a). Anxiety and working memory and the emotions. Cambridge, UK: Cambridge
capacity. Cognition and Emotion, 2, 145–154. University Press.
Darke, S. (1988b). Effects of anxiety on inferential Epstein, S. (1990). Cognitive-experiential self-
reasoning task-performance. Journal of Personal- theory. In L. A. Pervin (Ed.), Handbook of person-
ity and Social Psychology, 55, 499 –505. ality: Theory and research (pp. 165–192). New
Dawes, R. M., McTavish, J., & Shaklee, H. (1977). York/London: The Guilford Press.
Behavior, communication, and assumptions about Etcoff, N. (1999). Survival of the prettiest. New
other peoples behavior in a commons dilemma York: Doubleday.
situation. Journal of Personality and Social Psy- Faber, R. J., & Christenson, G. A. (1996). In the
chology, 35, 1–11. mood to buy: Differences in the mood states ex-
De Houwer, J., Thomas, S., & Baeyens, F. (2001). perienced by compulsive buyers and other con-
Associative learning of likes and dislikes: A re- sumers. Psychology & Marketing, 13, 803– 819.
view of 25 years of research on human evaluative Fehr, E., & Fischbacher, U. (2002). Why social pref-
conditioning. Psychological Bulletin, 127, 853– erences matter—The impact of non-selfish motives
869. on competition, cooperation and incentives. Eco-
Denes-Raj, V., & Epstein, S. (1994). Conflict be- nomic Journal, 112, C1–C33.
tween intuitive and rational processing—When Fehr, E., & Gachter, S. (2002). Altruistic punishment
people behave against their better judgment. Jour- in humans. Nature, 415, 137–140.
174 PHAM

Fellows, L. K., & Farah, M. J. (2005). Different nondepressed individuals. Cognition and Emo-
underlying impairments in decision-making fol- tion, 2, 185–200.
lowing ventromedial and dorsolateral frontal lobe Greene, J., & Haidt, J. (2002). How (and where) does
damage in humans. Cerebral Cortex, 15, 58 – 63. moral judgment work? Trends in Cognitive Sci-
Fessler, D. M. T., Pillsworth, E. G., & Flamson, T. J. ences, 6, 517–523.
(2004). Angry men and disgusted women: An evo- Greene, J., Sommerville, R. B., Nystrom, L. E.,
lutionary approach to the influence of emotions on Darley, J. M., & Cohen, J. D. (2001). An fMRI
risk taking. Organizational Behavior and Human investigation of emotional engagement in moral
Decision Processes, 95, 107–123. judgment. Science, 293, 2105–2108.
Fetherstonhaugh, D., Slovic, P., Johnson, S. M., & Greene, T. R., & Noice, H. (1988). Influence of
Friedrich, J. (1997). Insensitivity to the value of positive affect upon creative-thinking and prob-
human life: A study of psychophysical numbing. lem-solving in children. Psychological Re-
Journal of Risk and Uncertainty, 14, 283–300. ports, 63, 895– 898.
Fiedler, K. (1988). Emotional mood, cognitive style, Haidt, J. (2001). The emotional dog and its rational
and behavior regulation. In K. Fiedler & J. Forgas tail: A social intuitionist approach to moral judg-
(Eds.), Affect, cognition, and social behavior (pp. ment. Psychological Review, 108, 814 – 834.
100 –119). Toronto: C. J. Hogrefe. Hare, R. D. (1985). Comparison of procedures for the
Forgas, J. P. (1995). Mood and judgment: The affect assessment of psychopathy. Journal of Consulting
infusion model (AIM). Psychological Bulletin, and Clinical Psychology, 53, 7–16.
117, 39. Hartlage, S., Alloy, L. B., Vazquez, C., & Dykman,
Forgas, J. P. (1998). On being happy and mistaken: B. (1993). Automatic and effortful processing in
Mood effects on the fundamental attribution error. depression. Psychological Bulletin, 113, 247–278.
Journal of Personality and Social Psychology, 75, Hendrickx, L., Vlek, C., & Oppewal, H. (1989).
318 –331. Relative importance of scenario information and
Foster, C. A., Witcher, B. S., Campbell, W. K., & frequency information in the judgment of risk.
Green, J. D. (1998). Arousal and attraction: Evi-
Acta Psychologica, 72, 41– 63.
dence for automatic and controlled processes.
Higgins, E. T. (1987). Self-discrepancy—A theory
Journal of Personality and Social Psychology, 74,
relating self and affect. Psychological Review, 94,
86 –101.
319 –340.
Frank, R. H. (1988). Passions within reason: The
Hirshleifer, D., & Shumway, T. (2003). Good day
strategic role of the emotions. New York: Norton.
sunshine: Stock returns and the weather. Journal
Fredrickson, B. L. (1998). What good are positive
emotions? Journal of General Psychology, 2, 300 – of Finance, 58, 1009 –1032.
319. Hirt, E. R., Melton, R. J., McDonald, H. E., &
Fredrickson, B. L., & Kahneman, D. (1993). Dura- Harackiewicz, J. M. (1996). Processing goals, task
tion neglect in retrospective evaluations of affec- interest, and the mood-performance relationship:
tive episodes. Journal of Personality and Social A mediational analysis. Journal of Personality and
Psychology, 65, 45–55. Social Psychology, 71, 245–261.
Frijda, N. H. (1988). The laws of emotion. American Hockey, G. R. J., Maule, A. J., Clough, P. J., &
Psychologist, 43, 349 –358. Bdzola, L. (2000). Effects of negative mood states
Fry, P. S. (1975). Affect and resistance to temptation. on risk in everyday decision making. Cognition &
Developmental Psychology, 11, 466 – 472. Emotion, 14, 823– 855.
Goette, L. & Huffman, D. (2007). Affect and the Hockey, R. (1973). Changes in information-selection
motivational foundations of social capital. Review patterns in multisource monitoring as a function of
of General Psychology, 11, 142–154. induced arousal shifts. Journal of Experimental
Goldberg, M. E., & Gorn, G. J. (1987). Happy and Psychology, 101, 35– 42.
sad TV programs—How they affect reactions to Howard, N. (1993). The role of emotions in multior-
commercials. Journal of Consumer Research, 14, ganizational decision-making. Journal of the Op-
387– 403. erational Research Society, 44, 613– 623.
Gorn, G. J., Goldberg, M. E., & Basu, K. (1993). Hsee, C. K., & Kunreuther, H. C. (2000). The affec-
Mood, awareness, and product evaluation. Journal tion effect in insurance decisions. Journal of Risk
of Consumer Psychology, 2, 237–256. and Uncertainty, 20, 141–159.
Gorn, G. J., Pham, M. T., & Sin, L. Y. (2001). When Hsee, C. K., & Rottenstreich, Y. (2004). Music,
arousal influences ad evaluation and valence does pandas, and muggers: On the affective psychology
not (and vice versa). Journal of Consumer Psy- of value. Journal of Experimental Psychology-
chology, 11, 43–55. General, 133, 23–30.
Gotlib, I. H., McLachlan, A. L., & Katz, A. N. Hsee, C. K., Zhang, J., Yu, F., & Xi, Y. (2003). Lay
(1988). Biases in visual attention in depressed and rationalism and inconsistency between predicted
SPECIAL ISSUE: EMOTION AND RATIONALITY 175

experience and decision. Journal of Behavioral analysis of dollar responses to public issues. Jour-
Decision Making, 16, 257–272. nal of Risk and Uncertainty, 19, 203–235.
Humphreys, M. S., & Revelle, W. (1984). Personal- Kahneman, D., Schkade, D., & Sunstein, C. R.
ity, motivation, and performance—A theory of the (1998). Shared outrage and erratic awards: The
relationship between individual-differences and in- psychology of punitive damages. Journal of Risk
formation-processing. Psychological Review, 91, and Uncertainty, 16, 49 – 86.
153–184. Kamstra, M. J., Kramer, L. A., & Levi, M. D. (2003).
Isen, A. M. (2001). An influence of positive affect on Winter blues: A SAD stock market cycle. Ameri-
decision making in complex situations: Theoretical can Economic Review, 93, 324 –343.
issues with practical implications. Journal of Con- Keinan, G. (1987). Decision-making under stress—
sumer Psychology, 11, 75– 85. Scanning of alternatives under controllable and
Isen, A. M., & Daubman, K. A. (1984). The influence uncontrollable threats. Journal of Personality and
of affect on categorization. Journal of Personality Social Psychology, 52, 639 – 644.
and Social Psychology, 47, 1206 –1217. Keller, P. A., Lipkus, I. M., & Rimer, B. K. (2002).
Isen, A. M., Daubman, K. A., & Nowicki, G. P. Depressive realism and health risk accuracy: The
(1987). Positive affect facilitates creative problem- negative consequences of positive mood. Journal
solving. Journal of Personality and Social Psy- of Consumer Research, 29, 57– 69.
chology, 52, 1122–1131. Kirkpatrick, L. A., & Epstein, S. (1992). Cognitive
Isen, A. M., & Means, B. (1983). The influence of experiential self-theory and subjective-probabili-
positive affect on decision-making strategy. Social ty—Further evidence for 2 conceptual systems.
Cognition, 2, 18 –31. Journal of Personality and Social Psychology, 63,
Isen, A. M., Niedenthal, P. M., & Cantor, N. (1992). 534 –544.
An influence of positive affect on social categori- Knapp, A., & Clark, M. S. (1991). Some detrimental
zation. Motivation and Emotion, 16, 65–78. effects of negative mood on individuals ability to
Isen, A. M., & Shalker, T. E. (1982). The effect of solve resource dilemmas. Personality and Social
Psychology Bulletin, 17, 678 – 688.
feeling state on evaluation of positive, neutral, and
Kruglanski, A. W. (1989). The psychology of being
negative stimuli—When you accentuate the posi-
“right”: The problem of accuracy in social percep-
tive, do you eliminate the negative. Social Psy-
tion and cognition. Psychological Bulletin, 106,
chology Quarterly, 45, 58 – 63.
395– 409.
Isen, A. M., Shalker, T. E., Clark, M., & Karp, L.
Landman, J. (1987). Regret and elation following
(1978). Affect, accessibility of material in mem-
action and inaction: Affective responses to positive
ory, and behavior-cognitive loop. Journal of Per-
versus negative outcomes. Personality and Social
sonality and Social Psychology, 36, 1–12. Psychology Bulletin, 13, 524 –536.
Johnson, E. J., Hershey, J., Meszaros, J., & Langlois, J. H., Kalakanis, L., Rubenstein, A. J.,
Kunreuther, H. (1993). Framing, probability dis- Larson, A., Hallam, M., & Smoot, M. (2000).
tortions, and insurance decisions. Journal of Risk Maxims or myths of beauty? A meta-analytic and
and Uncertainty, 7, 35–51. theoretical review. Psychological Bulletin, 126,
Johnson, E. J., & Tversky, A. (1983). Affect, gener- 390 – 423.
alization, and the perception of risk. Journal of Lazarus, R. S. (1991). Emotion and adaptation. New
Personality and Social Psychology, 45, 20 –31. York: Oxford University Press.
Kahn, B., & Isen, A. M. (1993). The influence of LeDoux, J. E. (1996). The emotional brain: The
positive affect on variety seeking among safe, en- mysterious underpinnings of emotional life. New
joyable products. Journal of Consumer Re- York: Simon & Schuster.
search, 20, 257–270. Lee, L., & Ariely, D. (2006). In search of homo
Kahneman, D. (1994). New challenges to the ratio- economicus: Preference consistency, emotions,
nality assumption. Journal of Institutional and and cognition. Columbia: Columbia University.
Theoretical Economics-Zeitschrift Fur Die Gesa- Lehmann, D. R. (2001). The impact of altruism and
mte Staatswissenschaft, 150, 18 –36. envy on competitive behavior and satisfaction. In-
Kahneman, D., & Miller, D. T. (1986). Norm theo- ternational Journal of Research in Marketing, 18,
ry—Comparing reality to its alternatives. Psycho- 5–17.
logical Review, 93, 136 –153. Leith, K. P., & Baumeister, R. F. (1996). Why do bad
Kahneman, D., Ritov, I., Jacowitz, K. E., & Grant, P. moods increase self-defeating behavior? Emotion,
(1993). Stated willingness-to-pay for public- risk taking, and self-regulation. Journal of Person-
goods—A psychological perspective. Psychologi- ality and Social Psychology, 71, 1250 –1267.
cal Science, 4, 310 –315. Leland, J. W., & Grafman, J. (2005). Experimental
Kahneman, D., Ritov, I., & Schkade, D. (1999). tests of the somatic marker hypothesis. Games and
Economic preferences or attitude expressions? An Economic Behavior, 52, 386 – 409.
176 PHAM

Leon, M. R., & Revelle, W. (1985). Effects of anx- Mayer, J. D., Gaschke, Y. N., Braverman, D. L., &
iety on analogical reasoning—A test of 3 theoret- Evans, T. W. (1992). Mood-congruent judgment is
ical-models. Journal of Personality and Social a general effect. Journal of Personality and Social
Psychology, 49, 1302–1315. Psychology, 63, 119 –132.
Lerner, J. S., & Keltner, D. (2001). Fear, anger, and McClure, S. M., Laibson, D. I., Loewenstein, G., &
risk. Journal of Personality and Social Psychol- Cohen, J. D. (2004). Separate neural systems value
ogy, 81, 146 –159. immediate and delayed monetary rewards. Sci-
Lerner, J. S., Small, D. A., & Loewenstein, G. ence, 306, 503–507.
(2004). Heart strings and purse strings—Carryover Medvec, V. H., Madey, S. F., & Gilovich, T. (1995).
effects of emotions on economic decisions. Psy- When less is more - counterfactual thinking and
chological Science, 15, 337–341. satisfaction among olympic medalists. Journal of
Lewinsohn, P. M., Mischel, W., Chaplin, W., & Personality and Social Psychology, 69, 603– 610.
Barton, R. (1980). Social competence and depres- Mellers, B. A., Schwartz, A., Ho, K., & Ritov, I.
sion—The role of illusory self-perceptions. Jour- (1997). Decision affect theory: Emotional reac-
nal of Abnormal Psychology, 89, 203–212. tions to the outcomes of risky options. Psycholog-
Loewenstein, G. (1996). Out of control: Visceral ical Science, 8, 423– 429.
influences on behavior. Organizational Behavior Miniard, P. W., Bhatla, S., & Sirdeshmukh, D.
and Human Decision Processes, 65, 272–292. (1992). Mood as a determinant of postconsump-
Loewenstein, G. & Small, D. (2007). The scarecrow tion product evaluations. Journal of Consumer
and the tin man: The vicissitudes of human sym- Psychology, 1, 173–195.
pathy and caring. Review of General Psychology, Mitchell, D. G. V., Colledge, E., Leonard, A., &
11, 112–126. Blair, R. J. R. (2002). Risky decisions and re-
Loewenstein, G., Weber, E. U., Hsee, C. K., & sponse reversal: Is there evidence of orbitofrontal
Welch, N. (2001). Risk as feelings. Psychological cortex dysfunction in psychopathic individuals?
Bulletin, 127, 267–286. Neuropsychologia, 40, 2013–2022.
Lorber, M. F. (2004). Psychophysiology of aggres- Moll, J., de Oliveira-Souza, R., Eslinger, P. J.,
sion, psychopathy, and conduct problems: A meta- Bramati, I. E., Mourao-Miranda, J., Andreiuolo,
analysis. Psychological Bulletin, 130, 531–552. P. A., et al. (2002). The neural correlates of moral
Losel, F., & Schmucker, M. (2004). Psychopathy, sensitivity: A functional magnetic resonance im-
risk taking, and attention: A differentiated test of aging investigation of basic and moral emotions.
the somatic marker hypothesis. Journal of Abnor- Journal of Neuroscience, 22, 2730 –2736.
mal Psychology, 113, 522–529. Moll, J., de Oliveira-Souza, R., Moll, F. T., Ignacio,
Mackie, D. M., & Worth, L. T. (1989). Processing F. A., Bramati, I. E., Caparelli-Daquer, E. M., et al.
deficits and the mediation of positive affect in (2005). The moral affiliations of disgust—A func-
persuasion. Journal of Personality and Social Psy- tional MRI study. Cognitive and Behavioral Neu-
chology, 57, 27– 40. rology, 18, 68 –78.
Maia, T. V., & McClelland, J. L. (2004). A reexam- Monat, A., Averill, J. R., & Lazarus, R. S. (1972).
ination of the evidence for the somatic marker Anticipatory stress and coping reactions under var-
hypothesis: What participants really know in the ious conditions of uncertainty. Journal of Person-
Iowa gambling task. Proceedings of the National ality and Social Psychology, 24, 237–253.
Academy of Sciences, USA of the United States of Msetfi, R. M., Murphy, R. A., Simpson, J., &
America, 101, 16075–16080. Kornbrot, D. E. (2005). Depressive realism and
Mano, H. (1992). Judgments under distress— outcome density bias in contingency judgments:
Assessing the role of unpleasantness and arousal in The effect of the context and intertrial interval.
judgment formation. Organizational Behavior and Journal of Experimental Psychology-General,
Human Decision Processes, 52, 216 –245. 134, 10 –22.
Mano, H. (1994). Risk-taking, framing effects, and Mueller, J. H. (1977). Test anxiety, input modality,
affect. Organizational Behavior and Human Deci- and levels of organization in free-recall. Bulletin of
sion Processes, 57, 38 –58. the Psychonomic Society, 9, 67– 69.
Martin, D. J., Abramson, L. Y., & Alloy, L. B. Mueller, J. H. (1978). Effects of individual-
(1984). Illusion of control for self and others in differences in test anxiety and type of orienting
depressed and nondepressed college-students. task on levels of organization in free-recall. Jour-
Journal of Personality and Social Psychology, 46, nal of Research in Personality, 12, 100 –116.
125–136. Neufeldt, V. (1991). Webster’s New World Dictio-
Mattes, J., & Cantor, J. (1982). Enhancing responses nary (3rd ed.). New York: Prentice Hall.
to television advertisements via the transfer of Nisbett, R., & Ross, L. (1980). Human inference:
residual arousal from prior programming. Journal Strategies and shortcomings of social judgment.
of Broadcasting, 26, 553–566. Englewood Cliffs, NJ: Prentice Hall.
SPECIAL ISSUE: EMOTION AND RATIONALITY 177

Nowlis, S. M., & Shiv, B. (2005). The influence of on switching behavior. Organizational Behavior
consumer distractions on the effectiveness of food- and Human Decision Processes, 96, 23–37.
sampling programs. Journal of Marketing Re- Roese, N. J. (1997). Counterfactual thinking. Psycho-
search, 42, 157–168. logical Bulletin, 121, 133–148.
Oaksford, M., Morris, F., Grainger, B., & Williams, Roth, A. E. (1995). Introduction to experimental eco-
J. M. G. (1996). Mood, reasoning, and central exec- nomics. In J. H. Kagel & A. E. Roth (Eds.), The
utive processes. Journal of Experimental Psycholo- handbook of experimental economics (pp. 3–109).
gy-Learning Memory & Cognition, 22, 476 – 492. Princeton: Princeton University Press.
Ortony, A., Clore, G. L., & Collins, A. (1988). The Rottenstreich, Y., & Hsee, C. K. (2001). Money, kisses,
cognitive structure of emotions. Cambridge: Cam- and electric shocks: On the affective psychology of
bridge University Press risk. Psychological Science, 12, 185–190.
Ottati, V. C., & Isbell, L. M. (1996). Effects of mood Ruder, M., & Bless, H. (2003). Mood and the reli-
during exposure to target information on subse- ance on the ease of retrieval heuristic. Journal of
quently reported judgments: An on-line model of Personality and Social Psychology, 85, 20 –32.
misattribution and correction. Journal of Person- Ruderman, A. J. (1986). Dietary restraint—A theo-
ality and Social Psychology, 71, 39 –53. retical and empirical review. Psychological Bulle-
Pacini, R., Muir, F., & Epstein, S. (1998). Depressive tin, 99, 247–262.
realism from the perspective of cognitive- Sanbonmatsu, D. M., & Kardes, F. R. (1988). The
experiential self-theory. Journal of Personality effects of physiological arousal on information-
and Social Psychology, 74, 1056 –1068. processing and persuasion. Journal of Consumer
Patrick, C. J. (1994). Emotion and psychopathy— Research, 15, 379 –385.
Startling new insights. Psychophysiology, 31, Sanfey, A. G., Rilling, J. K., Aronson, J. A., Nystrom,
319 –330. L. E., & Cohen, J. D. (2003). The neural basis of
Peretz, I., Gagnon, L., & Bouchard, B. (1998). Music economic decision-making in the ultimatum game.
and emotion: Perceptual determinants, immediacy, Science, 300, 1755–1758.
Saunders, E. M. J. (1993). Stock-prices and Wall-
and isolation after brain damage. Cognition, 68, 111–
Street weather. American Economic Review, 83,
141.
1337–1345.
Pham, M. T. (1996). Cue representation and selection
Schmitt, W. A., Brinkley, C. A., & Newman, J. P.
effects of arousal on persuasion. Journal of Con-
(1999). Testing Damasio’s somatic marker hy-
sumer Research, 22, 373–387.
pothesis with psychopathic individuals: Risk tak-
Pham, M. T. (1998). Representativeness, relevance,
ers or risk averse? Journal of Abnormal Psychol-
and the use of feelings in decision making. Journal
ogy, 108, 538 –543.
of Consumer Research, 25, 144 –159. Schnall, S., Haidt, J., & Clore, G. L. (2006). Disgust
Pham, M. T. (2004). The logic of feeling. Journal of as embodied moral judgment. Unpublished manu-
Consumer Psychology, 14, 360 –369. script, University of Virginia.
Pham, M. T., Cohen, J. B., Pracejus, J. W., & Schwarz, J. C., & Pollack, P. R. (1977). Affect and
Hughes, G. D. (2001). Affect monitoring and the delay of gratification. Journal of Research in Per-
primacy of feelings in judgment. Journal of Con- sonality, 11, 147–164.
sumer Research, 28, 167–188. Schwarz, N. (1990). Feelings as information: Infor-
Pham, T. H., Philippot, P., & Rime, B. (2000). Sub- mational and motivational functions of affective
jective and autonomic responses to emotion induc- states. In R. M. Sorrentino & E. T. Higgins (Eds.),
tion in psychopaths. Encéphale, 26, 45–51. Handbook of motivation and cognition (Vol. 2, pp.
Pillutla, M. M., & Murnighan, J. K. (1996). Unfair- 521–561). New York: Guilford Press.
ness, anger, and spite: Emotional rejections of ul- Schwarz, N. (2002). Situated cognition and the wis-
timatum offers. Organizational Behavior and Hu- dom of feelings: Cognitive tuning. In L. F. Barrett
man Decision Processes, 68, 208 –224. & P. Salovey (Eds.), The wisdom in feelings: Psy-
Raghunathan, R., & Pham, M. T. (1999). All negative chological processes in emotional intelligence (pp.
moods are not equal: Motivational influences of 144 –166). New York: Guilford Press.
anxiety and sadness on decision making. Organi- Schwarz, N., & Clore, G. L. (1983). Mood, misattribu-
zational Behavior and Human Decision Pro- tion, and judgments of well-being—Informative and
cesses, 79, 56 –77. directive functions of affective states. Journal of Per-
Raghunathan, R., Pham, M. T., & Corfman, K. P. sonality and Social Psychology, 45, 513–523.
(2006). Informational properties of anxiety and Schwarz, N., & Clore, G. L. (1996). Feelings and
sadness, and displaced coping. Journal of Con- phenomenal experiences. In E. T. Higgins & A. W.
sumer Research, 32, 596 – 602. Kruglanski (Eds.), Social psychology: Handbook
Ratner, R. K., & Herbst, K. C. (2005). When good of basic principles (pp. 433– 465). New York: The
decisions have bad outcomes: The impact of affect Guilford Press.
178 PHAM

Seeman, G., & Schwarz, J. C. (1974). Affective state Tice, D. M., Bratslavsky, E., & Baumeister, R. F. (2001).
and preference for immediate versus delayed reward. Emotional distress regulation takes precedence over im-
Journal of Research in Personality, 7, 384 –394. pulse control: If you feel bad, do it! Journal of Person-
Sen, A. K. (1990). Rational behaviour. In J. Eatwell, ality and Social Psychology, 80, 53–67.
M. M. & N. P (Eds.), The New Palgrave: Utility Tiedens, L. Z., & Linton, S. (2001). Judgment under
and probability (pp. 198 –216). New York- emotional certainty and uncertainty: The effects of
London: Norton. specific emotions on information processing. Journal
Shiffman, S., & Waters, A. J. (2004). Negative affect of Personality and Social Psychology, 81, 973–988.
and smoking lapses: A prospective analysis. Journal Trafimow, D., Bromgard, I. K., Finlay, K. A., &
of Consulting and Clinical Psychology, 72, 192–201. Ketelaar, T. (2005). The role of affect in determin-
Shiv, B., & Fedorikhin, A. (1999). Heart and mind in ing the attributional weight of immoral behaviors.
conflict: The interplay of affect and cognition in Personality and Social Psychology Bulletin, 31,
consumer decision making. Journal of Consumer 935–948.
Research, 26, 278 –292. Tversky, A., & Griffin, D. (1991). Endowment and
Shiv, B., Loewenstein, G., Bechara, A., Damasio, H., contrast in judgments of well-being. In F. Strack,
& Damasio, A. R. (2005). Investment behavior and M. Argyle, & N. Schwarz (Eds.), Subjective well-
the negative side of emotion. Psychological Sci- being: An interdisciplinary perspective (pp. 101–
ence, 16, 435– 439. 118). Oxford: Pergamon Press.
Siemer, M., & Reisenzein, R. (1998). Effects of Vandijk, E., & Zeelenberg, M. (2006). The dampen-
mood on evaluative judgements: Influence of re- ing effect of uncertainty on positive and negative
duced processing capacity and mood salience. emotions. Journal of Behavioral Decision Mak-
Cognition & Emotion, 12, 783– 805. ing, 19, 171–176.
Silberman, E. K., Weingartner, H., & Post, R. M. Verplanken, B., Hofstee, G., & Janssen, H. J. W.
(1983). Thinking disorder in depression—Logic (1998). Accessibility of effective versus cognitive
and strategy in an abstract reasoning task. Archives components of attitudes. European Journal of So-
of General Psychiatry, 40, 775–780. cial Psychology, 28, 23–35.
Sinaceur, M., Heath, C., & Cole, S. (2005). Emo- Wheatley, T., & Haidt, J. (2005). Hypnotic disgust
tional and deliberative reactions to a public cri- makes moral judgments more severe. Psychologi-
sis—Mad Cow disease in France. Psychological cal Science, 16, 780 –784.
Science, 16, 247–254. White, G. L., Fishbein, S., & Rutsein, J. (1981). Passionate
Sinclair, R. C. (1988). Mood, categorization breadth, love and the misattribution of arousal. Journal of Per-
and performance-appraisal—The effects of order sonality and Social Psychology, 41, 56–62.
of information acquisition and affective state on Wilson, T. D., Lisle, D. J., Schooler, J. W., Hodges,
halo, accuracy, information-retrieval, and evalua- S. D., Klaaren, K. J., & LaFleur, S. J. (1993).
tions. Organizational Behavior and Human Deci- Introspecting about reasons can reduce post-choice
sion Processes, 42, 22– 46. satisfaction. Personality & Social Psychology Bul-
Sinclair, R. C., & Mark, M. M. (1995). The effects of letin, 19, 331–339.
mood state on judgmental accuracy—Processing Worth, L. T., & Mackie, D. M. (1987). Cognitive
strategy as a mechanism. Cognition & Emotion, 9, mediation of positive affect in persuasion. Social
417– 438. Cognition, 5, 76 –94.
Sinclair, R. C., Mark, M. M., & Clore, G. L. (1994). Wright, W. F., & Bower, G. H. (1992). Mood Effects
Mood-related persuasion depends on (mis)attribu- on subjective-probability assessment. Organiza-
tions. Social Cognition, 12, 309 –326. tional Behavior and Human Decision Pro-
Slovic, P., Finucane, M., Peters, E., & MacGregor, cesses, 52, 276 –291.
D. G. (2002). The affect heuristic. In T. Gilovich, D. Yeung, C. W. M., & Wyer, R. S. (2004). Affect,
Griffin & D. Kahneman (Eds.), Heuristics and bi- appraisal, and consumer judgment. Journal of
ases: The psychology of intuitive judgment (pp. 397– Consumer Research, 31, 412– 424.
420). New York: Cambridge University Press. Zajonc, R. B. (1980). Feeling and thinking—
Strack, F. (1992). The different routes to social judg- Preferences need no inferences. American Psy-
ments: Experiential versus informational strate- chologist, 35, 151–175.
gies. In L. L. M. A. Tesser (Ed.), The construction Zillmann, D. (1971). Excitation transfer in commu-
of social judgments (pp. 249 –275). Hillsdale, nication-mediated aggressive behavior. Journal of
N. J.: Lawrence Erlbawm Associates. Experimental Social Psychology, 7, 419 – 434.
Strongman, K. T. (1987). The psychology of emotion
(3rd ed.). Chichester: Wiley.
Sunstein, C. R. (2003). Terrorism and probability Received November 27, 2006
neglect. Journal of Risk and Uncertainty, 26, 121– Revision received November 27, 2006
136. Accepted November 28, 2006 䡲