Content uploaded by Gen Eickers
Author content
All content in this area was uploaded by Gen Eickers on Oct 18, 2021
Content may be subject to copyright.
opposite emotions such as happiness and sadness do not
unequivocally implicate distinct brain networks (Murphy,
Nimmo-Smith, & Lawrence, 2003). These results challenge
affect program theory’sfirst assumption that emotions can be
equated with central somatic responses, and instead point
toward the need for more open-ended categories that support
this sort of variability.
Second, Moors presents a series of methodological issues that
compromise studies supporting classical theories. Against affect
program theory, these include biases in both recognition and pro-
duction studies, and difficulties in determining the relation between
aspecificresponse(e.g.,afacialexpression) and a basic emotion.
She considers three strategies that might rescue affect program the-
ory but does not reject them explicitly. These strategies are (a)
insisting that empirical evidence does not unequivocally falsify
their theory, but rather calls for more sophisticated methods and
designs; (b) claiming that basic emotions might have more fine-
grained subsets with slightly different profiles; and (c) attributing
these failures to the use of static stimuli in a single modality rather
than multimodal, dynamic stimuli.
Even if Moors does not deliver a decisive verdict, we agree that
the objections that she reviews are important criticisms against
affect program theory. First, we grant that disunity has been well
established by empirical findings and underestimated by affect pro-
gram theory, though we argue next that skeptics have also underes-
timated unity; an adequate view must take both unity and disunity
into consideration. Second, even if there are more fine-grained sub-
sets for each basic emotion, and even if we use multimodal and
dynamic stimuli, affect program theory still posits categories that
are too rigid. For instance, Ekman (1992,Ekman&Cordaro,2011)
claimed that emotions are grouped in what he called “emotion
families”but that these families are still presumed to have a distinct
physiology, distinctive universal signals, and the like. We believe
that emotions should not be classified according to fixed signals
but rather to more flexible forms of embodiment—aviewthatwe
present next.
The third strand of criticism considered by Moors involved the
logic behind the design and interpretation of empirical research.
In Moors’sview,todeterminespecificity for each basic emotion,
we need to establish a relation between a basic emotion and the
hypothesized response. However, manipulating basic emotions
turns out to show only the relation between one or two compo-
nents (e.g., motor and cognitive components) or between one
component and an emotion label. Moors claims that, if affect pro-
gram theory is true, there should be concordance among multiple
components and that concordance should be higher for a given
emotion than it is for others. We think the evidence for multicom-
ponent concordance is somewhat greater than Moors implies, but
we also grant that physiological differentiation of the emotions is
imperfect. This puts pressure on the affect program theory as tra-
ditionally understood.
Fourth and last, she sides with Ortony and Turner (1990)
and states that traditional theories run the risk of affirming the
consequent. For instance, proponents of affect program theory
infer that their account is supported by evidence for universal
facial expressions. But even if facial expressions were universal,
they could be caused by different factors such as learning his-
tory. Thus, arguing that universality implies the truth of affect
program theory is a fallacy. Against this charge, we believe that
a more charitable interpretation is plausible. Instead of proving
affect program theory, universality could be seen as confirming
one of its hypotheses. Efforts to accommodate such evidence by
other theories may even be post hoc. Nonetheless, Moors also
considers a stronger argument against affect program theory,
one might reject the evidence for universality and argue that
facial expressions might still be culturally bound. This is an
objection that we partly agree with. Although we do not believe
that expressions are as open-ended as dimensional theories
hold, we do think that they are culturally influenced and thus
not entirely natural as affect program theory would have it.
From these arguments, Moors concludes that vernacular
discrete categories fail to be scientifically meaningful. Instead,
she submits that only components are real. In the next section,
we present objections against this conclusion and show why
these arguments point toward revision rather than elimination
of discrete categories.
Upshot: Revision Required, Not Elimination
Moors thinks that disunity leads us to the rejection of the scien-
tific status of discrete emotion categories, that is, that it points
toward eliminativism. Moors’s argument presupposes Carnap’s
view on the adequacy of scientific definitions. According to this
account, a scientific (prescriptive) definition is adequate in case
it robustly overlaps with the initial descriptive definition of the
phenomenon in question (similarity criterion), and it allows us
to generalize from one member to other members of the set
(fruitfulness criterion). Moors takes the latter to imply that the
set should be homogeneous in a nonsuperficial way, meaning
that members of the set should share either a mechanism or
structure.
This is precisely what disunity would establish. If disunity is
true, emotions—as captured by discrete categories—do not form
homogeneous sets. Thus, Moors thinks, these categories are not sci-
entifically adequate. Furthermore, she thinks that the components
proposed by dimensional appraisal theories and her version of PC
(which encompasses goal-directedness and emotional metaexper-
ience) also fail to make emotions—or rather, emotional episodes—
scientifically meaningful. This is because these components are
present in the production of both emotional and nonemotional
content, thus precluding us from generalizing from one emotion to
the others based on these features.
We think, however, that there are good reasons to resist elimi-
nativism. First, Moors’s argument for eliminativism rests on a
false dilemma. The argument assumes that either discrete catego-
ries map one-to-one onto physiological states or they are not sci-
entifically fruitful. There is, however, a third option, namely, that
discrete categories map one-to-many onto physiology while still
having enough unity to be useful for science. Many of the entities
that figure in scientific generalizations have variable instantia-
tion: genes, proteins, metals, commodities, economic demand,
migrations, kinship ties, stratification, and nation-states. We
don’t mean to imply that all these kinds are natural in the way
that “water”is putatively natural (having a common essence). In
fact, we think that all sciences, even physics and chemistry,
depend on theoretical constructions; what counts as a planet or
an element is stipulated. Scientists chose taxonomies that prove
to be useful in supporting generalizations. Likewise for emotions.
32 COMMENTARIES
Suppose we define fear as an embodied motivation state that
arises in response to danger. Such a functional definition
(embodied danger response) will be instantiated in different
ways (not just in the amygdala), but it might still prove scientifi-
cally useful, as it may allow generalizations, thus satisfying the
fruitfulness criterion embraced by Moors. For example, horror
film directors know how to scare people, even though fear is
manifested differently (and not in everyone); fear also impacts
perception and attention in numerous studies, and sustained or
extreme fear can cause trauma.
Second, we believe that there may be at least two interpretations
of disunity. One interpretation is completely consistent with the
thesis that emotions are natural kinds. Some natural kinds are said
to be homeostatic property clusters—collections of features that
tend to co-occur because there are underlying mechanisms that
make them mutually reinforcing (Boyd, 1991,1999). Homeostatic
property clusters can vary from case to case, exhibiting different
features in the cluster. On this interpretation, physiological vari-
able is compatible with the claim that emotions are natural kinds
(Scarantino, 2012a,2012b). On a second interpretation, disunity
might be used to establish that emotions do not form natural kinds
by challenging the assumption that each is governed by a single
underlying mechanism. But even this kind of disunity need not
entail that emotions are not real. The fact that a phenomenon
lacks mechanistic unity is not always an argument against its real-
ity. Consider psychiatric categories, such as depression. Depression
is defined as a constellation of different symptoms that may load
differently in different individuals and cross-culturally but is none-
theless painfully real and implicates various mechanisms at the
neural and bodily levels. Other examples abound. There are many
categories that support scientificgeneralizations,eventhoughthey
lack underlying mechanisms that promote unity: most social cate-
gories (gender, class, political orientation), artifact concepts (ham-
mer, boat), action types (running, swimming, dancing), many
biological categories (predator, species, mating display), and super-
ordinate-level categories (plants, liquids, and lipids). Some of these
are social constructs (e.g., social categories and artifacts), but they
are perfectly real. Thus, emotions can still be real even if they are
not natural kinds.
Consequently, instead of elimination we think we need to
revise discrete emotion categories and define them in ways that
allow them to be variably instantiated. Proposals for such revi-
sion are already available (Scarantino, 2012a,2015). In the fol-
lowing sections, we propose an embodied view that embraces
revisionism and defines emotions both as socially constructed
and embodied phenomena. If we are correct, we need not elimi-
nate discrete emotion categories or give up on the idea that
emotions are in fact real, albeit variable and culturally
influenced.
A Social Embodiment Approach to Discrete Emotions
Emotions and Social Affordances
As just noted, we embrace a view that defends emotions as
being both embodied and socially constructed. By arguing that
emotions are embodied states, we mean to say that emotions
necessarily come along with physical changes and are displayed
in physiological patterns. These patterns are not arbitrary.
They prepare the body for action. The actions in question can
be socially shaped. Consider anger, which can be defined as an
embodied defense response. Cultures may differ in what counts
as an offense and to act when offended. The same emotion can
be realized in different physical patterns. For example, people
respond differently to offenses in different contexts. These pat-
terns are recognizable and hence contribute to social affordan-
ces. Someone who displays anger by shouting, for instance,
indicates to others that the social context affords aggression,
and that, in turn, changes the social affordances for others in
the context. The impact of culture on emotion is essential to
emotions in the sense that culture and sociality inevitably play
a role in shaping emotional responses and responses to those
responses (i.e., what gets cued by an emotional display). Social
forces make emotions recognizable, and once recognized, the
emotion can exert a social influence. Thus, the social context in
which an emotion is displayed gets shaped by that emotion and
conversely.
Our claim that emotions are embodied states and thus play a
role in registering and generating social affordances does not
entail that bodily signals are universal. We disagree with the
argument that emotions are directly presented to the world, for
example, via facial expressions, and thereby immediately (and
universally) understandable (Gallagher, 2008). Rather the view
is that body changes prepare the organism for (culturally
tuned) actions and that these changes can be interpreted
(through cultural learning), and potentiate (culturally sanc-
tioned) responses. On this account, there is no one-to-one
mapping between emotions and faces, brain mechanisms, or
physiology in general, but the body plays a key role in emotion:
Each emotion involves action preparation and the resulting
projection of physical signals, which create social affordances.
Throughout this registration process, which is essentially a pro-
cess of taking in social cues that may help to choose, coordi-
nate, or even to motivate the upcoming action, some action
possibilities become more salient and seem more urgent. From
an observer perspective this changes affordances. If an observer
sees facial expressions or other physical signals, their choices
change, too. A flash of rage may cause either an aggressive
response or an apology.
Consider a case for ambiguous and context-dependent emo-
tional display: blushing. It appears to be quite unpredictable
when someone blushes. First, there are people who blush more
often than others. Some people with social anxiety tend to
blush a lot. Second, the blush can occur in response to a nega-
tive or a positive event. Say a friend tells you he will be getting
married soon and either you or he or both blush. Each of these
situations is imaginable even if this is clearly not an embarrass-
ing situation. Blushing does not have only one emotional corre-
late, but many. Pride, shame, and embarrassment are the main
examples, but there is even more diversity in reports about
what is felt during blushing (Crozier, 2006). Sometimes people
even report they feel happy when blushing or simply blush
because they got surprised. In some situations, subjects are not
even able to tell exactly what they feel. We draw here on trans-
actional theories (Griffiths & Scarantino, 2009; Scarantino,
2012b,2015), according to which emotions are a kind of mean-
ingful engagement with the world, highly influenced by the
social environment and in dynamic relation to it.
COMMENTARIES 33
All emotion terms correspond to many possible bodily pat-
terns but, pace Ekman (1992), none of these are unique to any
given emotion. To fulfill their social roles, responses to insults,
dangers, and losses must exhibit considerable plasticity. At the
same time, we do not mean to state that ambiguity of emotion-
expression mapping is open-ended (Russell, 1994). Moors
refers to Russell, who tends to focus exclusively on undifferenti-
ated arousal. When confronted with evidence that the body
does more than getting excited, he erroneously responds that
the expressive behaviors are open-endedly ambiguous. To
establish this, he shows that context affects the interpretation of
expressions. But this actually shows that expressions can have
fixed meanings in a context. The fact that the same expression
can have different meanings does not entail that, on a given
occasion, it is not the expression of a specific emotion. Plus,
ambiguity is not open-ended itself. To be able to register emo-
tions and act upon emotional expressions and social affordan-
ces, they must signal specific action tendencies. Expressions are
not arbitrary signs but indicators of how someone may behave.
Within a culture, each emotion may correspond to a limited set
of possible physical or physiological patterns. Ambiguity can
also derive from the fact that some emotions overlap physiolog-
ically, but there are usually differences as well (e.g., sadness and
fear may both be displayed by undistinguishable facial expres-
sions but differ in other physiological responses, such as heart
rate and perspiration). The category of emotions itself may be
open-ended (new emotions can be created and blends are com-
mon), but specific categories have predictable impact, once cul-
tural background, social context, and (as we see) individual
goals are specified.
Embodiment
In describing the social functions of emotions, we have been
referring to their efficacy as signals. This implies that emotions
are expressed in the face and body. With the standard affect
program theory, we think there is strong evidence suggesting
that each emotion is somatically realized. The two traditions
upon which Moors draws—dimensional appraisal theory and
PC—have tended to downplay the body. In its early formula-
tions, dimensional appraisal theory ignored the body, and early
formulations of PC allowed only one dimension of bodily
responsiveness: arousal. We think that there is much more
somatic specificity than these views tend to recognize, though
we stop short of the strong thesis that each emotion maps onto
a unique set of bodily symptoms.
Evidence linking emotions to the body comes in a variety of
forms. Most famously, there is work on distinctive facial expres-
sions, and new expressions are identified regularly. For example,
instead of a single positive expression, joy, there is work on awe
and amusement (Shiota et al., 2003), love and peace (Hejmadi et
al., 2000), pride (Tracy & Robbins, 2004), interest (Reeve, 1993),
desire (Gonzaga et al., 2006), and sensory pleasure (Wehrle et al.,
2000). There is also a literature showing that feedback from facial
expressions (Flack, 2006), posture (Duclos et al., 1989), and respira-
tion patterns (Philippot, Chapelle, & Blairy, 2002)canevokerecog-
nizable and distinct emotional states. In addition, there are dozens
of studies linking emotions to specificautonomicpatterns(e.g.,
Friedman, 2010;Kreibig,2010;Levenson,2014). Pattern classifiers
do a reasonably good job inferring a person’semotionalstatefrom
physiological measures, even in cases where the emotions in ques-
tion were elicited in different ways (Kragel & LeBar, 2013). Emo-
tions are associated with a variety of distinct neuroendocrine
responses as well (Henry, 1986). Linking these dimensions, there
are studies showing that facial expressions co-occur with other
somatic variables (Lerner, Dahl, Hariri, & Taylor, 2007;Price&
Harmon-Jones, 2015).
Moors is less impressed by such findings, and she cites meta-
analyses that aim to challenge the specificity of somatic
responses. We find some of these meta-analyses unconvincing.
Although aiming at critique, they explicitly acknowledge that a
surprising degree of physiological differentiation is possible
(Cacioppo, 2000; Quigley & Barrett, 2014). These meta-analy-
ses do establish that the same emotion category can be associ-
ated with different bodily responses in different contexts, and
this carries some weight against traditional affect program the-
ory. We strongly agree that pioneers of the approach, such as
Ekman, have exaggerated universality and underestimated vari-
ability in embodiment, but many proponents of discrete emo-
tions explicitly recognize context effects (e.g., Lang, 2014;
Roseman, 2011). Even those who work with animal models rec-
ognize that context matters, for example, those who work on
the flight–fight–freezing response (Fanselow, 1994). In fact,
there are four freezing mechanisms (Branda~
o et al., 2008).
Some proponents of discrete emotions also recognize that
embodiment can be shaped by culture, rejecting the universal-
ity of facial expressions (e.g., Prinz, 2004).
In the next section, we describe how the degree of specificity
shown in the literature, when combined with the social functions
of emotions, can allow for discrete emotions while rejecting the
untenable one-to-one mapping between emotions and bodily
states.
Discrete Emotions
Moors’s skepticism about discrete emotions derives, in part,
from her conviction that there are no biologically fixed somatic
patterns that can be neatly mapped onto emotion categories.
We agree with her conviction that there are no fixed somatic
patterns, but it does not follow from this argument that there
are no discrete emotions. By analogy, consider dancing. Danc-
ing is a culturally shaped, embodied activity with open-ended
variation. Dances of the same category can be instantiated in
numerous ways, and dances in different categories can overlap
so that decontextualized segments would be impossible to clas-
sify. Nevertheless, there is no reason to doubt that there are dis-
tinct styles of dance, which can be reliably recognized. Emotion,
of course, differs from dance, but the example establishes that
variable embodiment is compatible with discrete categories.
We think that emotions can be classified according to their
motivational role taken in conjunction with their embodiment.
Emotions are embodied responses that respond to situations
that bear on the well-being of an organism, such as threats,
losses, achievements, transgressions, and so on. Under cultural
and evolutionary forces, we have acquired a variety of behav-
ioral responses to each of these situations. In principle, we
could respond with any behavior to a threat or a loss, but some
responses become entrenched, and these are both potentiated
34 COMMENTARIES
and recognizable. When a person in a Western culture is
insulted, standing responses include various forms of symbolic
aggression (shouting, profanity, and bodily responses that pre-
pare the body for a violent altercation), and various forms of
withdrawal (storming away, sulking, and brooding). We call all
of these “anger.”They are functionally unified by their situa-
tional meaning. “Anger”is a superordinate category, and we
could label subtypes, each of which would have a range of vari-
ation, pivoting around prototypical features. We describe
someone as enraged or sullen. These embodiments can be rec-
ognized by others, especially when contextual information is
available. A sullen state may be mistaken for another form of
disengagement when considered physiologically, but when it
arises in response to an insult, it takes on a situational meeting
that can be discerned by others, especially within a culture.
The evidence for embodiment together with the evidence
that emotions have socially recognized motivational impact
suggest that emotional responsiveness is relatively constrained
within a culture. There are bodily responses that we can classify
with considerable reliability even when no context is given, and
these become even easier to interpret when they unfold in sit-
uations that have discernable import for an organism. Oppo-
nents of discrete emotions are impressed by the ambiguity of
expressive behavior, but their own research shows that contex-
tually available information can be used to arrive at consistent
interpretations (e.g., Carroll & Russell, 1996). Skeptics also
complain that somatic responses vary with induction methods,
but this makes perfect sense if we regard bodily responses as
adapted to specific contexts. If the skeptical position were true,
recognition of emotions by physiological variable should be
impossible both across induction methods and within.
Moors presents water as the gold standard (so to speak) of a
real kind. All samples of water are said to share a common
chemical essence (though even this example is controversial;
Chang, 2012). Emotions are not like that. Anger has many
physiological realizers, and emotions, as a category, engage
brain mechanisms that contribute to nonemotional behavior
(2.2). Emotions are functionally unified by the fact that each is
a bioculturally shaped, embodied response that prepares us for
dealing with situations that bear on well-being. Emotions are
more like tools than chemical compounds. Anger, like a ham-
mer, can take various forms, but those are constrained by cul-
tural conventions and suitability for its functional role.
Emotions need not be natural, much less have unifying essen-
ces, to be real. Even if emotions are social constructions in
some sense (e.g., if they are bodily habits acquired through
socialization), they may be perfectly real. There is a clear differ-
ence between constructed entities and nonexistent entities.
Moors’s Positive View
Dimensional Appraisals
We want to turn now to Moors’sownpositiveview.Weagree
with many of her conclusions, but we also want to indicate
some concerns. One of the elements that Moors intends to inte-
grate into her view is dimensionalappraisaltheory.Initstradi-
tional form, this theory claims that emotions are elicited through
asequenceofappraisalsandthateachemotioncorrespondstoa
constellation of different appraisal judgments rather than a single
molar judgment, such as “Iamindanger.”In addition, some
dimensional appraisal theories include physiological, behavioral,
and other noncognitive elements. Within a constructivist frame-
work, the use of dimensions is motivated by the idea that ver-
nacular emotion categories do not capture real natural processes,
meaning that emotions are to be described not as discrete entities
but as combinations of the values these dimensions can take on.
These combinations can be open-ended in their variation, thus
allowing infinite possible emotions.
The dimensions that Moors considers include appraisal judg-
ments (such as goal congruence or incongruence), action tenden-
cies, somatic responses, and overt behavior. The latter three are
classified as behavior related, and they all contribute to the gen-
eral experiential component ofemotions(Moors,thisissue,
p. 8). Although we agree with Moors about the behavioral
aspects of emotions, we take issue with the idea that the cogni-
tive appraisal dimensions are a proper approach to emotions.
First, evidence for dimensional theories is inconclusive. Several
reviews have discussed at length the tension between discrete and
dimensional approaches, concludingthatevidencedoesnotgreatly
favor one or the other approach (Hamann, 2012;Kragel&LaBar,
2016;Lench,Flores,&Bench,2011). The cognitive appraisals have
not been located in the brain, and there is no evidence that they are
subject to selective deficits.
Second, there is also an important body of evidence support-
ing discrete categories. Lench et al. (2011) showed that moder-
ate effect sizes differentiate various discrete emotions such as
happiness, sadness, and anger. Similarly, Kragel and LaBar
(2016), in their review on neuroimaging data, concluded that
emotions are best represented using discrete categories rather
than dimensional ones. This means that despite there being dis-
unity, there could still be specificity at the brain level that
should not be underestimated. In other words, things might
not be as disorganized as dimensional theories claim.
Third, dimensional appraisals imply an implausible degree
of cognitive sophistication for emotion elicitation. In many
cases emotions are perceptually triggered (a loud, sudden noise;
a beautiful scene; drinking spoiled milk; losing one’s balance;
listening to music). Here, it seems unlikely that we need to step
through cognitive appraisals, though cognition may play a role
in regulation and elaboration.
Last, we disagree with the thesis that postulating discrete emo-
tions lacks scientificutility.Thesecategorieshaveledtoimmensely
fruitful research programs, including countless studies in which
discrete emotion categories are shown to have differential impact
on decision making and behavior. As we argued earlier, we have
good reasons to resist the sort of eliminativism of discrete catego-
ries dimensional theories propose. Instead, we think we should
maintain discrete emotion categories but revise them to account
for disunity. Thus, the rejection of the scientificusefulnessofdis-
crete categories, which provides the argumentative basis for dimen-
sional theories, is indefensible. If our view is correct, then
dimensional views cannot even get off the ground.
Valence
The next ingredient that Moors is willing to countenance is
valence. This is an import from both dimensional appraisal
COMMENTARIES 35
theories and PC theories. Indeed, most emotion theorists try to
accommodate the intuition that emotions are valenced (for an
exception, see Solomon, 2001). We are inclined to agree with
Moors about the importance and reality of valence (see Prinz,
2010). Still, we want to register two notes of caution.
First, we take issue with those models that try to explain the
behavioral differences between negative and positive emotions
by “valence alone”(see Russell, 2003). Different negative emo-
tions influence behavior differently. This is especially clear
when one looks at emotional impacts on judgment. Here are
some examples: DeSteno et al. (2000) showed that sadness and
anger have different effects on perceived likelihood, Lerner and
Keltner (2001) showed that fear and anger have different
impact on risk assessment, Zeelenberg and Pieters (2004)
showed that regret and disappointment have different effects
on customer satisfaction, and Seidel and Prinz (2013) showed
that anger and disgust have different effects on moral judg-
ment. In line with her skepticism regarding discrete emotion
categories, Moors would reply that she can rely on appraisal
dimensions to explain such differential effects. We would favor
a more embodied approach, but we grant that cognitive pro-
cesses can contribute here. Our main point in the present con-
text is that valence has limited predictive power, but the
examples also underscore our more general conclusion that
there is scientific utility in discrete emotion categories. The fact
that, say, anger versus disgust induction can result in predict-
able different behavior testifies the reality of these constructs.
They have different roles in information processing. Given the
vast literature showing emotion-specific effects, eliminativism
seems unlikely.
Another concern is the assumption that valence can be
modeled by a single dimension. A recent review of brain imag-
ing results suggests that valence is implemented by a variety of
different mechanisms (Lindquist et al., 2016). Research looking
at neural circuitry and behavior indicates that there are a vari-
ety of functionally different reward systems and inhibition sys-
tems (e.g., Corr, 2013). There is also behavioral evidence that
emotions with the same valence can have opposite behavioral
effects. For example, disgust is a withdrawal emotion, and anger
is an approach emotion (Carver & Harmon-Jones, 2009).
Such differentiation raises the possibility that valence may
fractionate into multiple categories. Since the days of animal
learning, researchers have drawn a distinction between removal
of punishment, nonpunishment, and reward, which already
casts doubt on the idea that valence is a single dimension. Push-
ing this further, it might turn out that there are many forms of
valence. We can see this by reflecting on the most commonly
studied negative emotions. Anger causes approach, disgust
causes withdrawal, sadness causes inactivity, and fear can cause
all three of these responses (fight–flight–freezing). Admittedly,
all negative emotions share one thing: We exert effort to avoid
experiencing them. But this does not begin to capture the ways
in which negativity can be manifest.
This may appear to play into the hands of the skeptic. Per-
haps even valence should be eliminated. We see the upshot
somewhat differently. Efforts to replace discrete emotions with
simpler dimensions, such as valence, may lead us to overlook
distinctions in positive and negative responsiveness. By reflect-
ing on the specific functions of discrete emotions (e.g.,
aggressing against threats, avoiding contaminants, regrouping
after a loss, etc.), we can discover the need for a richer
approach.
Goals
In a third aspect of her theory, Moors wants to integrate goals. She
argues that goal-directed mechanisms are relevant for the ability to
readjust and regulate action tendencies. Moors’sargumentagainst
affect program theory assumes that such accounts must reject the
goal-directedness of emotions. Her exploration implies that
embodied approaches cannot incorporate goals. Moors’sown
attempt at a dynamic definition of emotions, as described in the
previous sections, speaks in favor of goal-directedness. That is,
goals normally contribute to emotion elicitation. We agree with
Moors in embracing the goal-directedness, the strategic element of
emotions. We now want to establish that emotions, in our embod-
ied constructionist view, need not be stimulus driven. Our view
combines elements of Moors’sviewwiththeembodiedapproach
described previously. Moors takes affect program theory as the
main embodied approach and rejects it as nativist and context
insensitive. We present an alternative by rejecting the aspect of con-
text insensitivity and by arguing that emotions are strategic.
To make this case, we argue that goal-sensitivity is consis-
tent with embodiment. First, embodied views commit to physi-
cal realization but are nevertheless compatible with many
theories of elicitation. Second, against Moors, embodiment
means that emotions can function as signals, and signaling is
most effective when strategic. The function of emotions (as
preparation for action) lends itself to signaling. This works
only if the emotional display implies an action tendency. The
translation of the emotional state into emotional display
depends on many factors. In some cases, it may hurt us in the
long run to display aggression when we are angry so that it is
strategically better not to display anger at all or to express it dif-
ferently. That is, (social) appropriateness of certain behaviors/
expressions plays a role in choosing emotional display.
There are other approaches that emphasize the strategic func-
tions of emotions (Frank, 2004;Frijda&Mesquita,1994;Keltner&
Haidt, 1999,2001;Ross&Dumouchel,2004). These approaches
tend to focus on social functions (such as cooperation), and they
also tend to emphasize evolved functions. This orientation implies
that emotions evolved because it isadvantageousforsocietyor
social groups. On our embodied view, we want to point out that
emotions can also have strategic advantages for an individual, and
these depend on personal goals in a specificcontext.Forexample,
we may choose to express anger when it is to our own advantage to
display aggression and express disappointment when it is better to
pout than to rage. Such emotions are often culturally tuned, and
their meaning depends on social interpretation. The strategy
deployed by an individual may reflect transient personal goals, but
it is also informed by cultural norms. Thus, there are societies in
which strong displays of anger are shunned (Briggs, 1970), and
there are cultural differences in how people respond to anger dis-
plays (Engelmann & Pogosyan, 2013).
The account we defend in this commentary considers emo-
tions to be more strategic than often assumed. As Ben-Ze’ev
(2000) points out, it is essential to consider the practical nature
of emotions when defining them, as they are not theoretical
36 COMMENTARIES
states. That practical nature includes motivation, as discussed
earlier, and signaling, because motivational preparation can be
perceived. Like other action-oriented states, emotions are not
mere reflexes; they are sensitive to goals. The practical nature
of emotions requires both goals and embodiment.
This last remark brings out both our agreement with Moors and
the main axis on which our views divide. We applaud her effort to
emphasize variability; context sensitivity; and, above all, goal direct-
edness. The traditional affect program theory fails to adequately
address these aspects of emotion. Moors’sefforttoaccommodate
these aspects includes both a negative thesis (skepticism about dis-
crete emotion categories) and a positive view about the building
blocks used to construct emotions (dimensional appraisals, valence,
and arousal). We reject the skeptical view (discrete emotions are
scientifically useful even if variable and constructed), and we think
bodily responses offer more promising building blocks than those
that Moors prefers. Where she takes constructivism to entail antire-
alism and (a qualified form of) cognitivism, we favor an approach
that preserves the embodied orientation of the affect program the-
ory while embracing the evidence that bodily changes take on pat-
terns and meanings that reflect cultural, social, and individual
situations. Some proponents of embodied approaches have made
themselves vulnerable to refutation by supposing that bodily
responses are fixed. In this commentary, we have tried to suggest
that embodiment and construction are compatible and that both
forestall skepticism about discrete emotions.
References
Ben-Ze’ev, A. (2000). The subtlety of emotions. Cambridge, MA: MIT
Press.
Boyd, R. (1991). Realism, anti-foundationalism and the enthusiasm for
natural kinds. Philosophical Studies,61, 127–148.
Boyd, R. (1999). Homeostasis, species, and higher taxa. In R. A. Wilson
(Ed.), Species: New interdisciplinary essays (pp. 141–185). Cambridge,
MA: MIT Press.
Branda~
o, M. L., Zanoveli, J. M., Ruiz-Martinez, R. C., Oliveira, L. C., &
Landeira-Fernandez, J. (2008). Different patterns of freezing behavior
organized in the periaqueductal gray of rats: Association with different
types of anxiety. Behavioral Brain Research,188,1–13.
Briggs, J. L. (1970). Never in anger; portrait of an Eskimo family. Cam-
bridge, MA: Harvard University Press.
Cacioppo, J. T., Berntson, G. G., Larsen, J. T., Poehlmann, K. M., & Ito, T.
A. (2000). The psychophysiology of emotion. In Handbook of emotions
(2nd ed., pp. 173–191). New York: Guilford Press.
Carroll, J. M., & Russell, J. A. (1996). Do facial expressions signal specific
emotions? Judging emotion from the face in context. Journal of Person-
ality and Social Psychology,70, 205–218.
Carver, C. S., & Harmon-Jones, E. (2009). Anger is an approach-related
affect: Evidence and implications. Psychological Bulletin,135, 183–204.
Chang, H. (2012). Is water H2O? Evidence, pluralism and realism. Dor-
drecht, the Netherlands: Springer.
Corr, P. J. (2013). Approach and avoidance behavior: Multiple systems and
their interactions. Emotion Review,5, 285–290.
Crozier, W. R. (2006). Blushing and the social emotions: The self unmasked.New
York: Palgrave Macmillan.
DeSteno, D., Petty, R. E., Wegener, D., & Rucker, D. T. (2000). Beyond
valence in the perception of likelihood: The role of emotion specificity.
Journal of Personality and Social Psychology,78, 397–416.
Duclos, S. E., Laird, J. D., Schneider, E., Sexter, M., Stern, L., & Van
Lighten, O. (1989). Emotion-specific effects of facial expressions and
postures on emotional experience. Journal of Personality and Social
Psychology,57, 100–108.
Ekman, P. (1992). An argument for basic emotions. Cognition & Emotion,
6(3), 169–200.
Ekman, P., & Cordaro, D. (2011). What is meant by calling emotions basic.
Emotion Review,3(4), 364–370.
Engelmann, J. B., & Pogosyan, M. (2013). Emotion perception across cul-
tures: The role of cognitive mechanisms. Frontiers in Psychology,4, 118.
Fanselow, M. S. (1994). Neural organization of the defensive behavior sys-
tem responsible for fear. Psychonomic Bulletin & Review,1, 429–438
Flack, W. (2006). Peripheral feedback effects of facial expressions, bodily
postures, and vocal expressions on emotional feelings. Cognition &
Emotion,20, 177–195.
Frank, R. (2004). Introducing moral emotions into models of rational
choice. In A. S. R. Manstead, N. Frijda, & A. Fischer (Eds.), Feelings
and emotions: The Amsterdam symposium (pp. 422–440). Cambridge,
UK: Cambridge University Press.
Friedman, B. H. (2010). Feelings and the body: The Jamesian perspective on
autonomic specificity of emotion. Biological Psychology,84,383–393.
Frijda, N. H., & Mesquita, B. (1994). The social roles and functions of emo-
tions. In S. Kitayama & H. R. Markus (Eds.), Emotion and culture:
Empirical studies of mutual influence. Washington, DC, American Psy-
chological Association.
Gallagher, S. (2008). Direct perception in the intersubjective context. Con-
sciousness and Cognition,17, 535–543.
Gonzaga, G. C., Turner, R. A., Keltner, D., Campos, B., & Altemus, M.
(2006). Romantic love and sexual desire in close relationships. Emotion,
6, 163–179.
Griffiths, P. E., Scarantino, A. (2009). Emotions in the wild: The situated
perspective on emotion. In P. Robbins & Murat Ayede (Eds.), Hand-
book of situated cognition (pp. 437–454). Cambridge, UK: Cambridge
University Press.
Hamann, S. (2012). Mapping discrete and dimensional emotions onto the
brain: Controversies and consensus. Trends in Cognitive Sciences,16,
458.
Hejmadi, A., Davidson, R. J., & Rozin, P. (2000). Exploring Hindu Indian
emotion expressions: Evidence for accurate recognition by Americans
and Indians. Psychological Science,11, 183–187.
Henry, J. P. (1986). Neuroendocrine patterns of emotional response. In R.
Plutchik & H. Kellerman (Eds.), Emotion, theory, research, and experi-
ence,Vol. 3 (pp. 37–60). Orlando, FL: Academic Press.
Johansen, J. P., Cain, C. K., Ostroff, L. E., & LeDoux, J. E. (2011). Molecular
mechanisms of fear learning and memory. Cell,147(3), 509–524.
Keltner, D., & Haidt, J. (1999). Social functions of emotions at four levels of
analysis. Cognition & Emotion,13, 505–521.
Keltner, D., & Haidt, J. (2001). Social functions of emotions. In T. Mayne
& G. A. Bonanno (Eds.), Emotions: Current issues and future directions
(pp. 192–213). New York, NY: Guilford Press.
Kragel, P. A., & LaBar, K. S. (2016). Decoding the nature of emotion in the
brain. Trends in Cognitive Sciences,20, 444–455.
Kragel, P. A., LaBar, K. S. (2013). Multivariate pattern classification reveals
autonomic and experiential representations of discrete emotions. Emo-
tion,13, 681–90.
Kreibig, S. D. (2010). Autonomic nervous system activity in emotion: A
review. Biological Psychology,84, 394–421.
Lang, P. J. (2014). Emotion’s response patterns: The brain and the auto-
nomic nervous system. Emotion Review,6, 93–99.
Lench, H. C., Flores, S. A., & Bench, S. W. (2011). Discrete emotions pre-
dict changes in cognition, judgment, experience, behavior, and physiol-
ogy: A meta-analysis of experimental emotion elicitations.
Psychological bulletin,137, 834–855.
Lerner, J., & Keltner, D. (2001). Fear, anger, and risk. Journal of Personality
and Social Psychology,81, 146–159.
Lerner, J., Dahl, R., Hariri, A., & Taylor, S. (2007). Facial expressions of
emotion reveal neuroendocrine and cardiovascular stress responses.
Biological Psychiatry,61, 253–260.
Levenson, R. W. (2014). The autonomic nervous system and emotion.
Emotion Review,6, 100–112.
Lindquist, K. A., Satpute, A. B., Wager, T. D., Weber, J., & Barrett, L. F.
(2016). The brain basis of positive and negative affect: Evidence from a
meta-analysis of the human neuroimaging literature. Cerebral Cortex,
26, 1910–1922.
COMMENTARIES 37
Lindquist, K. A., Wager, T. D., Kober, H., Bliss-Moreau, E., & Barrett, L. F.
(2012). The brain basis of emotion: A meta-analytic review. Behavioral and
Brain Sciences,35(3), 121–143.
Moors, A. (2012). Comparison of affect program theories, appraisal theo-
ries, and psychological construction theories. In P. Zachar & R. D. Ellis
(Eds.), Categorical versus dimensional models of affect: A seminar on
the theories of Panksepp and Russell (pp. 257–278). Amsterdam, the
Netherlands: John Benjamins.
Murphy, F. C., Nimmo-Smith, I., & Lawrence, A. D. (2003). Functional
neuroanatomy of emotions: A meta-analysis. Cognitive, Affective, &
Behavioral Neuroscience,3, 207–233.
€
Ohman, A., & Mineka, S. (2001). Fears, phobias, and preparedness:
Toward an evolved module of fear and fear learning. Psychological
Review,108, 483–522.
Ortony, A., & Turner, T. J. (1990). What’s basic about basic emotions? Psy-
chological Review,97, 315–331.
Philippot, P., Chapelle, G., & Blairy, S. (2002). Respiratory feedback in the
generation of emotion. Cognition & Emotion,16, 605–627.
Price, T. F., & Harmon-Jones, E., 2015. Embodied emotion: The influence
of manipulated facial and bodily states on emotive responses. WIREs
Cognitive Science,6, 461–473.
Prinz, J. (2004). Gut reactions: A perceptual theory of emotion.NewYork:
Oxford University Press.
Prinz, J. (2010). For valence. Emotion Review,2,5–13.
Quigley, K. S., & Barrett, L. F. (2014). Is there consistency and specificity of auto-
nomic changes during emotional episodes? Guidance from the conceptual
act theory and psychophysiology. Biological psychology,98,82–94.
Reeve, J. (1993). The face of interest. Motivation and Emotion,17, 353–375.
Roseman, I. J. (2011). Emotional behaviors, emotivational goals, emotion
strategies: Multiple levels of organization integrate variable and consis-
tent responses. Emotion Review,3, 434–443.
Ross, D., & Dumouchel, P. (2004). Emotions as strategic signals. Rational-
ity and Society,16(3), 251–86.
Russell, J. A. (1994). Is there universal recognition of emotion from facial
expression: A review of the cross-cultural studies. Psychological Bulle-
tin,115, 102–141.
Russell, J. A. (2003). Core affect and the psychological construction of
emotion. Psychological Review,110, 145–172.
Scarantino, A. (2012a). Discrete emotions. from folk psychology to causal
mechanisms. In P. Zachar and R. Ellis (Eds.), Categorical and dimen-
sional models of affect, (pp. 135–154). Amsterdam: Consciousness and
Emotion book series, John Benjamins.
Scarantino, A. (2012b). How to define emotions scientifically. Emotion
Review,4, 358–368.
Scarantino, A. (2015). Basic emotions, psychological construction, and the
problem of variability. In L. F. Barrett & J. A. Russell (Eds.), The psychologi-
cal construction of emotion (pp. 334–376). New York, NY: Guilford Press.
Seidel, A., & Prinz, J. (2013). Sound morality: Irritating and icky noises
amplify judgments in divergent moral domains. Cognition,127,1–5.
Shiota, M. N., Campos, B., & Keltner, D. (2003). The faces of positive emo-
tion: Prototype displays of awe, amusement, and pride. Annals of the
New York Academy of Sciences,1000, 296–299.
Solomon, R. C. (2001). Against valence. In R. C. Solomon (Ed.), Not pas-
sion’s slave (pp. 135–147). New York: Oxford University Press.
Tracy, J. L., & Robins, R. W. (2004). Show your pride evidence for a dis-
crete emotion expression. Psychological Science,15, 194–197.
Wehrle, T., Kaiser, S., Schmidt, S., & Scherer, K. R. (2000). Studying the
dynamics of emotional expression using synthesized facial muscle
movements. Journal of Personality and Social Psychology,78, 105–119.
Zeelenberg, M., & Pieters, R. (2004). Beyondvalenceincustomerdissatisfaction:
Areviewandnewfindings on behavioral responses to regret and disap-
pointment in failed services. Journal of Business Research,57,445–455.
38 COMMENTARIES
