Emotional Responses to Art: From Collation and Arousal to ...

Review of General Psychology
Copyright 2005 by the Educational Publishing Foundation
2005, Vol. 9, No. 4, 342–357
1089-2680/05/$12.00
DOI: 10.1037/1089-2680.9.4.342
Emotional Responses to Art: From Collation and Arousal
to Cognition and Emotion
Paul J. Silvia
University of North Carolina at Greensboro
The study of emotional responses to art has remained curiously detached from the
psychology of emotions. Historically, the leading tradition has been Daniel Berlyne’s
psychobiological model, embodied by the “new experimental aesthetics” movement of
the 1970s. That theory explained hedonic qualities of art by referring to arousal-
modifying “collative properties” of art, such as complexity, novelty, uncertainty, and
conflict. Berlyne’s influence on the experimental study of aesthetics has been enor-
mous, largely for the better but also for the worse. Berlyne’s suspicion of cognitive
psychology led to an unproductive perseveration on arousal as the mechanism of
“aesthetic responses.” This article describes how appraisal theories of emotion inform
the study of aesthetics. Appraisal theories make new predictions about emotional
responses to art, expand the domain of aesthetic emotions beyond positive emotions
such as interest and enjoyment, inform other theories (e.g., prototypicality models), and
reinterpret past findings.
Emotions and art are intimately related (Tan,
1971; Tomkins, 1962), experimental psycholo-
2000). From ancient to modern times, theories
gists began tackling hedonic qualities of art
of aesthetics have emphasized the role of art in
(Berlyne, 1971a, 1972, 1974). Since then, the
evoking, shaping, and modifying human feel-
psychology of emotion and the psychology of
ings. The experimental study of preferences,
art have had little contact (see Silvia, in press-b;
evaluations, and feelings related to art has a
Tan, 2000).
long history in psychology. Aesthetics is one of
A landmark in the modern study of emotional
the oldest areas of psychological research, dat-
responses to art is Daniel Berlyne’s (1971a,
ing to Fechner’s (1876) landmark work. Psy-
1974) development of the “new experimental
chology has had a steady interest in aesthetic
aesthetics.” This tradition, which still pervades
problems since then, but art has never received
contemporary research on art, emphasized con-
as much attention as one would expect (see
trolled laboratory research involving advanced
Berlyne, 1971a; Tan, 2000; Valentine, 1962).
behavioral-science methods. I begin with a his-
The study of art and the study of emotions, as
torical overview of Berlyne and the new exper-
areas of scientific inquiry, both languished dur-
imental aesthetics, with an emphasis on how his
ing much of the last century. It is not surprising
theory explained emotional responses to art.
that the behavioral emphasis on observable ac-
This review discusses the innovative features of
tion over inner experience would lead to a ne-
Berlyne’s work, examines the theoretical as-
glect of research on aesthetics. In an interesting
sumptions that motivated the behavior-theory
coincidence, both art and emotion resurfaced in
approach to aesthetics, and considers the limi-
psychology at about the same time. As emotion
tations of this research for explaining emotions
psychologists began developing theories of ba-
and art.
sic emotions (Ekman & Friesen, 1971; Izard,
I then turn to modern theories of emotions,
particularly
appraisal
theories
of
emotion
(Lazarus, 1991; Roseman & Smith, 2001;
Scherer, 2001a). To modern researchers, it is
Paul J. Silvia, Department of Psychology, University of
clear that the psychology of emotions is a good
North Carolina at Greensboro.
starting point for exploring emotional responses
Correspondence concerning this article should be ad-
to art. After seeing what appraisal theories can
dressed to Paul J. Silvia, Department of Psychology, P. O.
say about emotional responses to art, I review
Box 26170, University of North Carolina, Greensboro, NC
27402-6170. E-mail: p_silvia@uncg.edu
recent experiments that tested appraisal predic-
342

---

EMOTIONS AND ART
343
tions about interest in poetry and visual art
psychology that surrounded his work (see
(Silvia, 2005a, 2005b). I cannot hope to address
Bolles, 1967). Berlyne did not commit to a
or resolve all of the rich issues that follow from
strong Hullian model of drive, in which drive
intersecting appraisal theories with the study of
reduction is the sole source of reward (e.g.,
art. Instead, my goal is to demonstrate the fruit-
Hull, 1952). He did, however, embrace the con-
fulness of modern emotion theories for classic
ceptual lexicon of the Hullian tradition, espe-
aesthetic problems, thus, perhaps, stimulating a
cially the assumption that reward involves the
“new” new experimental aesthetics.
reduction of activated states (be they “arousal”
or “drive”) toward a preferred state (Cupchik,
Daniel Berlyne and the New Experimental
1988). In his early model, Berlyne (1960) as-
Aesthetics
sumed that people preferred states of low
arousal, but—in an intriguing twist— he as-
Berlyne’s approach to the psychology of art,
serted that low arousal came from moderate
labeled “the new experimental aesthetics” (Ber-
levels of arousal potential. Thus, stimuli with
lyne, 1974), was an extension of his earlier
low and high levels of complexity, novelty,
theories of curiosity and exploration (Berlyne,
1960, 1967). Berlyne’s theory of curiosity is
uncertainty, and conflict caused high levels of
better thought of as a series of theories, for
arousal (see Silvia, 2006, chap. 2). As clever as
Berlyne modified his theory’s presumed mech-
this model was, it fared poorly as research cast
anisms as psychobiological research clarified
doubt on the existence of a single system of
the neurological systems involved in percep-
arousal and reward (see Berlyne, 1967).
tion, arousal, and reward (see Silvia, 2006,
By the time of his psychobiological model of
chap. 2). Some features, however, remained
aesthetics, Berlyne (1967, 1971a) assumed that
constant in his work, particularly (a) an empha-
the hedonic qualities of art came from separate
sis on a class of stimulus features known as
biological systems of reward. The first sys-
“collative variables” (Berlyne, 1960, chap. 2)
tem—the primary reward system— generated
and (b) an emphasis on arousal as a mechanism
positive affect whenever arousal potential in-
of reward and preference.
creased. This system’s activation increased as
The class of collative variables consisted of
arousal potential increased, eventually leveling
stimulus factors such as complexity, novelty,
off at a high level. As a result, things became
uncertainty, and conflict. In an innovative anal-
more appealing as complexity, novelty, and so
ysis, Berlyne (1960) proposed that these seem-
forth increased. The second system—the pri-
ingly different features had two core similari-
mary aversion system— generated negative af-
ties. First, each involved comparing pieces of
fect whenever arousal potential increased. This
information, such as incoming information with
system’s activation also increased as arousal
expected information (e.g., novelty or uncer-
potential increased, but it had a higher absolute
tainty) or different informational elements
activation threshold than the primary reward
within a stimulus field (e.g., conflict or com-
system. Thus, the primary reward system
plexity). As Berlyne (1971a, p. 69) described it,
“kicked in” before the primary aversion system.
“the word ‘collative’ . . . adverts to the fact that,
The joint operation of these two systems created
in order to decide how novel, surprising, com-
an inverted-U curve. Stimuli move from neutral
plex, and so on, a pattern is, one must compare
to positive as arousal potential goes up, but they
or collate information from two or more
sources.” Second, each collative variable has
shift from positive to negative after arousal po-
the quality of arousal potential, the ability to
tential passes an optimal tipping point.1
affect the intensity of arousal. Thus, stimuli
high in complexity, novelty, uncertainty, and
1 As in his earlier theory, Berlyne (1971a) assumed that
conflict can increase arousal, although they will
decreases in arousal were rewarding. This was enabled
not necessarily do so.
through the operation of a third system—the secondary
The hedonic qualities of stimuli, artistic or
reward system—that inhibited the primary aversion system.
otherwise, were traced to changes in arousal.
When arousal potential declines, this system inhibits the
primary aversion system. This is experienced as pleasure,
Berlyne’s (1960) theorizing is clearly rooted in
because the primary reward system is no longer antagonized
the post-Hullian developments in motivation
by a coinciding aversion system.

---

344
SILVIA
Lest these psychobiological ideas seem de-
be general across kinds of art. If complexity is
tached from the study of emotions and art, it is
an abstract, structural feature of art, then the
worth noting that a strength of Berlyne’s theo-
quadratic effect of complexity on preference
rizing was his emphasis on continuities between
should be evident for visual art, music, film,
the experience of art and the experience of
poetry, literature, and any other aesthetic con-
artlessness. The psychological dynamics of re-
text that can be characterized on a simple– com-
ward were presumed to be the same for both.
plex dimension.
This assumption mainstreamed the psychology
The new experimental aesthetics was clearly
of art and enabled the tools of hard-nosed be-
fertile. The two primary books (Berlyne, 1971a,
havioral inquiry to be applied to aesthetic prob-
1974) are rich with experimental research that
lems. Nevertheless, the new experimental aes-
catalyzed a new interest in psychology and art.
thetics was an unlikely revolution. Even in
Within a few years, the effects of the collative
hindsight, it is surprising that art resurfaced in
variables on the hedonic qualities of art had
mainstream research through a behaviorist tra-
been tested with randomly generated polygons,
dition that eschewed cognitive and emotional
patterns, musical melodies, and film clips
constructs (Berlyne, 1975; Konecˇni, 1996).
(Evans & Day, 1971; Hare, 1974; Normore,
1974); with authentic visual art (Berlyne &
The New Experimental Aesthetics
Ogilvie, 1974; Cupchik, 1974); with samples of
adults and children (Bragg & Crozier, 1974);
Berlyne used these two classes of concepts—
with cross-cultural designs comparing Cana-
the collative variables and the arousal model of
dian, Indian, and Ugandan participants (Ber-
reward—to make predictions about emotional
lyne, 1976; Berlyne, Robbins, & Thompson,
responses to art. He assumed that the collative
1974); and with comparisons of untrained par-
variables were embodied in structural features
ticipants and experts trained in visual and mu-
of art. Paintings, music, literature, and films
sical arts (Crozier, 1974; Hare, 1974). These
could be described as varying in complexity,
studies measured a broad range of variables,
novelty, uncertainty, and conflict, particularly
including semantic differential ratings, time
when these variables are viewed as representing
spent viewing or listening to something, choices
abstract qualities of information (Berlyne,
between two art objects, physiological measures
1957). A tradition of information-theory models
of “arousal” (typically skin conductance; Evans
of aesthetics (e.g., Moles, 1966) lent weight to
& Day, 1971), and factor scores formed from
this supposition. If art can be understood in
condensing large sets of subjective responses.
terms of objective stimulus features such as
An unexpected finding that emerged in early
complexity and novelty, then emotional re-
research was the difference between ratings of
sponses to art can be explained with reference to
interest and ratings of preference (also referred
the broader hedonic effects of complexity and
to as pleasingness). Preference judgments often
novelty. Berlyne (1971a, p. 176) asserted that
followed a nonlinear function, whereas interest
“the stimulus properties that apparently govern
was typically a linear function of the collative
aesthetic appreciation consist essentially of those
variables. Interest was thus anomalous for the
that constitute what we have called ‘arousal po-
model, because the interaction of the three re-
tential.’” As a result, preferences for art can be
ward systems was expected to create inverted-U
framed in terms of how collative properties of art
functions. This was an early clue that emotional
affect the arousal systems of reward.
responses to art were more differentiated than
As an example of Berlyne’s specific predic-
simply “positive” or “rewarding.” Many re-
tions, preferences for art should follow an in-
searchers up to the present day have found this
verted-U pattern. Manipulating many levels of
effect (e.g., Ellsworth & Smith, 1988; Silvia,
complexity—such as by creating random poly-
2005b), about which I have more to say later.
gons with different numbers of sides—should
yield a quadratic effect on preference. The
The New Experimental Aesthetics
highly simple and highly complex polygons
in Hindsight
should be the least appealing, whereas the mod-
erately complex polygons should be the most
Berlyne’s (1971a) approach to experimental
appealing. Furthermore, this relationship should
aesthetics left an enormous mark on later re-

---

EMOTIONS AND ART
345
search. As time has past, it has been easier to
research that also open doors for future re-
gain perspective on Berlyne and his work (see
search. Emotion psychology has developed into
Cupchik, 1988; Konecˇni, 1996). Applying the
a powerful and integrative approach to many
tough-minded methods of behavioral research
psychological topics; below I explore its fruit-
to aesthetic problems accelerated the growth of
fulness for understanding emotional responses
knowledge about psychological aspects of art
to art.
and brought aesthetic problems into the main-
stream of psychology. At the same time, certain
elements of Berlyne’s thinking inhibited re-
Cognitive Appraisals and Emotional
search directions that would have proven fruit-
Experience
ful. Cupchik (1988) pointed out that Berlyne’s
Modern emotion psychology can offer much
antipathy toward the emergence of cognitive
to the study of aesthetics. Emotion psychology
psychology (see especially Berlyne, 1975)
emerged, in part, from the recognition that emo-
closed the door to studying how thoughts and
tions are not merely states of high arousal (e.g.,
emotions interact in aesthetic contexts. Else-
Duffy, 1934). Instead, emotions are evolved,
where (Silvia, 2006, chap. 2) I have reviewed
pancultural psychological mechanisms for deal-
the growth of Berlyne’s (1960, 1978) thinking
ing with “fundamental life tasks” (Ekman,
about arousal and reward, particularly his reluc-
1992). Modern theories offer sophisticated
tance to abandon a model of reward founded on
models of the nature, functions, and processes
shifts in arousal despite the changing tides of
of emotion. There are many theories of what
motivational psychology.
emotions are and where they come from (see
Were Berlyne alive today, he probably would
Silvia & Warburton, 2006). In recent years,
not be an arousal theorist. The concept of “gen-
appraisal theories of emotion have emerged as
eral arousal” has faded into history. It is now
a leading perspective on emotions (Ellsworth &
widely known that markers of “arousal” are
Scherer, 2003; Lazarus, 1991; Roseman, 2001;
essentially decoupled (see Neiss, 1988). For
Scherer, 2001a). Appraisal theories have a long
example, it is common to find decoupling
history, dating to early research on stress and
within different “arousal” measures (such as
coping (Schorr, 2001), and, at least according to
electrodermal and cardiovascular responses)
appraisal theorists, “there is, at present, no via-
and even within the same system (such as low
ble alternative to an appraisal (in the broad
correlations among heart rate, systolic blood
sense of the word) explanation for the general
pressure, and diastolic blood pressure; e.g.,
prediction of the elicitation and differentiation
Gendolla & Kru¨sken, 2001). For these and other
of emotions” (Scherer, 2001b, pp. 389 –390).
reasons, many researchers have closed the
books on arousal as a legitimate, explanatory
construct (Neiss, 1988). This is one reason why
Appraisals of Events
some theoretical debates—such as conflicts be-
tween Berlyne’s arousal model and a prototype
The central assumption of all appraisal theo-
model (e.g., Boselie, 1991; Martindale, Moore,
ries is that evaluations of events, not events
& Borkum, 1990; North & Hargreaves, 2000)—
themselves, are the local cause of emotional
are unproductive; the psychobiological assump-
experience (Roseman & Smith, 2001). This as-
tions of Berlyne’s arousal model are known to
sumption firmly establishes appraisal theories
be wrong.
within a subjective, contextual approach to un-
Berlyne’s arousal explanations have faded,
derstanding emotions. It may seem needlessly
but the fact remains that many experiments have
extreme to assume that events do not cause
revealed effects of novelty, complexity, and un-
emotions, but it is easy to demonstrate that
certainty on feelings of interest and enjoyment.
emotions arise from subjective appraisals of
If the collative variables do not affect prefer-
events (Roseman & Evdokas, 2004). In fact, it
ence by virtue of their effects on arousal, how
is hard to explain interperson and intraperson
do they affect aesthetic preferences? This leaves
variability in emotional experience by referring
the psychology of art with orphaned data in
to objective features of events. In any situation,
need of a new theory. In the following sections,
different people will respond with different
I propose some new ways of thinking about past
emotions. And, in similar situations, the same

---

346
SILVIA
person will have different emotions at different
one’s potential to cope with the event as high,
times. This situation is familiar to researchers in
such as through self-assertion or aggression (see
experimental aesthetics, given the wide vari-
Kuppens, Van Mechelen, Smits, & De Boeck,
ability in emotional responses to art. Thus, it
2003).
soon seems self-evident that objective events
Some emotions have small, simple appraisal
are poor explanations of emotions.
structures. The emotion of interest, for example,
A subtler reason why events do not causes
has two appraisals: an appraisal of something as
emotions concerns the diversity of events that
new, unexpected, or complex and an appraisal
cause the same emotion (Roseman & Smith,
of one’s ability to comprehend the new, com-
2001). For any given person, a vast set of events
plex thing (Silvia, 2005a, 2005b, 2006). Other
can lead to the same emotion. For example,
emotions have complex appraisal structures.
someone might be happy on (a) hearing that a
The appraisal structure of shame, for instance,
friend just had a baby, (b) listening to a new
involves appraising an event as (a) relevant, (b)
song by a favorite composer, (c) having a good
caused by the self, (c) urgent, (d) involving bad
dinner with old friends, (d) finding a new short-
outcomes, (e) committed through negligence,
cut for the morning commute, (e) having a
and (f) inconsistent with one’s personal stan-
paper accepted for publication, and (f) imagin-
dards (see Scherer, 2001a). Similar emotions
ing hypothetical triumphs over his or her nefar-
can be grouped into families according to their
ious foes. What objective, surface features do
similar
appraisal
structures
(Ellsworth
&
these events have in common? (Indeed, what
Scherer, 2003). For example, the emotions in
are the “objective features” of fictional, fantas-
the “hostility triad”—anger, disgust, and con-
tical events?) Appraisal theories, in contrast,
tempt (Izard, 1977)— have several appraisal
easily handle these problems because the under-
components in common.
lying subjective judgments—in this case, ap-
praising an event as congruent with a goal and
Some Quick Contrasts
as likely to continue (Lazarus, 2001)—are the
same in each situation. Different events will
This brief overview of appraisal theories has
create the same emotions if the person appraises
already highlighted some unique ideas relative
each event similarly. As before, the subjective
to the new experimental aesthetics. First, we see
psychological structure of an event is what is
contrasts with Berlyne’s emphasis on objective
critical to emotions.
stimulus features, such as objective complexity
and novelty, as causes of aesthetic responses.
Differentiation of Emotions
Appraisal theories would recast this by contend-
ing that an “objectively complex” object will
A second assumption of appraisal theories is
affect aesthetic responses only inasmuch as a
that emotions can be differentiated according to
person subjectively appraises the object as
the evaluations that cause them (Roseman &
“complex.” The person’s subjective appraisals
Smith, 2001). This simply means that different
mediate between emotions and the outer world.
emotions are brought about by different groups
Second, emotional responses to art are traced to
of evaluations, known as appraisal structures.
the unfolding of a set of cognitive appraisals,
Each emotion has a distinct appraisal structure
not to the operation of psychobiological reward
composed of a set of appraisal components.
and punishment systems (Berlyne, 1971a). A
Common appraisal components include ap-
variable’s “collative properties” are not enough,
praising something as being unexpected, rele-
according to appraisal theories, to create an
vant to a goal, controllable or uncontrollable,
emotion. A series of subjective evaluations is
inconsistent with personal standards, and one’s
required.
own or another person’s fault (see Roseman,
Finally, many emotions are of interest in the
2001; Scherer, 2001a; Smith & Ellsworth,
appraisal approach. The Berlyne tradition re-
1985; Weiner, 1986). The appraisal structure of
stricted itself to enjoyment, interest, and aver-
anger, for example, consists of (a) appraising an
sion. This small class of affective responses
event as incongruent with a goal or motive, (b)
follows from the model of motivation underly-
appraising an agent (typically another person)
ing the research. In arousal and drive models,
as blameworthy for the event, and (c) appraising
motivational outcomes were “rewarding” or

---

EMOTIONS AND ART
347
“aversive” (Berlyne, 1967). This leaves only
that cause a sense of dysfluent processing (Sil-
simple positive and negative states as possible
via, 2005b). Appraising something as new, un-
subjective reactions. In contrast, appraisal the-
familiar, uncertain, complex, inconsistent, in-
ories focus on a wide range of emotions such as
choate, and mysterious typifies this first ap-
happiness, interest, surprise, awe, anger, fear,
praisal. The similarity to Berlyne’s (1960) class
sadness, shame, guilt, disgust, contempt, and
of “collative variables” is apparent (see Silvia,
embarrassment (Ellsworth & Scherer, 2003).
2006, chap. 2). The second appraisal is a cop-
The psychology of art has had little to say about
ing-potential check (Scherer, 2001a), which re-
most of these emotions, despite their apparent
fers to an appraised ability to understand the
relevance to the experience of art, largely be-
new, unfamiliar, complex thing identified by the
cause of the assumptions of the collation-and-
first appraisal.2 A coping-potential check, un-
arousal model.
like the novelty check, has no equivalent in the
Berlyne tradition. This appraisal is explicitly
Evidence for Appraisals
subjective and metacognitive, because it refers
and Interest in Art
to judgments regarding comprehension, under-
standing, and meaningfulness (cf. Leder, Belke,
It is one thing to show that appraisal theories
Oeberst, & Augustin, 2004; Russell, 2003).
of emotions can say much about aesthetic prob-
Thus, the contrasts between an appraisal ap-
lems; it is quite another to develop testable
proach and a collation–arousal approach should
hypotheses that demonstrate the fruitfulness of
be more obvious.
an appraisal approach. This section reviews new
research rooted in appraisal theories on emo-
Basic Tests of the Appraisal Model
tional responses to art. This research examined
appraisals of interest (Silvia, 2006), the emotion
of Interest in Art
associated with exploration, intrinsic motiva-
Four experiments have tested how these two
tion, curiosity, and learning (Fredrickson, 1998;
appraisals jointly affect interest in art (Silvia,
Hidi, 1990, 2001; Krapp, 1999; Sansone &
2005b). The first study assessed interest in ran-
Smith, 2000; Silvia, 2001, 2005b; Tomkins,
domly generated polygons, a venerable staple of
1962). It is the prototypical emotion in a class of
experimental aesthetics research. Coping poten-
“epistemology-related” emotions (Ellsworth,
tial was operationalized with individual differ-
2003; Keltner & Haidt, 2003; Keltner & Shiota,
ences. People completed a brief scale that as-
2003). Interest is an important emotion in the
sessed their ability to understand complex art,
psychology of art. The Berlyne (1971a, 1974)
and then they viewed random polygons that
tradition has extensively researched interest in
ranged from simple (4 sides) to complex (160
response to art, and Tan (2000) proposed that
sides). People were asked to select the “most
interest is the emotion most central to aesthetic
interesting” polygon. Appraisals of ability to
experience. Of all of the emotions, then, interest
understand significantly predicted the complex-
seems like a good one with which to explore the
ity of the most interesting polygon. As people
usefulness of an appraisal approach to emo-
felt more able to understand complex art, they
tional aspects of aesthetics.
selected highly complex polygons as being the
The central assumption of an appraisal ap-
most interesting. A second group of participants
proach, as noted earlier, is that emotions arise
was asked to select the “most enjoyable” poly-
from subjective appraisals of events. In the case
gon. Appraisals of ability to understand pre-
of interest resulting from art, appraisal theories
dicted the level of complexity that people found
predict that people will experience interest
interesting but not the level that people found
when they make the appraisals that generate
interest. This naturally raises the question of
interest’s appraisal structure. Past theories and
2 A sequential model of appraisal, in which early apprais-
recent research (Ellsworth & Smith, 1988; Sil-
als provide inputs to others, is presumed in this model of
via, 2005) have shown that interest’s appraisal
interest and appraisal. Controversies over the sequential
structure consists of two primary components.
nature of appraisals have not yet been settled (see Roseman
& Smith, 2001; Scherer, 2001a), but early research
The first is a novelty check (Scherer, 2001a),
(Scherer, 1999) suggests it is reasonable to assume that
which refers to a class of disruptive variables
some appraisals inform others.

---

348
SILVIA
enjoyable. This is consistent with past research
to understand were unrelated to interest. For
on differences between interest and enjoyment
complex pictures, however, ability strongly pre-
(cf. Ellsworth & Smith, 1988; Russell, 1994;
dicted interest: Interest increased as appraised
Russell & George, 1990; Russell & Gray, 1991;
ability increased. The relationships between ap-
Silvia, 2006, chap. 1), and it supports the ap-
praisals and interest were unaffected when pos-
praisal model’s assumption that each emotion
sible “third-variable” confounds (trait curiosity
has a distinct appraisal structure.
and
positive
affectivity)
were
controlled
A second study examined interest in post-
statistically.
modern poetry. People were asked to read a
A fourth experiment showed that appraisals
poem by Scott MacLeod, a modern writer of
predicted behavioral expressions of interest. In
experimental language art. The poem, titled
this experiment viewing time, a measure popu-
“The Whitest Parts of the Body” (MacLeod,
larized by Berlyne (1971a, 1974), was used to
1999), is abstract, complex, and obscure. For
measure interest. People viewed random poly-
example, the poem’s first stanza reads “such
gons, ranging from simple to complex, on a
daring against men / with a throat so big /
computer screen. Their appraised ability to un-
separated by a hundred years / full of misfor-
derstand complex art had been measured earlier
tune: the bloody / flux. taken by a fit of madness
in the semester. People could view each poly-
/ prone to eating human flesh / and measured, in
gon for as long as they wished; when the image
due course, / by naturalists.” The final stanza
became boring, they could press a key to move
reads “it carries out terrible ravages / lost, and
to the next image. Appraisals of ability inter-
usually permanently / lost, among whom,
acted with the polygons’ complexity to predict
wringing her hands / and crying out / to heaven,
viewing times. People spent the most time
/ was the girl’s mother.” In a control condition,
viewing an image when the image was highly
people read the poem, appraised their ability to
complex and they felt able to understand com-
understand it, and reported their feelings of in-
plex art; the pattern is displayed in Figure 1.
terest. In a high-ability condition, people re-
This supports the appraisal model’s predictions
ceived information that unlocked the poem’s
with behavioral data, thus extending the past
meaning. They were told that the poem comes
studies beyond self-reports of interest.
from a book of poems about killer sharks. As
expected, people in the high-ability group found
the poem more interesting than did people in the
Within-Person Relationships of Appraisals
control group. Mediation analyses showed that
and Interest in Art
appraisals of ability to understand fully medi-
ated the effects of the manipulation on interest.
Another experiment extended this research
The extra information about the poem’s mean-
by examining within-person relationships be-
ing increased perceived ability to understand
tween appraisals and interest (Silvia, 2005a).
the poem; understanding, in turn, increased in-
Between-person effects show relationships be-
terest. This experiment thus shows a causal ef-
tween appraisals and interest within the sample
fect of appraised coping potential on interest in
as whole. Strictly speaking, though, appraisals
poetry.
should predict interest at the within-person level
A third experiment extended the appraisal
(Scherer, 2001a). Between-person and within-
model to visual art. Participants viewed pictures
person effects are mathematically distinct, and
taken from books and journals of modern “vi-
conventional between-person analyses need not
sual poetry”; the styles included concrete po-
replicate at the within-person level (see Nezlek,
etry, asemic writing, automatic art, and typo-
2001). In this study, people viewed more
graphic art (e.g., Harris, 2001; Huth, 1990;
than 30 pictures of abstract visual poetry. They
Morin, 2003; Selby, 2003; Topel, 2002). Half of
rated each image for interest and for appraisals
the pictures were fairly simple, and half were
of complexity and of ability to understand the
fairly complex (based on pretesting). For each
image. The large numbers of responses for each
picture, people gave ratings of interest and abil-
participant enabled an analysis of within-person
ity to understand the picture. As expected, in-
correlations.
terest depended on both complexity and coping
Using multilevel random-coefficient model-
potential. For simple pictures, ratings of ability
ing (Hox, 2002; Luke, 2004), I estimated the

---

EMOTIONS AND ART
349
Figure 1.
Effects of appraised ability to understand on time spent viewing random polygons
(in milliseconds) that varied in complexity (Silvia, 2005b, Experiment 4). People spent the
most time viewing a polygon when (a) the polygon was highly complex and (b) they felt able
to understand complex art.
within-person covariance of appraisals and in-
Millis, 2001). A typical finding in this area is
terest. Both appraisals strongly and significantly
that people with experience in art prefer works
predicted interest, thus providing a within-per-
higher in complexity than do people without
son confirmation of the appraisal predictions.
expertise in art (Hare, 1974; Walker, 1980).
Figure 2 shows how the two appraisals pre-
This finding challenges arousal models. One
dicted interest. This histogram depicts the dis-
needs to assume that the thresholds relating
tributions of unstandardized regression coeffi-
arousal and reward differ between experts and
cients for complexity and ability to understand.
novices or that complex stimuli are more or less
It shows that, for each person in the sample,
arousing to experts. An appraisal position, in
these appraisals positively predicted interest. It
contrast, easily explains this effect. Appraisal
also shows that these within-person relation-
theories, not surprisingly, trace group differ-
ships differed between people. Although com-
ences in emotional responses to differences in
plexity and coping potential positively pre-
appraisals (Roseman & Smith, 2001; van Ree-
dicted interest for everyone, they did so more
kum & Scherer, 1997). If people with training in
strongly for some people than for others. Ex-
art find some kinds of art more interesting, then
ploring why appraisal– emotion relationships
it must be because they are more likely to ap-
differ between people is an intriguing, cutting-
praise the art in ways that generate interest.
edge area of appraisal research (see Kuppens et
Knowledge about art will affect the emotional
al., 2003).
experience of art.
An experiment (Silvia, in press) tested the
The Appraisal Basis of Expertise
appraisal explanation for why artistic training
affects interest. It seems straightforward that
A final experiment (Silvia, in press) applied
training increases appraisals of coping potential,
the appraisal model to the study of artistic ex-
the evaluation of one’s resources and abilities
pertise. Many studies have revealed expert–
for understanding what is presented. This is,
novice differences in emotional responses to art
after all, the purpose of learning about the his-
(e.g., Hekkert & van Wieringen, 1996a, 1996b;
tory and theory of art. Given their higher ap-

---

350
SILVIA
Figure 2.
Histograms of within-person relationships between appraisals and interest. The
x-axis displays the size of within-person unstandardized regression coefficients that estimate
the effects of appraisals on interest. Thus, for the entire sample, appraisals of complexity and
appraisals of ability to understand had uniformly positive relationships with interest. More-
over, the range of effects was greater for appraisals of ability to understand.
praised ability to understand, experts should
Revisioning Past Research on
find the art more interesting. Students with ex-
Experimental Aesthetics
tensive art training or with no training viewed
simple and complex paintings. For each paint-
Thinking about emotional responses to art in
ing, they gave ratings of interest and of apprais-
terms of appraisals recasts some past research in
als. Figure 3 (top panel) shows the effects of
experimental aesthetics. Though not explicitly
expertise on interest. As expected, art experts
motivated by appraisal theories, many past ex-
and novices differed in how interesting they
periments demonstrate the value of an appraisal
found the complex visual art but not in how
approach. One example is the small body of
interesting they found the simple visual art.
work on how titles influence the experience of
Additional analyses examined whether these
art. Russell and Milne (1997) reasoned that peo-
group differences were due to appraisals. The
ple enjoy art more when they grasp its meaning.
two groups did not differ in how complex they
In one experiment, they presented reproductions
found the art. As in past research (see Crozier,
of paintings accompanied by their real titles,
1974), there was high agreement on appraised
accompanied by fake titles, or without titles.
complexity. However, the groups differed in
People judged the paintings as more meaningful
their appraised ability to understand the visual
when they read the real titles relative to fake
art. Figure 3 (bottom panel) shows how exper-
titles and no titles. Furthermore, the real titles
tise affected appraisals of ability to understand.
marginally increased enjoyment and interest in
As expected, people with artistic training felt
the paintings. In later experiments, Russell
more able to understand the visual art, and this
(2003) compared the effects of titles and de-
difference in understanding corresponded with
tailed descriptions of paintings. People who
differences in interest. This is an example of
read a paragraph describing the artist’s work
how the appraisal perspective can make new
and its relevance for the painting found the
predictions.
painting more meaningful and enjoyable than

---

EMOTIONS AND ART
351
connects this seemingly isolated body of work
to broader theoretical problems in the study of
aesthetics. These studies, which examined a
specific, isolated problem (how titles affect re-
sponses to art), can be brought under a broader
theoretical umbrella. This foreshadows the po-
tential for an appraisal approach to integrate
disparate pockets of research in experimental
aesthetics.
Some Implications of Appraisal Theories
for Emotional Responses to Art
So far, I have reviewed the basic assumptions
of appraisal theories and considered how they
can make new predictions about interest in art.
This section explores some of the more salient
implications of an appraisal model for under-
standing emotional responses to art.
Rethinking “Aesthetic Response”:
Interest, Enjoyment, and Distinct
Emotions
A central claim of appraisal theories is that
there are many distinct emotions (Lazarus,
1991; Roseman, 2001; Scherer, 2001a). These
emotions have different appraisal structures,
different functions, and usually different ex-
Figure 3.
Effects of artistic expertise on interest in simple
pressive and biological signatures. It is imprac-
and complex visual art (top) and on appraised ability to
tical, from an appraisal perspective, to speak of
understand the images (bottom).
global, undifferentiated emotional concepts.
Emotional concepts within the Berlyne tradi-
tion, however, have emphasized generic affec-
did people who received no information. Millis
tive terms, such as preference, reward value,
(2001) found that titles had the largest effect on
aesthetic response, hedonic response, and affec-
enjoyment when they fostered an elaborated,
tive response. These terms are too ambiguous
coherent representation of the painting; titles
given the different causes and effects of distinct
that simply described the painting’s central ob-
emotions. Does a painting that has “negative
jects and activities had no effect.
hedonic value” make someone sad? Anxious?
These experiments offer converging support
Disgusted or angry? Finding art disgusting is
for an appraisal model. People judged paintings
clearly different from finding art saddening. It is
as easier to understand when they could read the
unproductive to lump different emotions, each
paintings’ titles and descriptions. As a result,
with distinct causes and consequences, under an
people found the paintings more enjoyable and
umbrella emotional term.
interesting. These studies clearly parallel the
The study of interest and enjoyment shows
experiments described earlier, in which apprais-
the perils of ignoring differences between dis-
als of one’s ability to understand art made art
crete emotional states. Research conducted by
more interesting (Silvia, 2005a, 2005b). More-
Russell (1994; Russell & George, 1990; Russell
over, their manipulations provide additional ex-
& Gray, 1991) shows that ratings of interest
perimental support for a largely correlational
diverge from ratings of enjoyment and that ge-
literature. At the same time, the appraisal model
neric measures of preference mask these differ-

---