Content uploaded by Jean-Luc Jucker
Author content
All content in this area was uploaded by Jean-Luc Jucker on Feb 01, 2019
Content may be subject to copyright.
© Koninklijke Brill NV, Leiden, 2011 DOI: 10.1163/156853711X568716
Journal of Cognition and Culture 11 (2011) 115–136 brill.nl/jocc
Cognitive Constraints on the Visual Arts:
An Empirical Study of the Role of Perceived
Intentions in Appreciation Judgements
Jean-Luc Jucker* and Justin L. Barrett
Centre for Anthropology and Mind, University of Oxford, 64 Banbury Road,
Oxford OX2 6PN, UK
* Corresponding author, e-mail: jean-luc.jucker@anthro.ox.ac.uk
Abstract
What influences people’s appreciation of works of art? In this paper, we provide a new cognitive
approach to this big question, and the first empirical results in support of it. As a work of art
typically does not activate intuitive cognition for functional artefacts, it is represented as an
instance of non-verbal symbolic communication. By application of Sperber and Wilson’s
(1986/1995) Relevance eory of communication, we hypothesize that understanding the
artist’s intention plays a crucial role in intuitive art appreciation judgements. About 60 works of
fine art, representing a wide range of periods, genres and styles, were selected in the permanent
exhibitions at Tate Britain in London, and rated by more than 500 visitors for goodness and
understanding of the artist’s intention. Results suggest that works of art whose artist’s intention
is easy to understand tend to be preferred over those with more obscure intentions, even when
controlling for familiarity effects.
Keywords
Art, artefact, function, intention, Relevance eory, symbolic communication
Paintings, sculptures and installations are human-made objects or sets of
objects. What, then, characterizes works of art as a specific kind of artefact?
From infancy babies automatically represent artefacts partly in relation to how
they might be manipulated, apparently as a way to conceptualize their possible
use or function (Gibson, 1979/1986; Rochat, 1989). Similarly, adults assess
artefacts partly according to their function (Keil, 1989; Rips, 1989).1 is
springboard from manipulability to functional utility is absent from most
works of art however, as it is for printed text or signs. Typically, works of art
1 On the tendency of children and adults to think about objects in terms of purpose and
function, see also Kelemen (1999); Kelemen and Rosset (2009).
116 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
cannot be readily manipulated, and are neither created nor used to achieve
purely practical goals, which, for example, distinguishes them from tools, fur-
niture, or dwellings. In other words, although perceived as intentionally cre-
ated by human agents, the purpose of works of art is not clear, and cries out
for an explanation. For this reason, we hypothesize that works of art are intui-
tively processed as acts of non-verbal symbolic communication. If so, under-
standing the artist’s intention is at the core of art appreciation.
To clarify, let us contrast a tool with a work of art. A tool will be recog-
nized as such if it carries out a specific function; and if this tool carries out
this function, we directly infer that its maker intended it to do so. is tool,
then, may be assessed without speculation regarding the maker’s intention;
recognizing its function is sufficient to categorize it and use it. For a work of
art, the contrary is true: as it carries out no obvious precise function, it can-
not be assessed without speculation about the artist’s intention. In other
words, a work of art would be assumed to communicate something, which
would have to be inferred from the artist’s intention.
In this introduction, we review studies which show that intention plays an
important role in the assessment of artefacts, especially when their form or
function are ambiguous, such as in works of art. We then propose to apply
Sperber and Wilson’s (1986/1995) Relevance eory of communication to
the visual arts. We finally make a series of predictions for intuitive apprecia-
tion judgements in the visual arts.
Although theoretically applicable to all arts, our approach is addressed to
the visual arts only. at said, the words “art” and “visual arts” will be used as
synonyms in this article, as in common usage. Furthermore, our approach is
limited to works of art that are artefacts that lack practical functionality. We
believe, however, this circumscription to capture the bulk of what is com-
monly regarded as art.
Bloom’s eory of Artefact Categorization
Artefact categorization is a debated field with contrasting empirical results
(for a review including classic as well as recent studies, and a critique of arte-
fact categorization, see Malt and Sloman, 2007). Many researchers have
emphasized the importance of form and function in artefact categorization,
however. If an object is identical in form to another object, and if it carries
out the same function, it seems reasonable to assume that it will be catego-
rized in the same class of artefact. at said, Bloom (1996) showed the limi-
tations of this model. For instance, artefacts can be highly dissimilar in shape,
but still categorized under the same kind (e.g., chairs can have many forms,
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 117
but they are still considered as belonging to the same class, that of chairs).
Furthermore, taking into account the function of an artefact does not resolve
the problem, because two things can have the same function without being
categorized as the same kind (e.g., other things than chairs may be used to sit
on, but are not considered as belonging to the same class). Finally, an artefact
that has ceased to carry out its function is not necessarily considered as
belonging to a new kind (e.g., a broken chair is still, to a large extent, a
chair).
On the basis of an art theory (Levinson, 1979, 1993), Bloom (1996) then
proposed a new theory of artefact categorization, in which the decisive factor
is claimed to be the intention of the artefact’s maker. According to this the-
ory, an artefact actually is categorized as belonging to a kind X if, and only if,
its maker intended it to belong to that kind (or, more precisely, if it is recog-
nized that its maker intended it to belong to that kind). is does not mean,
of course, that considerations of form and function play no role in artefact
categorization; most of the time, form and function actually constitute good
indicators of what the intention of the artefact’s maker was (e.g., “If this thing
looks and is used like a chair, then it was intended to be a chair”). More sim-
ply, it means that taking into account the intention of the artefact’s maker
allows one to avoid some problems with classic approaches to artefact catego-
rization. As we have seen, similarity of form and function are not sufficient for
artefact categorization, because two objects may be dissimilar in form, but
belong to the same kind, and two objects may be similar in potential function,
but belong to different kinds. If one considers the intention of the artefact’s
maker, categorization appears to be less problematic: two things may be dis-
similar in form, but made with the same intention (and therefore considered
as belonging to the same kind), and two things may be similar in function, but
made with different intentions (and therefore considered as belonging to dif-
ferent kinds).
Bloom’s (1996) theory has been followed by a number of experimental
studies that support the idea that intention plays an important role in arte-
fact categorization and appreciation. A first set of developmental studies, in
which pictorial stimuli were used, seems particularly relevant to art. For
example, Bloom and Markson (1998) showed that, if similarity of form is
somehow important in children’s appreciation of pictures, it is neither neces-
sary nor sufficient. In one of the tasks of this study, children were asked to
draw objects similar in form, such as a balloon and a lollipop; after another
task, they had to name their drawings. Given the age of the participants
(3–4-year-old), the drawings were very simple, and so similar that it was
impossible to distinguish them on the only basis of form. However, a
118 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
significant percentage of the participants named the drawings correctly, sug-
gesting that children distinguished their creations by reference to what they
intended to represent. Some children also strongly protested when the experi-
menters named their drawings as something that they did not intend to rep-
resent (e.g., a balloon, when the children wanted to draw a lollipop).
In another study, Preissler and Bloom (2008) wanted to determine whether
this appeal to intentions also applies to the appreciation of pictures that chil-
dren did not draw by themselves. e participants (2-year-olds) were intro-
duced to two novel objects, similar in form. e experimenter then placed
these objects in two opaque boxes (one open, one closed), and drew a picture
which could represent either of them. In one condition, the experimenter
was staring into the open box while drawing; in another, she stared at the
wall behind the closed box. e drawing was then labelled with a novel name,
the two objects were removed from the boxes, and the children were asked to
indicate which of them could be named as the object represented in the pic-
ture. In the first condition, the children significantly extended the novel
name to the object that was in the open box; in the other condition, the
novel name was not significantly extended to the object in the closed box.
If these results confirm that children take intention into account in order
to understand representational displays, other studies investigated whether
this also applies to their appreciation of actual artefacts. For example, Gelman
and Bloom (2000) presented a series of nine objects to 3- and 5-year-olds.
e two conditions of these objects were their “intentional” or “accidental”
modes of production (e.g., a newspaper intentionally folded into the shape of
a hat, or a newspaper that, following a random accident, appears to be shaped
as a hat). Although the shapes of the objects were identical in the two condi-
tions, the items presented as intentionally created were significantly more
likely to be named as the “target” artefact (e.g., a hat), than the ones that
were supposed to be accidental.
Sperber and Wilson’s Relevance eory of Communication
Bloom’s (1996) theory of artefact categorization suggests that intention plays
an important role in intuitive judgements about artefacts, especially when
their form, function, or mode of production is ambiguous. From this point
of view, works of art that our approach targets typically are ambiguous. Being
human-made objects, works of art activate intuitive cognition for artefacts,
but at the same time frustrate functional expectations associated with arte-
facts; the creator’s intended function for his or her creation cannot be simply
“read off” of the work of art. How, then, are works of art represented by
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 119
human minds? We suggest that they are considered as acts of non-verbal sym-
bolic communication, in which case Sperber and Wilson’s (1986/1995) Rele-
vance eory might be a useful theoretical framework to approach them.
Relevance eory, as formulated by cognitive anthropologist Dan Sperber
and linguist Deirdre Wilson, essentially constitutes a development of the
Gricean “maxim of relevance”, and has been applied first in the field of prag-
matics. at said, Sperber and Wilson presented their theory as a general
model of human communication, and Yus Ramos (1998) showed that it was
successfully applied in other fields.
According to Relevance eory, to communicate is to make explicit an
intention (the intention to communicate, and the intention to communicate
something in particular), and successful communication occurs when this
intention is correctly inferred from the evidence, that is, from the utterance
or behaviour in question. However, as most of the time a number of different
inferences may be drawn from the evidence, communication is also con-
strained by the Principle of Relevance. According to the Principle of Rele-
vance, communicating goes along with an expectation of relevance: people
pay attention only to information which may have an effect in a given con-
text or, in cognitive terms, to information which is “worth processing”. Fur-
thermore, according to Sperber and Wilson, there are degrees of relevance,
and these may be described in terms of a cost-benefit relationship: the most
relevant propositions in a given discourse are those that (1) have the greatest
cognitive impact in the context and (2) are the easiest to process. In Sperber
and Wilson’s words, relevance is geared to “the processing of information
which is likely to bring about the greatest contribution to the mind’s general
cognitive goals at the smallest processing cost” (1995: 48).
We propose to apply Relevance eory to art appreciation. First of all, we
suggest that works of art, much like utterances, are intuitively assessed as acts
of symbolic communication: these particular artefacts, being intentionally
created through human agency, but lacking ordinary functionality, would be
thought to “mean something”. According to Relevance eory, human com-
munication carries an expectation of relevance; in our domain of interest, that
would mean that works of art are expected to communicate something that is
relevant or, in other words, worth processing. Furthermore, according to Rel-
evance eory, successful communication occurs when the speaker’s intention
is correctly inferred from the utterance. Understanding the artist’s intention
would, thus, be a crucial factor in assessing the relevance of a work of art.
It is easy to see the implications of such a model in the domain of art appre-
ciation. Some works of art, in which the artist’s intention is too difficult to
understand, would not be recognized as following the conventions of relevant
120 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
communication and, therefore, not highly appreciated. We might hear a viewer
exclaim, “I just don’t get it”, suggesting frustration with a failure of communi-
cation. Other works may present an easily understood “message” but the mes-
sage may fail to have much cognitive impact, contributing little inferentially.
Finally, works which strike a good balance between cognitive processing effort
and cognitive impact would be naturally more appreciated and successful.
Predictions
A Relevance eory approach provides resources for generating testable pre-
dictions for the cognitive constraints on appreciation judgements in the visual
arts. For this first study, we concentrated on the three following predictions
and controlled for familiarity effects at the same time.
1. Works of art whose artist’s intention is easy to understand will be appreciated
more highly than works of art whose artist’s intention is difficult to under-
stand. Following Relevance eory, the appreciation of a work of art
should depend on the relationship between (1) the cognitive cost that is
required to understand the artist’s intention and (2) the cognitive contri-
bution that understanding the artist’s intention could bring in return. If
true, a work of art whose artist’s intention is easy to understand should be
appreciated more highly than a work of art whose artist’s intention is dif-
ficult to understand, at least if both have the potential to bring back an
equal cognitive contribution. On the contrary, works of art that are not
clearly intentional or whose artist’s intention is not readily understand-
able, are unlikely to be considered as such or to be highly appreciated.
What makes a work of art “clearly intentional” and “readily understand-
able” however? We propose that the salience and the clarity of the artist’s
intention are influenced by at least two factors, which correspond to the
other predictions tested.
2. Works of art that are thought to have required a lot of effort and skill will be
appreciated more highly than works of art that are thought to have required
little effort and skill. Human products that appear to have required a lot of
effort and skill to produce seem to be naturally admired by people.2 In the
visual arts, it makes a difference to know that a painting was not achieved
in one day, but that it required several months of hard work.3 It also makes
2 On this point, see Algoe and Haidt (2009).
3 Kruger et al. (2001, p. 91) argued that “[perceived] effort is used as a heuristic for quality”.
In one study, two abstract paintings were considered better when they were thought to have
required more time when created (Kruger et al., 2004, Experiment 2).
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 121
a difference to realize that the same painting could not have been made by
anyone – even with more time at disposal – but only by a talented artist.
But why? From our point of view, the effort and skill that went into a work
of art do not only trigger respect and admiration. More importantly, they
constitute indicators and clarifiers of intentionality. For example, if a paint-
ing was allocated a lot of time and energy, it was certainly created on pur-
pose, and creating it “just so” mattered to the artist. As a consequence,
viewers may assume that every aspect of this work is unambiguously
intentional. e communicator (the artist) evidently thought that the
message mattered enough for this high degree of investment – or at least,
that is what the audience assumes according to Relevance eory. In con-
trast, a sketch or an unfinished painting may tacitly suggest that the com-
ponents of the painting are not all that important or deliberate, and that
the message is less relevant. Furthermore, if the effort that went into a
work of art indicates the presence of an intention, the skill that was con-
jointly required probably conveys the precise nature, or content of that
intention. As many beginners in art have bitterly experienced, there is
actually an important gap between intending to paint something and
actually painting it. e skilled artist, on the contrary, has the capacity to
convey exactly what he or she intends to convey, as if each brush stroke is
supporting and clarifying his or her precise intention.4
3. ere will be a difference in art appreciation between art specialists and non-
specialists. Because of their knowledge of art history and theory, specialists
are less rigid in their definition of art than non-specialists. For those who
know Marcel Duchamp’s Fountain and regard it as an iconic work of mod-
ern art, there is no problem to considering more recent, but similar candi-
dates, such as Tracey Emin’s My Bed, as works of art too; on the contrary,
for those who have never heard about readymades and their meaning in
modern art, it will be difficult to see more in these two works than a uri-
nal and an unmade bed. Furthermore, because of their training, art spe-
cialists can understand more easily what a work of art is about, even when
4 e amount of effort and skill that is thought to have gone into a work of art is certainly not
the only factor that helps the representation of the artist’s intention. For example, most works of
arts are accompanied by some type of information, such as titles given by the artists or
interpretative texts provided by gallery curators. Just as the amount of effort and skill which are
thought to have gone into a work of art, these pieces of information probably constitute
efficient indicators and clarifiers of intentionality. For example, an abstract painting may seem
meaningless until its title is taken into account, and a minimalist work of art may seem poor
until a notice explains what it took to actually make it. On this point see Leder et al. (2006), who
showed that elaborative titles improve the understanding of abstract paintings.
122 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
the artist’s intention is very obscure. As a direct consequence, they will tend
to be more “generous” in their judgements than non-specialists. For exam-
ple, they will know that an apparently mediocre piece actually required a
lot of effort and skill, whereas a non-specialist would be more inclined to
say that “a child could have done that”.5
As there is evidence that mere exposure to a work of art influences its appre-
ciation (Maslow, 1937; Temme, 1984; Cutting, 2003), we controlled for
familiarity effects alongside our three predictions.
Method
e design of the study was a within-between subjects questionnaire: we
selected a series of works of art that were all rated for goodness, familiarity,
effort, skill and intention, but by different groups of raters. As our predic-
tions were of two types, either the works of art (Predictions 1 and 2) or the
raters (Prediction 3) were used as the unit of analysis.
Works of Art
Fifty-seven works of visual art were used as our first units of analysis. Most of
them belonged to the collection of Tate Britain, and were part of its perma-
nent exhibitions in the Millbank building, London.6 e works selected cov-
ered a span of more than four centuries (from 1575 to 2008); they were
paintings and drawings (n=39), sculptures and reliefs (n=11), as well as instal-
lations and mixed types (n=7). Both figurative and abstract works were
included; a wide range of genres (e.g., portrait, landscape, still life, history
painting) and styles or schools were represented (e.g., baroque, romanticism,
modernism, conceptual art). e artists represented, most of them British,
included famous ones, such as Bacon, Constable, Hirst, Millais and Turner,
but also lesser known ones by the wide audience, such as Des Granges, Dion,
Gennari, Jasinski and Vézelay. Seven works were from female artists. As it
seemed unreasonable to ask the raters to rate the total number of works
selected, these were divided into four sets (3 sets of 15 items and 1 set of
5 Possible differences in appreciation of works of art between art specialists and non-specialists
were experimentally investigated by several researchers. A classic study is from Winston and
Cupchick (1992), who showed that experienced viewers tend to prefer complex and challenging
works of art, a quite commonsensical finding, which is consonant with our prediction.
6 Entry to Tate Britain is free, except for some temporary exhibitions. All the works selected
were displayed in free admission rooms.
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 123
12 items); these sets were arranged by chronological order, following a three-
period scheme traditionally used by Tate Britain to organize its exhibitions;
2 sets of 15 items were in the Historic period, 1 set of 15 items was in the
Modern period, and 1 set of 12 items was in the Contemporary period. For
the list of works selected and their arrangement in sets, see Appendix A.
Raters
Five hundred and fifty-two Tate Britain visitors agreed to rate one of the four
sets of works. However, 10 questionnaires were returned blank and seven
were not returned. Furthermore, seven raters under 12 years of age were not
included in the analyses. e resulting sample size was 528. More than two-
thirds of the questionnaires were complete (all works rated and all additional
questions answered). At least 37 (up to 48) raters rated the totality of works
in each set and on each variable. Fifty-four percent of the raters were female.
e mean age of the raters was 40 years (range 12–83; SD=16). Forty-four
nationalities were represented, with a majority of British raters (48%), fol-
lowed by Americans (9%), Australians (7%), Germans (6%) and French
(5%); more than two-thirds of the raters were native English speakers. Finally,
almost one-third of the raters reported some special training or expertise in the
Arts, more than two-thirds of those being visual arts specialists.7
Measures
e works of art were rated for (i) Goodness (“How much do you personally
like this work of art?”, 7-point scale anchored at 1 (not at all), 4 (neither like
nor dislike), and 7 (very much), (ii) Familiarity (“How familiar are you with
this work of art?”, 4-point scale anchored at 1 (never seen it or a picture of it
before), 2 (have seen it or a picture of it once before), 3 (have seen it or a pic-
ture of it a few times before), and 4 (have seen it or a picture of it many times
before), (iii) Effort (“How much effort do you think went into making this
work of art?”, 7-point scale anchored at 1 (practically no effort), 4 (moderate
effort), and 7 (extreme effort), (iv) Skill (“How much skill do you think went
into making this work of art?”, 7-point scale anchored at 1 (practically no
skill), 4 (moderate skill), and 7 (extreme skill), (v) Intention I (“How easy is
it for you to understand what this work of art is about?”, 7-point scale
anchored at 1 (not at all easy), 4 (moderately easy), and 7 (very easy), (vi)
Intention II (“How successful do you think the artist was in conveying what
7 Valid statistics are reported. Missing values were: Sex: n=26; Age: n=30; Nationality: n=25;
Native tongue: n=29; Expertise in the Arts: n=26.
124 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
this work of art is about?”, 7-point scale anchored at 1 (not at all successful),
4 (moderately successful), and 7 (completely successful).8
To ensure independence of observations, a within-between design was
used: each of the four sets of works was rated on each of the seven variables,
but by different groups of raters. A first group rated Goodness and Familiar-
ity, a second group Effort and Skill, and a third group Intention I and Inten-
tion II. As we had four sets of works, the total number of groups of raters was
12. With such a design, represented in Table 1, it was not possible for the rat-
ers to guess that we were looking for possible correlations between our depen-
dent variable (Goodness) and our main independent variables of interest
(Effort and Skill and Intention I and II).
Procedure
As the design of the study required three groups of raters for each of the four
sets of works, at least twelve types of questionnaire were requested. Originally,
it was intended to counterbalance the order of presentation of all works, which
would have given 24 types of questionnaire. However, it was not possible to
counterbalance Sets 3 and 4, because these were part of “one-way” exhibitions
(participants could visit them in one direction only, and it seemed inappropri-
ate to provide them with a “reversed” questionnaire). Sets 1 and 2 were coun-
terbalanced and hence the final number of questionnaire types used was 18.
Each questionnaire contained an introductory sheet with instructions; the list
of works (with small black-and-white illustrations to help accurately identify
the works) arranged by rooms that the participants were asked to rate (the
questions were repeated for each item); and three final sheets with additional
information, such as age, sex, involvement with the arts, and so on (as requested
by Tate Britain, a couple of questions used in their satisfaction polls were
added). e questionnaires were anonymous.
e questionnaires were distributed at two times, over a total of seven days.
Except Wednesday, when Tate Britain hosted a special event, each day of the
8 e measures (as well as the protocol) were tested in a pilot study in situ. e questions used
for the actual survey were those that had been the most successful among the pilot raters. For
example, the pilot raters clearly preferred the question “How much do you personally like this
work of art?” over “Do you consider that this is a good or a poor work of art?” for the Goodness
measure. A few pilot raters reported “embarrassment” with assessing the effort and skill that went
into a work of art, and during the actual survey this was a general trend among visual arts
specialists. Other pilot raters reported “boredom” with the Intention measures, especially in the
sets belonging to the Historic period (“It’s always easy to understand!”), but no similar reactions
were observed during the actual survey.
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 125
week and each opening hour were represented. e distribution of the differ-
ent questionnaire types followed a randomized plan. e visitors were
approached by an experimenter (wearing an official Tate badge) at one of the
entrances of a given set of works, and asked whether they would agree to par-
ticipate in a survey on art. If agreed, it was briefly explained what was
expected, and the participants were provided with a questionnaire, a clip-
board, and a Tate corporate pencil that they could keep at the end of the sur-
vey. e participants took this material with them, and rated the works
during their visit. ey could take as much time as they wanted, and returned
the completed questionnaires at one of the gallery information desks.9
Results
Either the works of art (Predictions 1 and 2) or the raters (Prediction 3) were
used as the unit of analysis, and three batteries of statistical tests were used.
First, we looked for possible differences between the works of art and between
9 Permission to conduct the survey was obtained from Tate Britain, and the protocol was
approved by the Research Ethics Committee of the University of Oxford.
Table 1
Correlational design
Group of raters Set of works
1234
1 G, F – – –
2 E, S – – –
3 I, II – – –
4 – G, F – –
5 – E, S – –
6 – I, II – –
7 – – G, F –
8 – – E, S –
9 – – I, II –
10 – – – G, F
11 – – – E, S
12 – – – I, II
G=Goodness; F=Familiarity; E=Effort; S=Skill; I=Intention I; II=Intention II.
126 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
the raters. Second, correlations and partial correlations were run for the works
of art. ird, the data for the works of art were introduced in a multiple linear
regression model.
Works of Art
Ratings for our first unit of analysis, the works of art, were coded in the num-
bers 1 to 7 (variables Goodness, Effort, Skill, Intention I and Intention II) and
1 to 4 (variable Familiarity), high scores indicating high appraisals. In order to
carry out the tests, the scores were averaged by work of art (N=57). In each
variable, the data were roughly normally distributed, although we observed
high means, which we attributed to the quality of the Tate Britain collections.
e variables Effort and Skill, and Intention I and II, were averaged as global
measures; this was done because these two pairs of ratings had been designed
as such, and also because their means, as expected, turned out to be very
similar (Effort: M=5.25, SD=0.72; Skill: M=5.18, SD=0.77; Intention I:
M=4.95, SD=0.94; Intention II: M=5.02, SD=0.85) and highly correlated
(Effort and Skill: r=0.84; Intention I and Intention II: r=0.96). e resulting
means for Effort and Skill (collapsed), and Intention I and II (collapsed), were
5.21 (SD=0.71) and 4.99 (SD=0.89), respectively. Means for Goodness and
Familiarity were 4.61 (SD=0.60) and 1.50 (SD=0.36), respectively.
A t-test found no significant difference between the orders of presentation
of the works that had been counterbalanced in Sets 1 and 2; this is reassuring
as Sets 3 and 4 had not been counterbalanced (as explained above). We
also looked for possible differences between the works that belong to the His-
toric period (Sets 1 and 2, n=30) and those that belong to the Modern-
Contemporary period (Sets 3 and 4, n=27). e Historic period means (with
standard deviations in parentheses) for Goodness, Familiarity, Effort and
Skill, and Intention I and II were 4.77 (0.52), 1.50 (0.35), 5.58 (0.43), and
5.40 (0.73), respectively, while those of the Modern-Contemporary period
were 4.43 (0.64), 1.49 (0.38), 4.80 (0.75), and 4.52 (0.83), respectively. A
two-tailed t-test found these differences to be significant for Effort and Skill
and Intention I and II,10 p<0.001, 95% CIs of the mean difference [0.45,
1.10], and [0.46, 1.29], respectively.11
10 e difference between the Historic and the Modern-Contemporary periods for Goodness,
with Historic rated more positively, did not reach significance after the Bonferroni correction
(uncorrected, p=0.035, 95% CI of the mean difference [0.02, 0.65]).
11 All results reported as significant are significant after Bonferroni correction.
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 127
Raters
To carry out the tests for our second unit of analysis, the scores were averaged
by raters (n=528). Just as for the works of art, the variables Effort and Skill,
and Intention I and II, were averaged as global measures. e data were
approximately normally distributed.
For each variable, a series of t-tests found no significant differences between
the sexes, the nationalities and the native tongues of the raters, and a one-way
ANOVA found no significant difference between the age groups12 of the rat-
ers. To test our hypothesis that expertise influences art appreciation (Predic-
tion 3), three groups of raters were compared: those who reported no expertise
in the Arts (n=352, hereafter Non-Specialists), those who reported expertise in
the visual arts (n=104, hereafter Visual Arts Specialists), and those who
reported expertise in other arts (n=46, hereafter Other Arts Specialists). A one-
way ANOVA found a main effect of Expertise on Familiarity (F(2, 157)=7.718,
r2=0.089), Effort and Skill (F(2, 163)=8.224, R2=0.091), and Intention I and
II (F(2, 173)=10.068, r2=0.104), p<0.001. Planned contrasts were run. As
expected, the independent variables were rated higher by Visual Arts Special-
ists (Familiarity: M=1.80, SD=0.67; Effort and Skill: M=5.78, SD=0.87;
Intention I and II: M=5.33, SD=0.79) than by Non-Specialists (Familiarity:
M=1.41, SD=0.46; Effort and Skill: M=5.00, SD=0.90; Intention I and II:
M=4.72, SD=0.88). ese differences were significant for each variable (Famil-
iarity: difference=0.39, 95% CI 0.19 to 0.58, t(157)=3.92; Effort and Skill:
difference=0.78, 95% CI 0.39 to 1.15, t(163)=4.03; Intention I and II: differ-
ence=0.61, 95% CI 0.29 to 0.93, t(173)=3.77, all p<0.001). Visual Arts Spe-
cialists’ ratings did not significantly differ from Other Arts Specialists.13
Correlations
Significant positive correlations were found between the dependent variable,
Goodness, and the independent variables Familiarity (r=0.435), Effort and
Skill (r=0.559), and Intention I and II (r=0.533), two-tailed, p<0.01. Since a
significant difference had been found between the groups of raters, correla-
tions were also run separately for Non-Specialists and Specialists (this last
group including Visual Arts Specialists and Other Arts Specialists, who did
not significantly differ). As presented in Table 2, the correlations of the Non-
Specialists group remained strong (Intention I and II: r=0.514), or became
stronger (Familiarity: r=0.453; Effort and Skill: r=0.571), two-tailed, p<0.01,
12 e four groups created were: 12–25 (n=112), 26–40 (n=150), 41–55 (n=135) and 56–83
(n=101).
13 All results reported as significant are significant after Bonferroni correction.
128 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
whereas those of the Specialists group decreased, but remained significant
(Effort and Skill: r=0.444, p<0.01; Intention I and II: r=0.288, p<0.05), apart
from Familiarity (r=0.210, p=0.116).
To test that the significant correlations between Goodness and our main
dependent variables of interest, Effort and Skill, and Intention I and II, were
not due to the effect of Familiarity, partial correlations were conducted. As
presented in Table 3, when controlling for Familiarity, the correlations
between Goodness and Effort and Skill (r=0.534), and between Goodness
and Intention I and II (r=0.453) were still significant (two-tailed, p<0.01).
Partial correlations were also run separately for Non-Specialists and Special-
ists. Much like before, these decreased but remained significant (Non-
Specialists: Effort and Skill: r=0.510; Intention I and II: r=0.419, two-tailed,
p<0.01; Specialists: Effort and Skill: r=0.429, two-tailed, p <0.01; Intention I
and II: r=0.267, two-tailed, p<0.05).
Correlations were finally run for Goodness and the average of Effort and
Skill and Intention I and II. We regarded the effort and skill that went into a
work of art as an indicator and clarifier of intention. Since their correspon-
dent measures turned out to be similar (Effort and Skill: M=5.21, SD=0.71;
Intention I and II: M=4.99, SD=0.89), and significantly correlated in each
group of raters (see Tables 2 and 3), we used them as a global measure of
intentionality, hereafter Deliberateness (M=5.10, SD=0.71). Strong signifi-
cant positive correlations were found between Goodness and Deliberateness
in each group of raters (Non-Specialists and Specialists: r=0.614; Non-
Specialists: r=0.614; Specialists: r=0.420), as well as when controlling for
Familiarity (Non-Specialists and Specialists: r=0.534; Non-Specialists:
r=0.534; Specialists: r=0.400), all tests two-tailed, p<0.01 (see also Fig. 1).
Multiple Linear Regression
Familiarity and Deliberateness were finally introduced in the multiple linear
regression model. As shown in Table 4, Deliberateness was the stronger
predictor of Goodness (β=0.529, t(55)=4.92, p<0.001), followed by Familiar-
ity (β=0.268, t(55)=2.49, p<0.05), this model significantly accounting for a
moderate proportion of the variance (R2=0.441, p<0.001). Deliberateness
remained the stronger predictor of Goodness when multiple linear regression
was run separately for Non-Specialists and Specialists, Familiarity losing sig-
nificance as a partial predictor in the last group.
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 129
Table 2
Pearson correlations between the main variables of the study
Variable Raters
All Non-specialists Specialists
G F E–S I–II G F E–S I–II G F E–S I–II
G – 0.435** 0.559** 0.533** – 0.453** 0.571** 0.514** – 0.210 0.444** 0.288*
F 0.435** – 0.203 0.344** 0.453** – 0.305* 0.369** 0.210 – 0.132 0.138
E–S 0.559** 0.203 – 0.570** 0.571** 0.305* – 0.537** 0.444** 0.132 – 0.444**
I–II 0.533** 0.344** 0.570** – 0.514** 0.369** 0.537** – 0.288* 0.138 0.444** –
G=Goodness; F=Familiarity; E-S=Effort and Skill; I–II=Intention I and Intention II. n=57, *p<0.05, **p<0.01.
Table 3
Partial correlations between the main variables of the study
Variable Raters
All Non-specialists Specialists
GE
–SI–II G E–SI–II G E–SI–II
G – 0.534** 0.453** – 0.510** 0.419** – 0.429** 0.267*
E–S 0.534** – 0.544** 0.510** – 0.479** 0.429** – 0.433**
I–II 0.453** 0.544** – 0.419** 0.479** – 0.267* 0.433** –
G=Goodness; F=Familiarity; E-S=Effort and Skill; I–II=Intention I and Intention II. Control variable:
Familiarity. n=57. **p<0.01; *p<0.05.
Discussion
e results clearly support the three tested predictions. To start, Goodness –
an indicator of how much a work of art was liked – significantly correlated
with Effort and Skill and with Intention I and II, even among Specialists. As
the partial correlations revealed, these relationships cannot be reduced to the
influence of Familiarity, and it therefore seems that works of art that are
thought to have required a lot of effort and skill, and whose artist’s intention
is easy to understand, tend to be preferred (Predictions 1 and 2). Similarly, a
general measure of the accessibility of the artist’s intentions, Deliberateness,
which combined Effort and Skill and Intention I and II, was a stronger pre-
dictor of Goodness than was Familiarity, both for Non-Specialists and
Specialists.
130 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
Table 4
Regression analysis for deliberateness and familiarity predicting goodness
Predictor
variable Raters
All Non-specialists Specialists
BSEBβBSEBβBSEBβ
D 0.451 0.091 0.529*** 0.473 0.102 0.516*** 0.390 0.121 0.396**
F 0.443 0.177 0.268* 0.500 0.221 0.252* 0.195 0.164 0.148
D=Deliberateness; F=Familiarity. n=57. *p<0.05; **p<0.01; ***p<0.001. All r2=0.441, p<0.001; Non-
Specialists r2=0.431, p<0.001; Specialists r2=0.444, p<0.01.
6.5
6.5
5.5
5.5
4.5
4.5
Deliberateness
Godness
3.5
3.5
2.5
2.5
6
6
5
5
4
4
3
3
7
7
Figure 1. Linear correlation between Goodness and Deliberateness (all groups
of raters) (n=57, R2=0.377).
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 131
Regarding Prediction 3, a significant difference was found between Non-
Specialists and Specialists, the latter tending to rate Familiarity, Effort and
Skill and Intention I and II higher. is finding is consonant with the idea
that visual arts specialists, because of their training, are better equipped to
understand the artist’s intention and to determine how much effort and skill
went into a work of art. As a result, they probably get more cognitive contri-
bution from difficult works, and may like a broader range of art.
A last result has to be discussed here. e works from the Historic period
(Sets 1 and 2) were rated significantly higher than the works from the
Modern-Contemporary period (Sets 3 and 4), and this in all dependent vari-
ables (Goodness was not significant after a Bonferroni correction, but was
trending in the same direction). Initially, we had no hypotheses between
artistic periods: there is nothing that would “prevent” a work from the mod-
ern period from displaying a lot of effort and skill, from being easily under-
standable, and, therefore, from being appreciated as much as a work from the
historic period. We know, however, that many works of modern and contem-
porary art are abstract and conceptual. In other words, the artist’s intention is
generally more difficult to understand than in historic art, and if one takes
that into account, the difference between periods makes sense.14
We end this discussion by mentioning a limitation of the study. As we
have seen, Sperber and Wilson (1986/1995) defined relevance in terms of a
relationship between cognitive cost and cognitive impact. In our domain of
interest, this means that understanding the artist’s intention is not sufficient
to appreciate a work of art. For example, a work of art whose artist’s inten-
tion is easy to understand, but which has little or no cognitive impact, should
not be highly appreciated. In contrast, we predicted a linear relationship
between understanding the artist’s intention and appreciating a work of art.
We did so as the works of art selected at Tate Britain belong to carefully
vetted “high art”. In other words, these works of art have already surmounted
the triviality barrier, and they all have the potential to bring a significant cog-
nitive contribution in return of the cost required to process them. is is why
we predicted that, at an equal potential cognitive contribution, works of art
whose artist’s intention is easy to understand should be preferred over works
14 is difference may also be due to the place where the survey was conducted. Although
Tate Britain displays historic as well as modern and contemporary works of art, one might
argue that it attracts fewer visitors who like modern and contemporary art, especially in
comparison with Tate Modern, also located in London.
132 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
of art whose artist’s intention is difficult to understand, and this is what our
findings suggest.
We finally acknowledge that our findings may be consistent with several
interpretations. For example, one may argue that the visitors had already
reached an appreciation judgement when they rated Intention (as well as
Effort and Skill). In other words, they would have considered that some
works of art are easy to understand just because they liked them in the first
instance, as a post-hoc justification to reduce cognitive dissonance. is is
possible, although the design of the study was aimed at avoiding such post-
hoc constructions: Goodness and Intention were rated by different groups of
visitors, and it was therefore not possible for them to be aware that we were
looking for a relationship between these variables. Further, this interpretation
would not be able to account for the difference which we found between
periods. If some works of art were considered easy to understand just because
they were appreciated in the first instance, we could not explain why the
works of art from the Historic period tended to be preferred over those of the
Modern-Contemporary period.
Conclusion
Works of art are ambiguous artefacts. As objects they trigger expectations of
practical use or function, but at the same time frustrate these expectations. As
human-made products they are assumed to have a purpose, but this purpose
is not clear. As this is the case, works of art are probably assessed as inten-
tional acts of non-verbal symbolic communication. If this is true, Bloom’s
(1996) theory of artefact categorization and Sperber and Wilson’s (1986/
1995) Relevance eory of communication allow us to generate similar fruit-
ful predictions for the appreciation of works of art.
According to Bloom, understanding the maker’s intention matters in the
assessment of artefacts, especially when their form or function is ambiguous
(Bloom and Markson, 1998; Gelman and Bloom, 2000; Preissler and Bloom,
2008). As works of art typically are artefacts that lack practical functionality,
understanding the artist’s intention should play an important role in intuitive
art appreciation judgements. According to Sperber and Wilson, intentional
acts of communication bear a greater or lesser cognitive contribution at a
higher or lower processing cost, a relationship that constrains their relevance
and their success (Sperber and Wilson, 1986/1995). If works of art are intui-
tively represented as acts of communication, their appreciation should also be
constrained by the principle of cognitive relevance. For example, the artist’s
intention – assumed to be represented in the work of art and to communi-
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 133
cate something – should be perceived as more or less worthy of attention,
and understanding it should be more or less easy to achieve, both things influ-
encing the appreciation of the work of art.
e evidence presented above supports these predictions. First, we found a
relationship between how much the works of art were liked and how easy they
were to understand, which suggests that intuitive art judgements involve spec-
ulation about the artist’s intention. Second, we found that works of art whose
artist’s intentions were easy to understand tended to be rated more favourably,
which suggests that works of art are at least partly assessed in terms of cogni-
tive relevance.
e factors that influence art appreciation are not limited to understanding
the artist’s intention and certainly are numerous. Aesthetic appraisal, personal
preferences and cultural trends to name a few – as well as mere exposure, exper-
tise, and lower-level cognitive processes15 – all play a role in people’s judge-
ments of works of art. Moreover, intentionalism in art appreciation – or the
idea that considering the artist’s intention is necessary to understand and judge
a work of art – has been criticized by art theorists (e.g., Wimsatt and Beardsley,
1946/1999). We think however that understanding the artist’s intention is at
the core of art appreciation, as far as one does not misread what we mean by
this. For example, we do not pretend that all art is meaningful (what is the
“meaning” of a Rothko’s multiform painting?), and many contemporary art-
ists would deny that their works convey a precise “message” that would need
to be “understood”. All we say is that works of art, because they cannot be
approached in terms of practical use or function, will be automatically assessed
as acts of symbolic communication and, therefore, will trigger speculations
about the artist’s intention (“Why did he or she make that?”) – whether
speculating about the artist’s intention is justified or not, and whether the art-
ist actually intended to communicate something or not. We have seen that
such a constraint appears to be strong, and that taking it into account has the
potential to shed light on how people deal with works of art.
Acknowledgements
We are grateful to former Director Stephen Deuchar for having allowed us to
conduct the survey at Tate Britain, where we also thank Melanie Greenwood
and the staff for assistance. We thank Paul Bloom, Michael Kubovy and an
anonymous reviewer for their helpful comments on an earlier version.
15 On this last point, see Solso (1994).
134 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
References
Algoe, S. B. and Haidt, J. (2009). Witnessing excellence in action: e ‘other-praising’ emo-
tions of elevation, gratitude, and admiration. Journal of Positive Psychology 4, 105-127.
Bloom, P. (1996). Intention, history, and artifact concepts. Cognition 60, 1-29.
Bloom, P. and Markson, L. (1998). Intention and analogy in children’s naming of pictorial rep-
resentations. Psychological Science 9, 200–204.
Cutting, J. E. (2003). Gustave Caillebotte, French impressionism, and mere exposure. Psy-
chonomic Bulletin and Review 10, 319-343.
Gelman, S. A. and Bloom, P. (2000). Young children are sensitive to how an object was created
when deciding what to name it. Cognition 76, 91-103.
Gibson, J. J. (1979/1986). e ecological approach to visual perception. Lawrence Erlbaum, Hills-
dale, NJ.
Keil, F. C. (1989). Concepts, kinds, and cognitive development. MIT Press, Cambridge, MA.
Kelemen, D. (1999). Function, goals and intention: Children’s teleological reasoning about
objects. Trends in Cognitive Sciences 3, 461-468.
Kelemen, D. and Rosset, E. (2009). e human function compunction: Teleological explana-
tion in adults. Cognition 111, 138-143.
Kruger, J., Wirtz, D., Van Boven, L. and Altermatt, T. W. (2004). e effort heuristic. Journal of
Experimental Social Psychology 40, 91–98.
Leder, H., Carbon, C. and Ripsas, A. (2006). Entitling art: Influence of title information on
understanding and appreciation of paintings. Acta Psychologica 121, 176-198.
Levinson, J. (1979). Defining art historically. e British Journal of Aesthetics 19, 232-250.
——. (1993). Extending art historically. e Journal of Aesthetics and Art Criticism 51, 411-423.
Malt, B. C. and Sloman, S. A. (2007). Artifact categorization: e Good, the bad, and the ugly.
In E. Margolis and S. Laurence (Eds.), Creations of the mind: eories of artifacts and their
representation, pp. 85-123. Oxford University Press, Oxford.
Maslow, A. H. (1937). e influence of familiarization on preference. Journal of Experimental
Psychology 21, 162-180.
Preissler, M. A. and Bloom, P. (2008). Two-year-olds use artist intention to understand draw-
ings. Cognition 106, 512-518.
Rips, L. J. (1989). Similarity, typicality, and categorization. In Vosinadou, S. and Ortony, A.
(Eds.), Similarity and analogical reasoning (pp. 21-59). Cambridge University Press, New
York, NY.
Rochat, P. (1989). Object manipulation and exploration in 2- to 5-month-old infants. Develop-
mental Psychology 25, 871-884.
Solso, R. L. (1994). Cognition and the visual arts. MIT Press, Cambridge, MA.
Sperber, D. and Wilson, D. (1986/1995). Relevance: Communication and cognition. Blackwell,
Malden, MA.
Temme, J. (1984). Effect of mere exposure, cognitive set, and task expectations on aesthetic
appreciation. In Crozier, W. R. and Chapman, A. J. (Eds.), Cognitive processes in the percep-
tion of art (pp. 389-410). Elsevier Science, Amsterdam.
Wimsatt, W. K. and Beardsley, M. C. (1946/1999). e intentional fallacy. In Warburton, N.
(Ed.), Philosophy: Basic readings, pp. 344-356. Routledge, London.
Winston, A. S. and Cupchick, G. C. (1992). e evaluation of high art and popular art by
naive and experienced viewers. Visual Arts Research 18, 1-14.
Yus Ramos, F. (1998). A decade of relevance theory. Journal of Pragmatics 30, 305-345.
J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136 135
Appendix A: List of Works of Art Selected
Set 1 (Historic, 15 works):
Cornelius Johnson, 1. Apolonius Veth, 1644, Painting
Nicholas Hilliard 2. , Queen Elizabeth I, 1575, Painting
David Des Granges, 3. e Saltonstall Family, 1637, Painting
Benedetto Gennari, 4. Elizabeth Panton, 1689, Painting
Benjamin West, 5. Sir omas Beauchamp-Proctor, Bt, 1777, Painting
omas Gainsborough, 6. e Baillie Family, 1784, Painting
Giovanni Antonio Canal, 7. London: e Old Horse Guards, 1749, Painting
Samuel Scott, 8. An Arch of Westminster Bridge, 1750, Painting
Johan Zoffany, 9. David Garrick in ‘e Farmer’s Return’, 1762, Painting
Benjamin West, 10. Cleombrotus Ordered into Banishment by Leonidas II, 1768, Painting
Joshua Reynolds (Sir), 11. Self-Portrait as a Deaf Man, 1775, Painting
Joseph Wright (of Derby), 12. Vesuvius in Eruption, 1780, Painting
George Stubbs, 13. A Couple of Foxhounds, 1792, Painting
Richard Wilson, 14. Llyn-y-Cau, Cader Idris, 1774, Painting
omas Gainsborough, 15. Gypsy Encampment, Sunset, 1780, Painting
Set 2 (Historic, 15 works):
James Ward, 16. Gordale Scar, 1814, Painting
John Singleton Copley, 17. e Death of Major Peirson, 6 January 1781, 1783, Painting
Joseph Mallord William Turner, 18. St Benedetto, 1843, Painting
John Constable, 19. Sketch for ‘Hadleigh Castle’, 1829, Painting
John Constable, 20. Cloud Study, 1822, Painting
John Constable, 21. e Revd Dr James Andrew, 1818, Painting
John Everett Millais (Sir, Bt), 22. Ophelia, 1852, Painting
William Holman Hunt, 23. e Ship, 1875, Painting
Ford Madox Brown, 24. Jesus Washing Peter’s Feet, 1856, Painting
John Brett, 25. e British Channel Seen from the Dorsetshire Cliffs, 1871, Painting
Edward Coley Burne-Jones (Sir, Bt), 26. e Golden Stairs, 1880, Painting
George Frederic Watts, 27. Hope, 1886, Painting
Jozef Pankiewicz, 28. Swans in the Saxon Garden in Warsaw, 1894, Painting
Lawrence Alma-Tadema (Sir), 29. Portrait of Ignacy Jan Paderewski, 1891, Painting
Feliks Jasinski, 30. Paolo and Francesca, 1903, Etching
Set 3 (Modern, 15 works):
David Bomberg, 31. Tregor and Tregoff, Cornwall, 1947, Painting
Frank Auerbach, 32. Primrose Hill, 1968, Painting
Francis Bacon, 33. Study for a Portrait, 1952, Painting
Henry Moore (OM, CH), 34. Recumbent Figure, 1938, Sculpture
Paule Vézelay, 35. Lines in Space No. 3, 1936, Relief
Paul Nash, 36. Voyages of the Moon, 1937, Painting
Patrick Heron, 37. Horizontal Stripe Painting, 1958, Painting
Peter Lanyon, 38. Zennor Storm, 1958, Painting
Terry Frost (Sir), 39. Black and White Movement, 1952, Painting
Victor Pasmore, 40. Synthetic Construction (White and Black), 1966, Relief
Mary Martin, 41. Inversions, 1966, Relief
Stephen Gilbert, 42. Structure 14c, 1961, Sculpture
Henri Gaudier-Brzeska, 43. Bird Swallowing a Fish, 1914, Sculpture
136 J.-L. Jucker, J. L. Barrett / Journal of Cognition and Culture 11 (2011) 115–136
Constantin Brancusi, 44. Prometheus, 1912, Sculpture
David Jones, 45. Flora in Calix – Light, 1950, Graphite and watercolour on paper
Set 4 (Contemporary, 12 works):
Jeremy Deller, 46. e History of the World, 2004, Wall drawing/installation
Simon Patterson, 47. e Great Bear, 1992, On paper, print
Mark Dion, 48. Tate ames Dig, 1999, Installation
Simon Starling, 49. Work, Made-ready, 2001, Installation
Fiona Rae, 50. Untitled (yellow), 1990, Painting
Rebecca Warren, 51. In e Bois, 2005, Sculpture
Rebecca Warren, 52. Come, Helga, 2006, Sculpture
Gillian Carnegie, 53. Black Square, 2008, Painting
Gillian Carnegie, 54. rsXXII – /8-7, 2007, Painting
Damien Hirst, 55. Pharmacy, 1992, Installation
Damien Hirst, 56. e Acquired Inability to Escape, 1991, Sculpture
Jake and Dinos Chapman, 57. e Chapman Family Collection, 2002, Sculpture
is order corresponds to the order that was used in the questionnaires when not counterbalanced.
