ArticlePDF Available

Ernst H. Gombrich, Pictorial Representation, and Some Issues in Art Education

Authors:

Abstract

The Journal of Aesthetic Education 38.4 (2004) 32-45 This essay will deal with different ways of conceptualizing pictorial representation in art education and their implications. The philosophical issues involved in pictorial representation have fascinated philosophers since the time of Plato and Aristotle. In the first half of the twentieth century, the issue of representation became a centerpiece of various theories among art historians in an attempt to explain representational styles of non-Western cultures that were vastly different from Western styles, and also in order to reinterpret less-valued Western styles that had deviated from the Classical and Renaissance traditions. In recent years, the field of aesthetics has seen a resurgence of interest in the issue, probably because of the demise of Formalism and the returning interest in representation among postmodernists. In art education, interest in pictorial representation has been limited to research on children's drawing (mostly unsolicited drawing) and to the alleged benefits of drawing activity on children's emotional and cognitive development. To understand art images made by adult artists, especially in the Western tradition, art educators must refer to art history or aesthetics. For non-Western images, an anthropological or sociological approach seems appropriate because, more than Western images, they seem to demand knowledge about the culture in which they originated. Art education therefore does not have one comprehensive model of pictorial representation to encompass all these different kinds of images. For more than a decade, the dominant theory of children's pictorial representation in art education has been a symbolic model, and it thrived and is thriving in the context of ever-growing cognitive claims on art. Replacing the long-held creative self-expression model of Viktor Lowenfeld of the 1960s and 1970s, the advocates for this model advanced their theory based on Nelson Goodman's symbolic interpretation of pictorial representation. While their theory illuminated what had not been revealed by the previous creative self-expression model, its long-held dominance in the field lends it an aura of self-evidence and authority. My aim in this essay is to reintroduce Ernst H. Gombrich's theory of pictorial representation in order to balance the dominant theory and to provide a broad perspective for viewing children's drawings and adults' art images together. Despite its potential contribution to art education, Gombrich's theory has not been adequately interpreted for art educators. When Gombrich's theory was introduced in art education in the past, he was treated, along with Goodman, as a prime advocate of the concept of representation as a convention. His attention to cultural, psychological, biological, sociological, and art historical dimensions of pictorial representation has never been fully explored by art educators. Gombrich was first and foremost an art historian and theorist of art history of solid classical scholarship, but he was also a cultural historian equipped with extensive knowledge of diverse fields such as cognitive psychology, ethology, psychoanalysis, and the philosophy of science. He was truly a scholar with an interdisciplinary penchant, which makes it difficult for his readers to categorize him. In addition, his elusive writing style tends to limit access to his theory. In this sense, Leslie Cunliffe's recent reintroduction of Gombrich's theories to art education is notable. Cunliffe covered the whole gamut of Gombrich's theories, including that on representation, by contextualizing him as a Social-Constructivist, and he expounded many potential benefits that art educators may gain through understanding and adopting his theories. My scope will be narrower than Cunliffe's. I will deal with Gombrich's theory of representation insofar as it illuminates the nature of children's drawings and other art images and helps us understand them within a coherent system. The order of discussion will be as follows: first, the linguistic model derived from Goodman and its continuing influence on art education; second, Gombrich's theory of pictorial representation, including his discussions of the conventional and biological aspects of representation, and of the role of the function of the image in determining representational style; and finally, implications of Gombrich's theory for art education. For the last few decades, the...
Nanyoung Kim is Assistant Professor of Art Education at the School of Art and De-
sign at East Carolina University. Her recent publications include two articles in Vi-
sual Arts Research and The Journal of Gender Issues in Art and Education.
Journal of Aesthetic Education, Vol. 38, No. 4, Winter 2004
©2004 Board of Trustees of the University of Illinois
Ernst H. Gombrich, Pictorial Representation,
and Some Issues in Art Education
NANYOUNG KIM
Introduction
This essay will deal with different ways of conceptualizing pictorial repre-
sentation in art education and their implications. The philosophical issues
involved in pictorial representation have fascinated philosophers since the
time of Plato and Aristotle. In the first half of the twentieth century, the is-
sue of representation became a centerpiece of various theories among art
historians in an attempt to explain representational styles of non-Western
cultures that were vastly different from Western styles, and also in order to
reinterpret less-valued Western styles that had deviated from the Classical
and Renaissance traditions. In recent years, the field of aesthetics has seen a
resurgence of interest in the issue, probably because of the demise of For-
malism and the returning interest in representation among postmodernists.
In art education, interest in pictorial representation has been limited to
research on children’s drawing (mostly unsolicited drawing) and to the al-
leged benefits of drawing activity on children’s emotional and cognitive de-
velopment. To understand art images made by adult artists, especially in
the Western tradition, art educators must refer to art history or aesthetics.
For non-Western images, an anthropological or sociological approach seems
appropriate because, more than Western images, they seem to demand
knowledge about the culture in which they originated. Art education there-
fore does not have one comprehensive model of pictorial representation to
encompass all these different kinds of images.
For more than a decade, the dominant theory of children’s pictorial rep-
resentation in art education has been a symbolic model, and it thrived and
is thriving in the context of ever-growing cognitive claims on art. Replacing
the long-held creative self-expression model of Viktor Lowenfeld of the
1960s and 1970s, the advocates for this model advanced their theory based
on Nelson Goodman’s symbolic interpretation of pictorial representation.1
04.32-45_JAE.38.4 7/20/06, 4:25 PM32
Gombrich, Pictorial Representation, and Art Education 33
While their theory illuminated what had not been revealed by the previous
creative self-expression model, its long-held dominance in the field lends it
an aura of self-evidence and authority.
My aim in this essay is to reintroduce Ernst H. Gombrich’s theory of
pictorial representation in order to balance the dominant theory and to
provide a broad perspective for viewing children’s drawings and adults’
art images together.2 Despite its potential contribution to art education,
Gombrich’s theory has not been adequately interpreted for art educators.
When Gombrich’s theory was introduced in art education in the past, he
was treated, along with Goodman, as a prime advocate of the concept of
representation as a convention. His attention to cultural, psychological, bio-
logical, sociological, and art historical dimensions of pictorial representa-
tion has never been fully explored by art educators. Gombrich was first and
foremost an art historian and theorist of art history of solid classical scholar-
ship, but he was also a cultural historian equipped with extensive knowl-
edge of diverse fields such as cognitive psychology, ethology, psychoanalysis,
and the philosophy of science. He was truly a scholar with an interdiscipli-
nary penchant, which makes it difficult for his readers to categorize him. In
addition, his elusive writing style tends to limit access to his theory.
In this sense, Leslie Cunliffe’s recent reintroduction of Gombrich’s
theories to art education is notable.3 Cunliffe covered the whole gamut of
Gombrich’s theories, including that on representation, by contextualizing
him as a Social-Constructivist, and he expounded many potential benefits
that art educators may gain through understanding and adopting his theo-
ries. My scope will be narrower than Cunliffe’s. I will deal with Gombrich’s
theory of representation insofar as it illuminates the nature of children’s
drawings and other art images and helps us understand them within a co-
herent system. The order of discussion will be as follows: first, the linguistic
model derived from Goodman and its continuing influence on art educa-
tion; second, Gombrich’s theory of pictorial representation, including his
discussions of the conventional and biological aspects of representation,
and of the role of the function of the image in determining representational
style; and finally, implications of Gombrich’s theory for art education.
A Dominant Linguistic Model of Pictorial Representation in Art
Education
For the last few decades, the most influential model of pictorial representation
in art education has been Goodman’s linguistic model as espoused in his
classic work, Languages of Art.4 Goodman’s main thesis in the book is that
works of art are symbols and that representation and expression should be
understood in terms of semantic relationships such as reference or denotation.
Goodman was influenced by Gombrich’s Art and Illusion and by Con-
structivist psychologists.5 In Art and Illusion, Gombrich repudiated traditional
04.32-45_JAE.38.4 7/20/06, 4:25 PM33
34 Nanyoung Kim
imitation theory by emphasizing the extent to which representation and
perception are influenced by cognitive structure, and how strongly repre-
sentational conventions affect the artist. However, Goodman went further
than Gombrich by arguing that all representations are conventional sym-
bolic schema and that therefore even
pictures in perspective, like any others, have to be read; and the abil-
ity to read has to be acquired.…Realistic representation…depends
not upon imitation or illusion or information but upon inculcation.
Almost any picture may represent almost anything (LA, 14, 38).
For Goodman, pictorial representations are symbols: “a picture, to repre-
sent an object, must be a symbol for it, stand for it, refer to it… more par-
ticularly denote it” (LA, 5). Although it is beyond the scope of this essay to
deal with several decades’ debates in the field of aesthetics, Goodman’s ex-
treme relativist position elicited much criticism. However, it also gave great
impetus to the field of analytical aesthetics for reframing many issues
in the philosophy of art, whether by eliciting admiration or rebuttals.6
Goodman’s approach, which treated images as conventional symbols,
was not confined to pictorial representation but was extended to character-
ize art in general and other cultural products. Through these languages,
Goodman argued, we construct worlds, any one of which is no better than
any other, just different.7 Goodman’s influence on the vocabulary of art edu-
cation came through Project Zero, a research program he founded in 1967 at
Harvard’s Graduate School of Education.8 Its results have been dissemi-
nated mostly by its researchers, such as Howard Gardner, David Perkins,
and Jessica Davis. One paragraph in Davis’s work shows how much the
theoretical model of Project Zero is based on Goodman’s philosophical
stance:
Project Zero researchers considered the development of facilities in
the construction and communication of meaning through art as the
acquisition of a kind of literacy. Through internal symbols or repre-
sentations, the individual child or producer of art constructs a world
view. Through external symbols or representations, the individual
shares a world view.9
It is not difficult to see that there is a close relationship between Gardner’s
theory of “multiple intelligences,” which has enormous popularity in art
education and education in general, and Goodman’s position on arts as lan-
guage.10 In Gardner’s theory, traditional differentiations between cognition,
emotion, perception, and sensation tend to be left out in the discussion, and
the cognitive status of the arts is elevated as much as other intelligences.
Another notable source of Goodman’s influence in art education can be
found in writings by Marjorie and Brent Wilson.11 According to the Wilsons,
“children draw in order to symbolically explore their worlds.”12 Brent Wilson
emphasized the importance of spontaneous drawing because “spontaneous
04.32-45_JAE.38.4 7/20/06, 4:25 PM34
Gombrich, Pictorial Representation, and Art Education 35
drawing discloses a set of symbols through which the child might present
and experiment with personal and developing ideas about himself and his
world.”13 Drawing should have a place also in the school “because of its
contribution to students’ cognitive processes, to their competence and skill
in the use of a valuable symbol system.”14 One of the unique ways in
which drawing expands our knowledge of the world is that unlike words,
“pictures…have a somewhat isomorphic similarity to the everyday objects
of experience.…Moreover, these concrete presentations are filled with
information.”15
Goodman’s model of art and pictorial representation as a symbol system
seems to have been accepted by art educators because it fitted well with the
desire among art educators to free art from the previous expressionist model
of art education and bestow on it a new cognitive status.16 Discipline-based
Art Education (DBAE) has advocated the cognitive status of art for more
than two decades. Since David Ecker’s early argument, the concept of art
making as visual problem solving has taken root among art educators.17
Michael Parsons emphasizes that the appreciation of art is a cognitive act as
much as an emotional and perceptual event, and Ralph Smith argued that
analyzing artworks, studying art history, and clarifying art concepts can
legitimately belong to the humanities.18
The most serious problem in considering pictorial representation as a
symbol system is that in doing so one tends to neglect the differences between
pictorial representation and other symbol systems, such as written words,
musical notes, or mathematical equations.19 For this reason, Goodman in-
troduced the concept of a syntactically “dense,” “undifferentiated,” and
“replete” symbol system for pictorial description, and that of a syntactically
“disjoint” and “differentiated” symbol system for linguistic description (LA,
148–154, 225–232) However, this distinction addresses only the formal as-
pects of symbol, not the representational content that the symbol refers to,
nor their relationship. This system of thought leaves out questions about dif-
ferent styles of images, as it is irrelevant to ask why English is different from
French. Even more important, Goodman’s theory cannot address the psy-
chological and perceptual involvement of the makers and spectators of pic-
torial representation, and therefore the enormously powerful effect of images
in various human cultures as well as their functions in children’s drawing.
Gombrich’s Theory of Pictorial Representation20
The Primacy of Schema and Schematic Style
For about five hundred years until the middle of the nineteenth century the
prevalent theory of pictorial representation in Western culture was imita-
tion theory, which originated in the fourth century BC in Greece and was
04.32-45_JAE.38.4 7/20/06, 4:25 PM35
36 Nanyoung Kim
revived in the fifteenth-century Italian Renaissance. According to this
theory, realistic image is natural and primary because a picture is assumed
to be a copy of the visual appearance of reality. When art historians had to
explain the nonrealistic representational styles of some non-Western cul-
tures in the nineteenth century, their question was framed as, “Why has this
style deviated from Western realistic representation?” It was then easy to
ascribe the “deviation” to the supposedly underdeveloped mind or skill of
the non-Western people.
One of Gombrich’s great contributions as an art historian is that he
noticed the implications of the ubiquitous schematic styles in world cul-
tures and the power of tradition even in realistic representation in the West-
ern tradition, and he developed a general theory to explain both of them.21
The core of his theory is his conception of the human being as an active
agent in forming his or her experiences in the world. In the 1950s and 1960s,
when Gombrich advanced his theory, this approach had been taken by Ge-
stalt psychologists and by many ethologists, and indeed Gombrich referred
to their theories. During recent decades since the demise of Behaviorism,
this theoretical orientation has gained substantial support from cognitive
psychology.22
According to Gombrich, picture making does not start with the record-
ing of a visual impression on the retina, but with constructing a graphic
structure that shows our mind’s basic conceptual understanding of the
world. Gombrich called this structure “schema.” Even though Gombrich
rarely used children’s drawing as an example, the presence of schema is
most apparent in children’s drawing. Many scholars have studied the sche-
mata of children and untrained adults, although they have named it differ-
ently: Rudolf Arnheim calls it “visual concept”; Wilson, Al Hurwitz, and
Wilson call it “intrinsic graphic biases”; and Anna Kindler and Bernard
Darras call it “initial imagery.”23 Earlier schemata are more universal and
simpler, not because younger children have more power of abstraction, but
because their schemata reflect the early, less differentiated, stage of their
cognitive categorization.
Because the schema portrays our conceptual categorizations and the per-
ceptual constancies of the world by means of a crude graphic equivalent, it
functions like a symbol. At the turn of the twentieth century, the first inves-
tigators of children’s unsolicited drawing, such as James Sully and Earl
Barnes, already noticed the symbolic function of children’s image.24 In art
education literature the concept survived in the name of the “symbolic” or
“schematic” graphic developmental stage, even though for a long time, un-
der Lowenfeld’s influence, art educators were more interested in the al-
leged benefits of drawing as an activity rather than in the nature of draw-
ings as drawn marks. When the symbolic aspect of children’s drawings was
04.32-45_JAE.38.4 7/20/06, 4:25 PM36
Gombrich, Pictorial Representation, and Art Education 37
picked up again through the Wilsons’s application of Goodman’s symbol
theory, children’s drawing schemata became invested with cognitive power
through which children investigate and construct their world.
Like linguistic symbols, children’s schemata of things do not have to be
invented for each object by each child. They are recycled, imitated con-
stantly, and transmitted through generations of children in the context of
story telling. Wilson found that some features of children’s drawings are
culture-specific because of children’s constant copying behavior through gen-
erations.25 Normal adults’ drawing is basically the same as children’s draw-
ing, even though it might be more detailed and more differentiated. The
context in which the adults’ schemata are imitated and transmitted from
generation to generation is the society’s artistic tradition, where image mak-
ing usually takes place. The power and dominance of traditional schemata
on the individual artist will be stronger in this adult artistic tradition than
in the children’s social group, because in most traditional societies the form
of image or the procedure of image making is often invested with meaning
and values that resist change.
Realism is not exempt from the power of schema. In Art and Illusion
Gombrich explained, referring to many historical documents, how the West-
ern realistic image-making tradition was built upon utilizing conventional
schemata. What is exceptional about Greek (and Renaissance) naturalism is
that against that tenacity of schema, artists started correcting the schema by
comparing it against nature, and they continued the process like “a great
collective enterprise” in which each artist or development served as a step-
ping stone for the next artist or development. Why did Greek artists ever
begin this process? In answering this question, Gombrich differs from Plato.
Form Follows Function
According to Gombrich, the motive force underlying the growth of natural-
ism was not the wish to imitate natural appearance as such but rather a de-
sire to fulfill a specific function expected from the image by its beholders.
Gombrich argued that the function of Greek image was specifically to de-
pict mythical stories vividly, not just “what” happened but “how” it hap-
pened, as if those events were seen by an “eye witness.” Gombrich termed
this the “visual” or “dramatic evocation of a mythical event.”26 He assumed
that this unique Greek function of image was influenced by Greek drama,
which was based on Homer’s narratives, older than the onset of visual real-
ism. The Greek public’s expectation regarding visual representation was
prepared by drama, in which human situations and emotions are convinc-
ingly portrayed as if they were happening in a real situation to participants
with whom we can empathize. To bring about the viewer’s empathy in a
visual representation, the facial expressions and gestures of the figures in
04.32-45_JAE.38.4 7/20/06, 4:25 PM37
38 Nanyoung Kim
the picture and the space in which they move should convincingly convey
their emotions and the content of the story.
In other societies, pictorial representation performed different functions.
The most prevalent “conceptual style” is one that is well suited to picto-
graphic descriptions. Egyptian images were used diagrammatically, like a
written text or a map. When the West adopted Christianity, the earlier real-
ism was gradually discarded and the style became schematic because the
function of image changed; it was to teach Bible stories to the illiterate pub-
lic. When image comes to function more symbolically, like a statement (de-
scribing what happened, not showing how it happened), its legibility be-
comes more important than its evocative power. In that case a reduced
naturalism serves the function better: the form changes toward emphasizing
diagrammatic completeness, clarity, and distinctiveness of schematic fea-
tures. This is also the function of many images found in the advertisements,
diagrams, and pictographic road signs of modern culture.
In summary, Gombrich contended that there is a fundamental continuity
and similarity among different styles of representation. He did this by dem-
onstrating the existence of and need for schema in picture making and by
arguing that the function of image endowed by a given society is a major
cause of different styles.27 Schema reflects our cognitive structure in its
brevity and conceptuality, and reveals the societal nature of picture making
in its continuity and conventionality. The argument that schema is symbolic
and conventional, however, does not imply that the relationship between
the symbol and what it symbolizes is arbitrary. What most differentiates
Gombrich from Goodman is Gombrich’s emphasis on the psychological
effects of pictorial representation on the viewer.
Human Response to Pictorial Representation
While Gombrich regarded the schema as a common factor in all picture
making, he established another continuity in picture perceiving, namely the
human natural tendency to endow an image with a “presence” (AI, 113).
Gombrich used the word “illusion” for this perceptual tendency. Because
he applied the word widely to images ranging from the schematic to the
naturalistic, it engendered a host of misinterpretations.28
The status of pictorial representation as imitation, as in Plato’s theory for
example, is not compatible with how pictorial representations affect humans.
Gombrich postulated that image originated as a substitute for the original
in terms of its function. The common factor shared by the original and its
image is their similar perceptual and psychological effects on the viewer, or
similarities in terms of the viewer’s response. It is important to note that
this claim does not require the overall visual resemblance on the image’s
side that imitation theory assumes, and it goes deeper into the biological
04.32-45_JAE.38.4 7/20/06, 4:25 PM38
Gombrich, Pictorial Representation, and Art Education 39
realm, beyond the cultural dimension of conventionality. Gombrich’s
argument on this is well shown in his reflection on a hobbyhorse:
The “first” hobbyhorse…was probably no image at all. Just a stick
which qualified as a horse because one could ride on it. The common
factor was function rather than form.…Or more precisely, that formal
aspect which fulfilled the minimum requirement for the performance
of the function—for any “ridable” object could serve as a horse….[I]n
this sense, “substitutes” reach deep into biological functions that are
common to man and animal. The cat runs after the ball as if it were a
mouse. The baby sucks its thumb as if it were the breast. In a sense
the ball “represents” a mouse to the cat, the thumb a breast to the
baby. But here too “representation” does not depend on formal simi-
larities, beyond the minimum requirements of function.…An “im-
age” in this biological sense is not an imitation of an object’s external
form but an imitation of certain privileged or relevant aspects.29
In this sense, the relationship we have with an image is not passive:
there is a “beholders’ share” in this relationship that we bring to the percep-
tion of image. This psychological phenomenon is based again on the con-
structive nature of human perception. According to Gombrich, we actively
“project” interpretations into the visual stimuli, test them, and confirm or
reject them. Some such interpretation is innate, some gained through expe-
rience. According to a given interpretation we fill in what is not there, orga-
nize what is there, and neglect or pay attention to a certain part, and our
gaze hovers around the image to make sense of it in terms of our experience
of the real world.
When we are not aware of the process, we may even believe that the im-
age is somehow alive. Myths and stories about image making from around
the world almost always tell of how the pictorial representation came alive.
Gombrich cited the Greek myth about Pygmalion, who fell in love with the
sculpture he had made and by his prayer prompted Venus to bring the
statue to life. Some such stories also tell of how an image could exert the
same magical power as the original and had to be chained or left incom-
plete out of fear that it might come alive.30 David Freedberg amassed a
huge amount of evidence in Western history that testifies to this peculiar
effect of image on the beholder, which has been a great part of picture use
among the general population.31 Even for those of us living in a modern
industrialized society, where the magic of image seems to have disappeared
and visual art is not explicitly produced for that purpose, this propensity
does not disappear as advertisements and commercial images are incessantly
tapping this unique aspect of human experience with the image.
According to Gombrich, the degree of resemblance and the belief in
“Pygmalion’s power of image” tend to have an inverse correlation. When a
society believes in the magic of the image, the image need not be complete
or realistic because the belief fills the gap. Therefore,
04.32-45_JAE.38.4 7/20/06, 4:25 PM39
40 Nanyoung Kim
it is when art withdrew from the Pygmalion phase of action that it
had to cast around for means to strengthen the illusion and to create
the twilight realm of suspended disbelief which the Greeks first
explored (AI, 206).
In a similar vein, in Egyptian culture the pictographic style of Egyptian im-
ages does not necessarily hinder their function as substitutes for real things
for the deceased.
This view of pictorial representation can explain the typical representa-
tional styles of indigenous masks and sculptures. In the first half of the
twentieth century, tribal arts were admired by Modernist artists for their
“abstract” or “expressive” qualities. According to Gombrich, an “expres-
sive” quality is not the result of the expression of the artist, but the exag-
geration of relevant features, “relevant” to fulfilling the magical function of
the image in the given society. The so-called abstract quality derived from a
basic schema that did not have to be developed into realism in that culture.
Implications for Art Education
Even though Gombrich was not an art educator and did not deal with
children’s pictorial representation specifically, his theory can serve art edu-
cators in several ways (AI, 26).32 First of all, his theory provides a compre-
hensive model for understanding children’s and adults’ pictorial represen-
tation. The need for schema is a common factor in all picture making, and
the variety of representational styles derives from the various functions of
schema. Among world cultures, realism is rather an exception and nonreal-
istic rendering is ubiquitous because of the nature of picture making. The
concept of schema gives us an unbiased perspective on non-Western im-
ages: as much as Western images, they are the results of problem solving, of
purposeful, reasoned human activities carried out in order to fulfill certain
functions in the given society or context. In multicultural art education set-
tings we introduce to children many images from different cultures. It is
important then for art educators to understand that the logic of image mak-
ing depends on the image’s use, not on the skill or psychology of the maker.
The weird, scary face of an African mask was probably made that way in
order to scare away evil spirits; similarly, many of the guardian animals
and monsters keeping the gates of ancient palaces and temples around the
world do not at all appear to be kind.33
Second, Gombrich’s attention to the viewer’s response toward pictorial
representation reminds us how strongly images can affect our perception
and psyche. Without this understanding we cannot duly comprehend the
explosion of images appearing in modern visual culture and the powerful
lure exerted by certain images on the viewer. The fascination of fakes sub-
stituting for originals, often discussed in postmodern contexts, might be
04.32-45_JAE.38.4 7/20/06, 4:25 PM40
Gombrich, Pictorial Representation, and Art Education 41
grounded in this deeply rooted human relationship with the image.34 If we
ignore this primitive aspect of our response to pictorial representation, we
cannot fully understand the actual roles that pictorial representation has
played in history as well as in our own visual culture.
Finally, Gombrich’s theory sheds light on children’s spontaneous draw-
ing activity. In this light, we can see that the schematic style of children’s
drawing serves the function of their image making perfectly well, and in
that sense children’s drawings can be likened to certain images that adult
artists make, such as certain features of Egyptian images. Because concep-
tual clarity and differentiation are needed for conveying a story line, evoca-
tive realistic rendering is neither needed nor desirable, even though it is
also highly unachievable by children.
However, Gombrich’s emphasis on the human response toward the im-
age also reminds us that children’s drawings might be much more than a
matter of conceptual symbols to the children themselves; children are en-
gaged emotionally when they draw. Their drawings suggest much more to
children than linguistic symbols about the experiential quality of things
that they have experienced or would experience in imaginative worlds.
I suggest that children’s spontaneous drawing is best understood in the
context of make-believe. Indeed, many writers in art education put chil-
dren’s drawing in the context of make-believe.35 However, that has never
been a real focus in explaining children’s drawing, maybe for fear of giving
the impression that drawing is mere make-believe play. Kendall Walton criti-
cized the recent dominance of the linguistic model for explaining represen-
tational arts and put forward his thesis that representational artworks are
“props” for make-believe. “In order to understand paintings, plays, films,
and novels, we must look first at dolls, hobbyhorses, toy trucks, and teddy
bears.”36 As a prop and prompter, the pictorial representation gives the
imagination a substance. Walton’s thesis curiously reminds us of many ex-
amples, such as the hobbyhorse, that Gombrich gave to explain the relation-
ship between image and image maker. Although the image is meager and
ambiguous in regard to the statement function in which language excels, its
“arousal” function is incomparably superior to a sentence.
Try to say sentence [“The cat sits on the mat.”] to a child and then
show him the picture [of the cat sitting on the mat] and your respect
for the image will soon be restored. The sentence will leave the child
unmoved; the image may delight him almost as much as the real cat.
Exchange the picture for a toy cat and the child may be ready to hug
the toy and take it to bed. The toy cat arouses the same reactions as a
real cat—possibly even stronger ones, since it is more docile and
easier to cuddle.37
Defining children’s spontaneous drawing in terms of make-believe has
several ramifications. First and foremost, this will drastically reduce the
04.32-45_JAE.38.4 7/20/06, 4:25 PM41
42 Nanyoung Kim
importance of drawing activity for children’s development that has been as-
sumed by many art educators.38 I disagree with the Wilsons’ argument that
children spontaneously draw to “know” the world through the symbols
they make. A child who does not make drawings can find other ways to
make his or her imaginative world vivid and still develop a sound under-
standing of the world. What is typically cognitive about children’s sponta-
neous drawing does not derive from the fact that children use images as
visual symbols that refer to objects or give information, but from the fact
that their drawings, as substitutes or prompters, activate their memories
and imaginations vividly and freely. Therefore what is crucial for reaping
those benefits of drawing is not the drawing act itself, but all the other vari-
ables and conditions that make the drawing act possible and enjoyable: all
the direct or indirect experiences that children can draw from, all the re-
sources and references that prompt their imagination, and all the meaning-
ful interactions that children have with other children and teachers. If those
conditions are met, there must after all be emotional and cognitive benefits,
as Lowenfeld and the Wilsons have argued for.
Second, this perspective gives us a plausible explanation for why as chil-
dren grow older they do not draw as much as before, or simply stop draw-
ing39: it may be that as children grow older they do not have as much devel-
opmental need for make-believe or story-telling activities as they did before,
and that their categories become more differentiated and the simple images
therefore no longer excite their imagination. It is therefore reasonable to ar-
gue that as much as drawing is normal for young children, ceasing to draw
is perfectly normal for older children. In addition, images from mass media
compete with children’s own image making. Eventually they will start going
to movies and reading novels, as we all did.
Finally, this argument will make some art educators worry about the sta-
tus of art in school. The current situation is not favorable to art educators
even if there are cognitive claims on drawing or on art. It is my opinion that
the cognitive benefit of spontaneous drawing as framed by the Wilsons is
dubious, and that spontaneous drawing does not have to be related to a
school art program. Art can have rich benefits for children’s education and
well-being, including kinesthetic, perceptual, emotional, cognitive, aesthetic,
and even spiritual or socio-ethical benefits, but it depends on how we teach
what.
Besides the benefits to art educators, Gombrich’s theory can benefit any-
body who studies visual art seriously. The soundest aspect of Gombrich’s
theory is his integration of culture and biology in explaining art phenom-
ena. He paid attention to the enormous power of tradition, but warned us
against an extreme relativism or social determinism. His diverse expertise
in different disciplines not only gives explanatory power to his theory but
also provides us a role model to emulate as a researcher. One of his most
04.32-45_JAE.38.4 7/20/06, 4:25 PM42
Gombrich, Pictorial Representation, and Art Education 43
unique scholarly characteristics was the critical acuity with which he with-
stood the most popular art theory of his time, namely Expressionism, which
by now has lost most of its intellectual charm. In a word, Gombrich gives us a
broad and critical perspective that will not easily yield to the rise and demise
of intellectual fashion. 40
NOTES
1. Viktor Lowenfeld, Creative and Mental Growth, 3d ed. (New York: Macmillan,
1960) and Nelson Goodman, Languages of Art: An Approach to a Theory of Symbols
(London: Oxford University Press, 1966).
2. Ernst Hans Gombrich was born in Vienna in 1909 and died in London, Novem-
ber 2001, at age 92. For general information about Gombrich and his works, visit
the Gombrich Archive at http://www.gombrich.co.uk/
3. Leslie Cunliffe, “Gombrich on Art: A Social-Constructivist Interpretation of His
Work and its Relevance to Education,” Journal of Aesthetic Education 32, no. 4
(1998): 61–77.
4. Nelson Goodman, Languages of Art: An Approach to a Theory of Symbols (London:
Oxford University Press, 1968). This book will be cited as LA in the text for all
subsequent references.
5. Ernst H. Gombrich, Art and Illusion: A Study in the Psychology of Pictorial Repre-
sentation, 4th ed. (Princeton, N.J.: Princeton University Press, 1969). This book
will be cited as AI in the text for all subsequent references. It was first published
in 1960 based on his A.W. Mellon Lectures given in 1956 at the National Gallery
in Washington, DC. Goodman cited Gombrich in his Languages of Art, p. 7, 10,
12n., 16n., 33, 75n. Besides, Goodman cited works by R.L. Gregory, J.S. Bruner,
and R. Arnheim.
6. Jenefer Robinson, “Language of Art at the Turn of the Century” in “Symposium:
The Legacy of Nelson Goodman,” Journal of Aesthetics and Art Criticism 58, no. 3
(Summer 2000): 213–18.
7. Nelson Goodman, Ways of Worldmaking (Indianapolis: Hackett, 1978).
8. For the formation of Project Zero, see Goodman, Of Mind and Other Matters
(Cambridge: Harvard University Press, 1984). For a recent report on Project
Zero, see Howard Gardner, “Project Zero: Nelson Goodman’s Legacy in Arts
Education,” in “Symposium: The Legacy of Nelson Goodman,” 245–49.
9. Jessica Davis, “The ‘U’ and the Wheel of ‘C’: Development and Devaluation of
Graphic Symbolization and the Cognitive Approach at Harvard Project Zero,”
in Child Development in Art, ed. A. Kindler (Reston, Va.: NAEA, 1997), 45–58.
10. Howard. Gardner, Frames of Mind: The Theory of Multiple Intelligences (New York:
Basic Books, 1983).
11. Marjorie Wilson and Brent Wilson, Teaching Children to Draw (Englewood Cliffs,
N.J.: Prentice Hall, 1982); Brent Wilson, Al Hurwitz, and Marjorie Wilson, Teach-
ing Drawing from Art (Worchester, Mass.: Davis, 1987); Brent Wilson and
Marjorie Wilson, “Recycling Symbols: A Basic Cognitive Process in the Arts,” in
The Arts, Cognition, and Basic Skills, ed. Stanley S. Madeja (St. Louis: CEMREL,
1978), 89–109.
12. Marjorie Wilson and Brent Wilson, Teaching Children to Draw, 19.
13. Ibid., xv.
14. Brent Wilson, Al Hurwitz, and Marjorie Wilson, Teaching Drawing from Art, 10.
15. Ibid., 14.
16. Gardner specifically mentioned how much talks about arts are focused on
“emotion, spirit, mystery, the ineffable and unanalyzable.” See Gardner,
“Project Zero: Nelson Goodman’s Legacy in Arts Education,” 247.
04.32-45_JAE.38.4 7/20/06, 4:25 PM43
44 Nanyoung Kim
17. David W. Ecker, “The Artistic Process as Qualitative Problem Solving,” in Read-
ings in Art Education, ed. E.W. Eisner and D.W. Ecker (Waltham: Braisdell, 1963),
57–80.
18. Michael J. Parsons, How We Understand Art: A Cognitive Developmental Account
of Aesthetic Experience (Cambridge: Cambridge University Press, 1987) and
Albert W. Levi and Ralph A. Smith, Art Education: A Critical Necessity (Urbana:
University of Illinois Press, 1991).
19. Michael J. Parsons also saw that the paradigm of the cognitive status of art in art
education was related to Goodman and the researchers of Project Zero. How-
ever, his view is the opposite of this author’s. He interpreted their premises to
lie in the differences among symbol systems and therefore to be logically unfa-
vorable to the justification of integrated curriculum promoted in recent art
education. See Michael Parsons, “Integrated Curriculum and Our Paradigm of
Cognition in the Arts,” Studies in Art Education 39, no. 2 (1998): 103–16.
20. Gombrich’s theory of pictorial representation is mainly discussed in his Art and
Illusion; The Image and the Eye (Oxford: Phaidon Press, 1982); “Meditations on a
Hobby Horse or the Roots of Artistic Form,” in Meditations on a Hobby Horse and
Other Essays on the Theory of Art, 3d ed. (Oxford: Phaidon Press, 1978), 1–11; and
The Uses of Images (London: Phaidon Press, 1999).
21. I am not sure whether Gombrich would agree with this statement, because he
fought against “grand general theories” of his time such as those emphasizing
“spirit of age,” “class conflict,” or “cultural evolution,” as an agent of change. His
avoidance of systematization of his theory, however, is a substantial obstacle for
the reader in grasping his theory as a whole.
22. For example, see Cognitive Approaches to Human Perception, ed. S. Ballesteros
(Hillsdale, N.J.: Lawrence Erlbaum, 1994).
23. Rudolf Arnheim, Art and Visual Perception (Berkeley: California University
Press, 1974); Wilson, Hurwitz, and Wilson, Teaching Drawing from Art; and Anna
M. Kindler and Bernard Darras, “Map of Artistic Development,” in Child Devel-
opment in Art, ed. Anna Kindler (Reston, Va.: NAEA, 1997), 17–44.
24. Arthur D. Efland, A History of Art Education (New York and London: Teachers
College Press, 1990), p. 161.
25. Brent Wilson, “The Artistic Tower of Babel: Inextricable Links between Culture
and Graphic Development,” Visual Arts Research 21 (1985): 90–104.
26. Gombrich, “Paintings on Walls,” in The Uses of Images, 14–47.
27. In Gombrich’s theory, there are other factors that contribute to the differences in
representational style, such as characteristics of initial schema, values of a given
society metaphorically translated into formal characteristics, and social pres-
sure. See Gombrich, “Visual Metaphors of Value in Art,” in Meditations on Hobby
Horse, 12–29; and Gombrich, “The Logic of Vanity Fair,” in Ideals and Idols (Ox-
ford: Phaidon Press, 1979), 60–92.
28. See for example, Richard Wollheim, “Art and Illusion,” British Journal of Aesthet-
ics 3, no. 1 (1963): 26–27; and T. Wilkerson, “Art, Illusion, and Aspects,” British
Journal of Aesthetics 18, no. 1 (1978): 5–58.
29. Gombrich, “Meditations on a Hobby Horse,” 4–6
30. James Elkins, On Pictures and the Words that Fail Them (Cambridge: Cambridge
University Press, 1998).
31. David Freedberg, The Power of Images: Studies in the History and Theory of Response
(Chicago: The University of Chicago Press, 1989).
32. In terms of children’s drawing, Gombrich agrees with Arnheim’s view in
Arnheim, “On Growth,” in Art and Visual Perception.
33. Gombrich, “Sculpture for Outdoors,” in The Uses of Images, 136–61.
34. Kristine G. Congdon and Dug Blandy, “Approaching the Real and Fake: Living
Life in the Fifth World,” Studies in Art Education 42, no. 3 (2001): 266–78.
35. Wilson and Wilson, Teaching Drawing from Art and Clair Golomb, The Child’s Cre-
ation of a Pictorial World (Berkeley: University of California Press, 1992).
04.32-45_JAE.38.4 7/20/06, 4:25 PM44
Gombrich, Pictorial Representation, and Art Education 45
36. Kendall Walton, Mimesis as Make-Believe (Cambridge: Harvard University Press,
1990), 12.
37. Gombrich, “The Visual Image: Its Place in Communication,” in The Image and the
Eye, 139.
38. Phil Pearson has recently questioned the developmental necessity of drawing
for children. See Pearson, “Toward a Theory of Children’s Drawing as Social
Practice,” Studies in Art Education 42, no. 2 (Summer 2001): 348–65.
39. Ibid., 352. Pearson also said that many children do not produce drawing. I can
support this claim from my childhood experience.
40. Gombrich, “The Logic of Vanity Fair,” in Ideals and Idols, 60–92.
04.32-45_JAE.38.4 7/20/06, 4:25 PM45
ResearchGate has not been able to resolve any citations for this publication.