THINKING THROUGH DRAWING: PRACTICE INTO KNOWLEDGE 61
A Multi-Stage Attention Hypothesis of Drawing Ability
Justin Ostrofsky
Brooklyn College of the City University of New York
Aaron Kozbelt
Brooklyn College and
The Graduate Center of the City University of New York
Introduction
The behavior of observational drawing is a spe-
cific type of artistic production where a person
attempts to render an external visual stimulus with
the aim of achieving visual accuracy. Cohen and
Bennett (1997) defined a visually accurate drawing
as “one that can be recognized as a particular object
at a particular time and in a particular space, ren-
dered with little addition of visual detail that can-
not be seen in the object represented or with little
deletion of visual detail” (p. 609). One of the most
fascinating aspects of this behavior involves the
vast individual differences in the ability to produce
visually accurate drawings (Cohen, 2005; Cohen &
Jones, 2008; Cohen & Earls, 2010; Kozbelt, 2001;
Ostrofsky, Kozbelt & Seidel, in press). When asked
to draw a visible object or scene, most adults expe-
rience tremendous difficulty. The ability to produce
accurate drawings normally comes only after great
expenditure of effort through years of intensive
training and practice.
Of central concern to this paper is the problem
of understanding realistic drawing ability. How can
we explain why skilled artists are capable of pro-
ducing highly accurate drawings while unskilled
non-artists are highly prone to making substantial
drawing errors? What underlying skills and strate-
gies lead to the ability to create more accurate draw-
ings? Since the process of observational drawing
starts with acquiring visual information from the
model being drawn, psychologists have primar-
ily focused on perceptual processing in explaining
individual differences in drawing accuracy (e.g.,
Cohen & Bennett, 1997; Cohen & Jones, 2008;
Kozbelt, 2001; Kozbelt & Seeley, 2007). However,
the exact perceptual processes that are related to
drawing ability are not currently well understood,
and researchers’ continued focus on early stages of
visual processing may well lead to biased or invalid
conclusions.
In contrast, this paper outlines a multi-stage
attention hypothesis, a tentative understanding of
the perceptual processes that are related to indi-
vidual differences in drawing ability. Briefly, this
hypothesis argues that drawing ability is influ-
enced by two factors: (1) what visual information
is selected versus not selected to be included in the
drawing; (2) the degree to which the visual system
enhances the processing of selected visual informa-
tion and suppresses the processing of non-selected
visual information during the activity of drawing.
Some empirical evidence in support of this
hypothesis stems from an experiment that was
recently conducted in our lab (Ostrofsky, Kozbelt,
& Seidel, in press). The experiment examined 15
trained artists specializing in observational draw-
ing and 33 non-artists who reported no training or
extensive experience in observational drawing. We
assessed participants’ observational drawing ability
by asking them to make a free-hand copy of a pho-
tograph of an octopus using pencil and paper. The
participants were given 15 minutes to complete their
drawing and were instructed to draw the octopus
as accurately as possible using whatever techniques
they felt comfortable with. To quantify the accuracy
of these drawings, three trained artists indepen-