46 TEACHERs COLLEGE COLUmbIA UNIvERsITy
What We Illustrate When We Draw
defined precisely: Each thing conveys clear infor-
mation about what it is. These are known recogniz-
able objects, things we can all categorize easily and
efficiently. They share as many aspects as possible
with Eleanor Rosch’s basic level prototypes (Rosch
et. al., 1976)—a person will be drawn with two legs
even if only one can be seen, teddy bears have two
symmetrical rounded ears, no matter how battered
and asymmetrical they actually are, and hands have
five distinct fingers, no matter how folded up... The
differences of conceptually distinct objects are inten-
sified—say a cup and vase side by side of the same
color, in the same lighting will be drawn, and seen,
as lit differently. Boundaries between objects are
also drawn even if they are not visible. In Eye and
Brain, Richard Gregory writes, “the visual separa-
tions of objects are not given simply by borders of
light on the retinas. Separation into objects is given
by various rules, and by knowledge. Sharp borders
are rather rare, except for line drawings, which are
not typical.” (Gregory, 1997, p. 6) The categori-
cal perception first illustrated in voice onset stud-
ies, seems present in the way novice adult drawers
draw inside and between linguistically defined
shapes: if they draw a shadow, they outline it, dis-
tinguish it from its neighbors and fill it in evenly
dark throughout, something rarely apparent. The
teddy bear’s fur will be described by an unvarying
pattern of mark from edge to edge, though textures
appear and disappear with swells and in changes
of light: non-drawers draw things categorical, logi-
cal, not random. They have difficulty seeing visual
continuity between objects that they have separately
defined: So though there may be no visible change
in the shadow on the peach as it spreads over the
table, we see peach and table. It is a monumental
step in drawing to be able to see the peach-table
shadow as one thing. Novice drawers’ images are
cognitively economical, and priviledge categorical
information over visible. (The same tactics and line,
can also be seen in the work of experienced draw-
ers when they first sit down, or if they haven’t done
it in a while. And cartoon artists exploit an even
line, clarity and recognizability brilliantly—think
of Mickey Mouse, Charlie Brown, Spiderman—but
with an idiosyncratic taste for specific details, and
an overall control of composition not common
in beginner drawings.) A non-drawer comes to a
drawing class using vision the way they need to to
function in the world, and this is illustrated in their
drawings.Normative visual clarification
I think that the way that untrained drawers draw
probably reveals much more about mental repre-
sentations (however one wants to describe them), or
higher level visual processing, than their ability to
observe raw data. In most peoples’ drawings, a street
of buildings is drawn with perfect squares, rather
than sides that recede back in a squished shape; a
circular wine bottle mouth is drawn with a circle
rather than an ellipse. Often a beginner cannot see
the discrepancies in their own representations of
the external visible world—this was where I first
noticed a pattern. They judge that it’s “not good,”
or “not right,” and they get irritated with their own
drawing. I can point out the difference between
what they drew and what is in front of them, but at
first this is very, very hard to see, not quite believed,
dismissed, and then though they might see it in an
instance, for a while it does not transfer to other
similar instances. Perhaps it’s some kind of cog-
nitive dissonance. I guess
because they are drawing
what they perceive with their
minds’ eyes, it is difficult to
see what lies out there. Per-
haps this is an attentional
bottleneck, or like super-
imposed images on a film,
or the duck-rabbit, or other
“ambiguous” images, we can
only see one interpretation
at a time. Or perhaps the
magnificent capacity of the
brain that allows for object
and color constancy and
Figure 2.