THINKING THROUGH DRAWING: PRACTICE INTO KNOWLEDGE

THINKING THROUGH DRAWING: PRACTICE INTO KNOWLEDGE 81

michelle Fava

Here the artist has evaluated the accuracy of her
measurements, and is making a decision to change
a portion of the drawing. She mentions many sepa-
rate features, and spatial relationships, so there is a
complex interaction between perception and deci-
sion making. Figure 1 shows an example of how the
double coding was employed.
This solution allowed each transcript to be
coded twice. The two coding schemes were separate
but related, suggesting two distinct but interrelated
types of cognition. These are (provisionally) labelled
“strategic thought” and “visual attention.” These cat-
egories are described in more detail in the following
section, in which the codes and categories are con-
sidered in relation to time.

A working model of drawing and cognition
We can consider the example of transcript from
the section above in relation to time. This is illus-
trated in Figure 2. It is clear that strategic state-
ments relate to the immediate past or future, while
visual attention (while often employed in strategy)
takes place momentarily. This dual temporality is
represented by the two axes of the model.
The model below in Figure 3 is constructed in
the same way, using the commonest types of state-
ments from the dataset of transcripts from all five
artists. It does not include all the categories from the
coding scheme, as these would be too numerous for
clarity, particularly in the visual attentive dimension.

The horizontal arrow represents the time spent

making the drawing. Statements reflecting meta-

cognition are more general, or refer to moments

in time outside of that time-span, for example they

might judge the difficulty of the task, the progres-

sion of the artists own skill, or compare the draw-

ing with one made previously. For this reason, they

are not directly linked to the timeline. It is unclear

whether they should still be considered part of the

horizontal axis.

Using the model to describe expert behaviours

While the incompleteness of verbal reports as

data mean the model cannot be used to exhaustively

describe, or quantify the frequency of, thought pro-

cesses, it is still useful in a number of ways. It pro-

vides a structure within which to describe drawing

strategies, and to consider other theories of drawing

and cognition. It allows us to consider the afore-

mentioned dichotomy as two domains in which

learning happens (in addition to other domains, e.g.

schematic knowledge, psychomotor, and affective).

Two expert behaviours were identified. These

are skills relating to meta-cognitive control in these

two domains: evaluative strategy, including periods

of non-judgemental cognitive activity; and control

of visual attention (or “attentional strategy” to use

Kozbelt’s (2010) term). I will briefly describe these

two skills in relation to the case studies.

The existence of periods of non-judgemental

Statement Strategic thought Visual attention

Ok. So that’s the sort of

compositional size that I want the

head to fill.

Evaluation

(against subgoal)

Compositional size

That’s the sort of space I’m gonna

use on the page.

Decision Overall composition

Ok. The problem there is I’ve just

gone too long.

Evaluation (of part) Global view

Relative length

So I just need to reassess Decision

the shape of the nose, the chin and

the mouth in relation to the eyes,

Subgoal Shape

Feature

Configuration of points

so it’s that sort of triangle bit

(gestures) in the centre of the head.

Input only Constructed shape

Global view

Figure 1. Example of coded transcript with two categories^3 (AR)