I
M
A
G
E
B
Y
F
X
Q
U
A
D
R
O
used by other embodied people.
This design-and-build activity is
profound. It was not long ago in hu-
man history that giving form to the
formless was considered the purview of
the divine. In fact, the English verb “to
create” comes from the Latin “creare,”
which means to bring “form out of
nothing.” When we design and build
systems, we bring form out of nothing.
Unfortunately, unlike the divine,
we cannot anticipate all the ways our
designs will affect the people who en-
counter them. And when a mismatch
arises, the world can become a very rig-
idly embodied place (see Figure 1).
Many of the great breakthroughs
in interactive computing have come
as improved embodiments capable of
transforming the way people experi-
ence the digital world. Sutherland’s
interactive display and light pen in
SketchPad, 31 Engelbart’s and English’s
mouse in NLS, 4 and Apple’s iPhone
all represent breakthrough embodi-
ments. But a vital engineering insight
is that they, as with all interactive
technologies, include certain “ability
assumptions” that must be met by hu-
man users. These assumptions are of-
ten unstated but alienating if they can-
not be met.
An everyday example makes the
point. In the student union building
at the University of Washington in
Seattle, wall-mounted touchscreens
function as information kiosks for
visitors (see Figure 2). In the on-screen
operating instructions, a particular
word stands out—“just,” as in, “just
touch the screen.” In fact, touching
the screen requires many abilities, in-
cluding closing one’s hand, extending
one’s index finger, elevating one’s arm,
seeing the target, landing accurately,
holding steady, and lifting without
sliding—along with the ability to read
and understand the instructions in the
first place. There is clearly no “just”
about it.
Where do ability assumptions come
from? Designers and developers make
assumptions from their own abilities,
from the ones they imagine other peo-
ple have, or the ones of the supposed
“average user.” 22 Unfortunately, each
source of such assumptions is flawed.
The first two are prone to bias and un-
representative; the third, insidious for
its statistical façade, does not reflect
the diversity of human life.

On that point, Rose25 offered an anecdote from the U.S. Air Force. After World War II, it frequently lost pilots and planes in peacetime crashes— incredibly, 17 on one particular day—so it decided to redesign its cockpits to reduce “pilot error.” Air Force engineers measured 4,063 pilots along 140 dimensions, averaging these values to create cockpits to fit the mathematically average pilot. But a young Air Force scien-