| |
There
is No Such Thing as Information Design
One
of the ways we increase the returns we obtain from understanding the world
is by adapting our vocabulary to reflect our increasing knowledge. Humanity
never would have progressed beyond building the most primitive of mechanical
engines if careful thinkers had not learned to distinguish between energy,
force, work, and power. These words are still used loosely in everyday
speech, but professional mechanical designers and physicists use them
carefully and with well-defined meanings. Professionals in the field of
information-related design will have to learn to be equally careful. We
must not allow sloppy thinking to muddy the deep waters we find here at
the meeting of art, psychology, and electronic technology. Things are
difficult enough as it is.
I
am often embarrassed when supposedly technically savvy authors abuse terms
such as information, digital, and binary, seemingly
unaware of their precise meanings and implications. If our field is to
advance, we must, without displacing creativity and aesthetics, make sure
our terminology is clear.
As
a curmudgeon, I am delighted to point out that the popular term, Information
Design, is a misnomer. Information cannot be designed; what can be
designed are the modes of transfer and the representations of information.
This is inherent in the nature of information, and it is important for
designers to keep the concepts of information and meaning
distinct.
Information
is an abstraction from any meaning a message might have and from any particular
form a message might take. In the 1940s, the founder of information theory,
Claude Shannon, moved information from the realm of philosophers to that
of physicists, showing that the term could be given a clear definition.
Not only could we define it, but, he demonstrated, we could also quantify
it and treat information as a part of physics, something that I found
amazing and eye opening when I first read his works. One of the ideas
that Shannon established is that any information can be represented by
a sequence of the elementary particles of information, which we now call
bits. The unit of information is a two-way choice; we can model
a bit mentally as yes or no, green or red, on or off, zero or one, and
so on. Bits was coined by the mathematician John W. Tukey as a
contraction of BInary digiTS, that is, zeros and ones, the most
commonly used mental model of the smallest unit of information. (Shannon
and Weaver 1963: 9)
We
can always represent information as a sequence of bits, or, equivalently,
as a sequence of characters in a text or a string of numbers in base ten
(or any other base). These mechanical rearrangements do not change the
information in any way. There is nothing special, with regard to the content,
about the binary (base two) representation of information, though some
people treat numbers in base two as something almost mystical.
Similarly,
there is nothing special about the digital representation of information.
If I take some information and translate it into a non-digital (i.e. analog)
form, say as voltages on a conductor or the frequencies of a radio signal,
that information remains unchanged. We can convert the information back
and forth between analog and digital representations at will. In fact,
every time we use a modem to send email over a telephone line we convert
digital information into analog form; yet nothing happens to the information
from losing its digital cachet. The information is the same however it
is represented, and the identical message eventually appears on the recipient's
display with no evidence of its having been in analog form.
You
should therefore disregard any use of the term digital when that
term is used to imply that the digital nature of some information affects
its content or impact. The form of information storage or transmittal-whether
digital or analog, binary bits or decimal digits, or in some other guise-is
irrelevant to the issue of conveying meaning to people.
You
might ask: if any information can be represented as a finite sequence
of symbols, can we represent other aspects of the world, such as a 1-centimeter
cube of pure gold or an infinitely long non-repeating decimal like the
square root of two, as information? We can. I just did. That information
can take the form of a bunch of numbers-or even the form of words on a
page-does not limit the content the information might convey.
Even
though information is an abstraction that is independent of form and therefore
information cannot be designed, the way in which we represent information
to others is of crucial importance in communicating the meaning of the
information. The representation of the information is the plastic medium
with which we work. It would have been more appropriate to call this field
Designing Information Representation.
So,
how do we represent particular information so that it has a desired effect
on the recipient? That it is possible to define the issue (and the field)
in this way depends on the fact that we can represent the same information
in different ways, and that some representations are more effective than
others. In addition, the particular effect we want to achieve can vary
a great deal.
This
is where we-graphic designers, computer-interface specialists, artists,
musicians, sound technologists, lighting directors, cognitive psychologists,
type designers, ergonomicists, and even mathematicians and physicists-come
in. It is our job as designers to create effective representations of
information for human consumption.
Learning
how to represent information effectively requires us to travel along two
complementary paths. One is the apprenticeship route, along which we learn
from the example of current and past practitioners. The other path is
the theoretical route, since understanding some theory-and how to apply
it-can shortcut much trial and error. I use theory in the sense
that scientists usually do, to mean an established body of knowledge,
rather than in the everyday sense of a tentative idea.
Most
of us learn how to represent information by observing the work of experts
and working side by side with them. This the apprenticeship path. To leverage
further our designing efforts we can build on the foundations of information
theory and cognitive psychology that underlie the theoretical path. The
right place to start on the second path is with an examination of the
audience for our information; people.
Ergonomics
and Cognetics
In
spite of technological advances, people's access to external information
has not expanded beyond their optical, auditory, haptic, olfactory, vestibular,
and (that favorite of infants) gustatory senses. Therefore, the first
responsibility of the designer is to know the limitations and capabilities
of the human senses; a body of knowledge usually considered part of ergonomics
and human factors. A practitioner should know or have immediately to hand,
for example, the angular resolution of the human eye and how large lettering
must be for various color combinations, light levels, distances, and contrast
ratios. We need to know, to list some examples, how fast can the eye move
from point to point, how long does it take to find a target visually,
under what conditions can we hear and understand speech (e.g. in terms
of signal to noise ratio), what range of frequencies are easiest to hear,
which typefaces are more readable, how much or how little pressure can
we feel on our fingertips and elsewhere on our bodies, and how much force
a human hand can exert. In addition, we must know how all this data is
affected by human variables such as age and health.
At
times, common sense and common practice are sufficient guides to this
knowledge. But be careful: Time and again, hallowed practices have turned
out to be less than optimal, and designers ignorant of the teachings of
ergonomics have often failed to create products that work within the physical
and mental limitations of human users. Don Norman's delightful book, The
Design of Everyday Things, is full of examples of designer disasters.
As
a designer, you may be called upon to create a computer kiosk or a web
interface. To do so efficiently and effectively is difficult if you are
not familiar with the pertinent results from cognitive psychology, such
as the classic book The Psychology of Human–Computer Interaction
by Card, Moran, and Newell (1983). Hundreds of millions of computers have
unpleasant interfaces (Microsoft's Windows comes to mind) because they
require abilities humans simply do not possess, proving that their interface
designers were ignorant of fundamental facts about the human mind. The
sheer volume of jokes about computers reflects our near-universal dissatisfaction
with their interfaces. Of course, it is not only cognitive psychology
that is useful. When our information representations have emotional content,
it is useful to have studied what little is known about human emotions
from an engineering point of view. For example, there is no direct correlation
between color and emotional content; the meaning of color is culturally
mediated. Red is not universally a sign of danger; in some oriental cultures
it is white, not black that is associated with death; and so on. The naïve
designer who relies on his intuition (i.e. his cultural biases) in these
regards is likely to find his creations failing unexpectedly when they
have to work with a variety of users.
I
am not suggesting that designers should study the neurophysiology of the
brain. While such studies can lead to useful knowledge about how people
absorb and react to representations of information, how the brain functions
internally is not directly relevant to our field. We can treat the brain
as a black box and safely ignore all the twaddle about the parts of the
brain and what they do. Even where this physiological information is technically
correct (which is not always the case), it is of no direct use to us;
our only interest as information presenters in the human brain is how
it reacts wholistically, that is, how people respond or behave.
The
currently fashionable right brain–left brain distinctions are particularly
silly. Does it matter to us, as designers, that certain abilities are
located at such and such a location? Would it make any difference if the
right brain–left brain abilities were localized in the top and bottom
of the brain rather than the right and left halves? I think not. Useful
information takes an operational form.
Art or Science?
Every
major field of human endeavor is a mix of art and science. Theoretical
physicists and professional mathematicians speak of the aesthetics of
their work, and are driven by concerns about elegance and beauty. Is a
painter any less an artist for knowing perspective, understanding Josef
Albers elegant experiments and demonstrations about color, or being aware
of chemical incompatibilities between various kinds of paint?
Designing
the presentation of information, by the same token, partakes of the nature
of both art and science. Edward Tufte's books reflect such a blend of knowledge.
In one of them, he outlines his five principles for designing graphics (1983:
105):
- Above all else show the data
- Maximize the data-ink ratio
- Erase non-data ink
- Erase redundant data ink
- Revise and edit
The
first four are (mostly) science-based. However, the last, "revise and
edit" tells us not only to check repeatedly that the first four conditions
are met, but also to apply our aesthetic judgment to the final work. Describing
electronic displays in terms of numbers of pixels no more makes our work
technical than the fact that the frequencies of consecutive notes of the
tempered chromatic scale have a ratio of the twelfth root of two makes
musical composition a hard science.
Summary
Some
practitioners in the field illogically called "Information Design" try
to drape themselves with the mantle of technological modernity, recklessly
pinning the drapery together with abused jargon and misunderstood concepts.
The gaps that are left prove more revealing than they know. This does
not mean that their work as designers is bad, it just means that we cannot
learn much from what they say about their work.
I
predict that just as astronomy moved from the ancient tales (lovely, but
totally fictitious) of how the constellations were formed to the modern
physical models of the universe (breath-takingly beautiful, and even more
awe-inspiring), the most valuable improvements in the effectiveness of
representations of information will come from scientific analyses of human
performance. I do not imply any abandonment of art and the artistic impulse.
I used to spend hours tediously learning the formal motions used in conducting
the various musical meters so that those actions would become automatic;
now, when I conduct an ensemble, I do not spend any mental energy on technique,
but can focus entirely on communicating the music through the baton. For
the same reason, a designer who works with information should have a firm
grasp of human factors. Once mastered, that knowledge can be swept into
the unconscious, where it can guide you unobtrusively, leaving you free
to concentrate on the emotive and aesthetic sides of your work.
The
designer must, to use Wittgenstein's words, "throw away the ladder after
he has climbed up it" (1961:151). It has been the role of my essay to
point out that it can be dangerous to dispose of the ladder any sooner.
References
Card,
Stuart; Moran, Thomas; and Newell, Allen. The Psychology of Human-Computer
Interaction. Hillsdale NJ: Lawrence Erlbaum Associates, 1983
Norman,
Donald. The Design of Everyday Things. New York: Basic Books, 1988
Shannon,
Claude and Weaver, Warren. The Mathematical Theory of Communication.
Urbana IL: University of Illinois Press, 1963.
Tufte,
Edward. The Visual Display of Quantitative Information. Cheshire,
CT: Graphics Press, 1983
Wittgenstein,
Ludwig. Tractatus Logico-Philosophicus. London: Routledge & Kegan
Paul, 1961
Jef
Raskin
Biography
November 30, 2001
Email the editor
with any questions or comments.
 |
|
|
|
|
