This series of seven interactive musical compositions is intended for serious play and wonderment about tools. Each piece is an abstract visual sound scoring language capable of playing itself back as the user gives input. The series is several attempted embodiments of my conceptual response to a recent HCI conclusion, as written in CWC 2004:
"Icon interpretation is inherently ambiguous because the relationship between icon and function is not determined by a set of well-defined syntactic and phonological rules."
Icon==Function is a science fiction, showing a hypothetical dystopia / utopia. What would an interface be like if its iconography were determined by a well-defined ruleset?
The software art was conceptualised by JT in a series of long-winded email discussions with David England and curator Michael Connor. The sketches were then interpreted by Josh, Matthew Phillips, and David Lu in a number of programming languages. The code is hosted by Sourceforge.
Each piece in the series is a simplified programming language used to
instruct the computer's functional output. A whimsical code editor is
exposed to the user as a visual toy that interacts with a mouse and
keyboard. No other labeling or indirect tagging is used to mark the
mini-programs written with each mini-language. The languages are made
of simple geometries - capable of representing themselves as their own
icon. The function's code can be seen as its own graphical icon. The
well-defined syntax of iconography in each representation is precisely
the instruction language controlling the sound output. When a user
browses a small set of thumbnails looking for a state to load and
play, the user is not seeing a textual name given to the work, nor a
proxy graphic that attempts to explain the function by some cultural
mapping. Instead, the user sees the code, itself. In the name
"Icon==Function," a double equality operator is taken from
modern text-based programming to communication that the icon and the
function are the same as one another. If a single equality sign were
used, it would signify undesirably that the function in these pieces
are somehow being assigned to the icon, and that the icon is a
transparent address for the function. The art series aims to collapse
the two ideas into one just for the sake of both conceptual argument
and exploratory scientific research.
Just Language Enough
In computer programming, the word "function" is used to
describe a collection of instructions that perform a coherent task. A
function hides the code inside of it from attention, and allows its
code to be invoked only by calling the function's name. In creating
the seven visual languages in this series, my biggest aim was to
produce the beginnings of language-ness from my interactive
geometries. The aim of adding the modern features of 'real'
programming code (such as memory and logical flow of control) was
somewhat irrelevant. In most cases, a simple linear list of
instructions sufficed in order to make the artistic point, and to
provide the software user with an experience that was close enough to
programming that they would gain some sense of authorship with their
constructions. A stronger focus was also placed on making the
experience playful in a game-like way, rather than providing the user
with an industrially competent solution.
Usability of the Pie are
unfamiliar components that interface designers hope to channel their
users to embrace through play or necessity. This is the part of the
experience that I am calling the exploration.
In the pieces of Icon==Function, the interfaces I am recycling come
from CAD and desktop graphics software applications like a paint
program or paste-up application. The colored square game is a slight
variation of a pixel editor. The triangle music interface is a
modified vector-line drawing tool. "Gear Sound" and
"Nested Circles" both rely on circle-drawing tools. Although
HexBall heavily resembles a video game, the most basic interface is a
pixel editor -- and so on. That is the intuition aspect. The
exploration aspect is what happens to those graphics creation
interfaces that result in something more than simply the circle that
was drawn, or the pixel that was toggled. No one originally expects
music to come from their rectangle dragging tool, nor would they
expect the element they drew to bounce around on the screen and obey
gravity after they drew it. However, this expectation should also
become intuitiho are already familiar with painting and graphics
applications. Even if draggable rectangles exist as part of widespread
commercial GUI OSes as a 2D spatial "multi-select" for file
icons, it is questionable what percentage of the user population
considers that an integral part of their computing lifestyles and
would recognize it in a new program, or know to check for it. My
mother (age 59, health care manager) has had considerable difficulty
getting into each piece as she must be told to actually drag the mouse
instead of just clicking it. After being reminded to drag, she spends
most of her time struggling with the touch pad on my laptop, and
eventually concludes that the piece is visually beautiful, without
delving deeply into trying new combinations. It is interesting for me
to see her interact with the pieces because she doesn't seem to
understand the language, nor is she searching for one. On the other
hand, my brother (musician in his early 20s) is better familiar not
only with graphics programs, but video games and software development
as well. He has witnessed my software interface experiments since the
mid-1990s and already knows that I expect him to explore, given little
or no instruction. My brother will start by doing "the sensor
dance." The sensor dance is what happens when a human is trying
all the different entry points into the interaction that he or she can
think of, in hopes to get any sort of response out of the unrecognized
interface. While an ambiguous physical installation will cause a
person to wave their arms in the air, move close and far, shout, clap,
and touch, a desktop computer based work (restricted to a mouse and
keyboard) evokes a different kind of sensor dance. The person will
click around on everything, dragging the mouse in different places and
pressing different keys. In my own observations, people have also been
known to re-launch the application to see if it does something new or
to check to see if it's broken. My brother is quick to engage in a neactive language, creating new instances and enjoying the outcomes
as his own achievements. Since he is a musician, he's familiar with
the idea of embracing a system and recombining its parts to produce a
new expression. One could say that my interfaces are
more intuitive for him. He is familiar on more levels than just
see-and-remember. I think these outer, more anthropological scopes of
understanding usability could make the problem clearer than simply
analyzing the language that translates between icon and function, or
icon-system and function-system.
Artist and coder: Josh Nimoy, coder: David Lu, coder: Matthew Phillips, Commission from FACT (www.fact.co.uk) and JMU HCI (www.hci-fun.org.uk) curated by Michael Connor & Marta Ruperez. HCI research headed by David England.
