Samu Mielonen
"I propose
solutions that nobody wants, to problems that don't exist."
- Alvin Lustig, architect, designer, educator and printer
One could argue that most of interface design is evolutionary in nature. Most current graphical user interfaces in computers are mere variations of the direct manipulation interface designed by Xerox researchers, and Alan Sutherland and Douglas Engelbart before them. The direct manipulation notion of noun-verb interactions, where the user first selects an object and then carries out an action on that object, is the same whatever system is used: Windows, MacOS, X-windows, BeOS, the browser metaphor or most multimedia. Only slight variations can be found in the visual or auditive representation of the basic idea of direct manipulation: Icons, Windows, Mice, sounds and Pointers (WIMP) may change in shape, size and color, but they all retain the same characteristics and roles.
After their introduction to the general public by the Apple Macintosh, the WIMP-interfaces have evolved gradually. User interface design has concentrated on creating understandable dialogue orders with windows, readable layouts with user interface objects or novel component interactions with new widgets. In UI design lingo this has extended to such "innovations" as non-modal dialog, tab-layout window, slider widget or tearable menu. Most of these innovations have been variations of the original Mac (or Star) interface. In effect, design in user interfaces for the personal computer has been very evolutionary for the past 15 years: no major breakthroughs have swept the landscape of general computing, although there have been a few contenders, like speech recognition which is now poised to make it's second debut (sic).
Rather than arguing that the activity of design is "only" evolutionary, one could argue that it is a process of gradual change interrupted by abrupt leaps due to fundamental shifts in the surrounding environment. In paleontology, this pattern has been called punctuated equilibrium: the static equilibrium of speciation is punctuated by sudden changes in the fossil record, changes that cannot be explained by gradual genetic evolution. It seems that the design of things often follows a similar pattern: novel break-through innovations are followed by longer periods of gradual improvements and variations on the initial theme.
If one is willing to accept this analogy between design and speciation, interesting questions arise. First of all, what kind of process leads to a revolutionary design concept? Not even the process of evolutionary design is too well understood and definitely not formalized, so understanding the revolutionary process would be even harder. Secondly, from a societal viewpoint, it is crucial to try to understand what kinds of environmental change allow revolutionary conceptions - or is it perhaps so that ground breaking concepts provoke changes in the environment?
Both Thomas Kuhn's "The Structure of Scientific Revolutions" and the punctuated equilibrium literature favor the idea of changes in landscape being the speciating cause, whether the landscape be conceptual or physical/genetic. Kuhn talks about anomalies in the explanatory power of paradigms, prompting the scientist to look for alternative explanations, some of which grow into new paradigms. The paleontologists talk about gene drifts that are caused by random mutations, but which stay apart from the homogenizing effect of genetic evolution due to either physical or somatic isolation. In effect, new species 'concepts' are born out of chance inside the existing species and then brought to surface by changes in the actual living environment.
Let us backtrack a bit. If we assume that punctuated equilibrium of species evolution is analogical to birth of new major concepts in design, UI design history should include evidence of abrupt changes in the conceptual landscape of the discipline. A brief glance at the history of human computer interaction reveals several such cases. The initial computers were used (or programmed) through rewiring the connections physically with jump wires. This cumbersome form of programming evolved slowly into punch cards, which originally simulated the placement of wires inside the computer. The next change in computer use was introduced by the command line interface with its verbal input, which became the dominant interface model for nearly two decades. This in turn was quickly overthrown by the graphical user interface that practically spawned the personal computer revolution. Now we've had 15 years of gradual improvement on the GUI model and no major UI breakthroughs.
While it can be argued that the process of design, when viewed from a longer historical perspective is in fact in punctuated equilibrium, we still don't know how and when the shifts from evolution to revolution occur. One could suggest, however, that revolutionary design happens under some specific conditions, while the diffusion of innovations is determined by sudden shifts in the equilibria of previous solutions and the new solutions, a kind of discontinuity.
What are these discontinuities, and could they be manufactured? An interesting question for any designer attempting to shape the world through innovation. If the discontinuities could be identified and profited from, or even better, manufactured, the project of revolutionary design would become much more rewarding. The discontinuities can be seen as societal changes in the patterns of behaviour of the actors for whom the design is targeted. While continuums are re-inforced by standards and organisational structures discontinuities are strengthened by the apparent failures of these standards (as in Kuhn's anomalies) or the superiority of new design solutions (as in traditional technology diffusion models).
I propose a third point of rupture that may break the lineage of a traditional design solutions- regardless of the superiority of or satisfaction with either the new or the prevailing solution. This rupture may be characterised as 'postmodern anxiety' or feeling of uncertainty about the conditions in a world without absolute truths. People change their preferred solutions, not because somebody is better (at least in the traditional technology or design sense), but because they are unsure whether the current solution really is the perfect one.
So, how does one go about making a revolutionary design? And if one succeeds, will the solution automatically be better than an evolutionary one? I don't think so. Design happens as punctuated equilibrium and both facets of the design process - both evolutionary and revolutionary - serve an important role in the defining and redefining of solutions. However, revolutionary design can be welcome in fields where the process of innovation has become stagnated and where all effort is driven towards further specialisation. A good example is the field of user interface design.
Revolutionary designs may be hard to achieve and the process of getting there may be unknown, but previous rvolutionary solutions offer some clues. Mental imagery techniques may be used to conjure up solutions that may initially seem out of place, but prove to be useful on closer inspection. It has been suggested that abductive reasoning, a mode of thinking very useful for coming up with novel ideas, is often visual in nature. Furthermore, one may use existing solutions or technologies in new contexts and combinations. This kind of innovation could be termed structural, since it usually reorganises the existing pieces of the puzzle. Finally, one may rely on pure chance to produce odd combinations and train oneself to separate the potential combinations from the uninteresting ones. These two abilities, producing surprising - even contradictory concepts and then recognising their potential, have been nominated as the two key factors in creativity by cognitive scientists.
Whatever the technique, the guiding principle could be to design for a person who does not exist and whose needs are not yet articulated by any existing solutions. This involves designing solutions that people do not want, to problems that do not exist, which is exactly what is needed. After all, design cannot take a break from the evolutionary cycle if it does not leave behind the previous solutions and the users who have been configured by those solutions. Therefore I suggest that any designer attempting to break from the mould should create their own user, with needs, problems and solutions that do not yet exist.