Information Visualization and the Next Generation Workspace

Pictures have gotten a bad rap for a long time. Arnheim points out that in the Hebrew tradition, the story of the long hostility toward images begins with the destruction of a piece of sculpture, the golden calf which Abraham burnt in the fire, ground to powder, strew across the water and made the children of Israel drink of. In modern times, many disciplines rely on pictures and diagrams. Indeed, much of modern math and science relies on the invention of the Cartesian plane. A typical example is provided by many economics textbooks which are littered with supply, demand, and utility curves. Much of the intuitive reasoning and explanation is expressed in terms of moving curves or intersection points around. Moreover, stories of the role of visual thinking underly many modern scientific discoveries, for example there are stories of Einstein’s riding of particles in his mind’s eye. Visual representations of complex phenomena can enhance or extend our ability to understand these phenomena. This insight has fueled the application of interactive 3d graphics to the visualization of phenomena studied by scientist of many stripes. As with the microscope and the telescope, and the stroboscopic camera and time-lapsed photography, the scientist is able to see phenomena of scales miniscule to galactic, and of durations instantaneous to glacial. Thus over the last few years scientific visualizationhas been applied to molecular modeling, atmospheric simulations, medical imaging, and the study of fluids and flows, surfaces and volumes. Yet in all of these cases, there is an essential connection to the underlying natural structure of universe. More recently, graphical techniques have been increasingly applied to problems of visualizing volumes of information of a more artificial or human-made nature: the data and documents of corporationsand countries, of stock markets and sports statistics, of goods and services, of computer file systems and the global information network. The work of Card, Mackinlay, and Robertson of Xerox PARC in particular has stimulated a new area of human-computer interaction called informationvisualization, which is about the building of interactive, visual representations of large amounts of abstract information. To be sure, the boundary between scientific and information visualization is not clearly demarcated—since for example, coloring of a terrain according to temperature seems hardly different from marking the points of a map according to the population or pollution levels—but rather a continuum in which the dimensions of the physical world gradually let go of the dimensions of the picture. During the last few years, researchers have designed a number of so-called ”information visualizations” which can be thought of as new kinds of sensemaking devices. They are what Don Norman calls ”cognitive artifacts” — artificial devices designed to maintain, display, or operate upon information. As such, a visualization is a represention not just a presentation. They aren’t just pretty pictures with which we persuade our managers or our funders. They are tools to extend our ability to think and make sense of the universe. Three aspects of visualizations are particularly important. They are, well of course, visual representations. The power of such representations partially lies in (as we often chant in our group at Xerox PARC) ”offloading the cognitive onto the perceptual.” If the equipment that exists between the surface of the eye and the depth of the mind can be shanghaied into the service of extracting sense or meaning from the arrangements of information in space, then more of our conscious effort can be applied to higher order processes of thinking about meaning and action. Second, recent work has focussed on increasing the amount of information that can be examined and manipulated without increasing the burden on the user. It is more and more common to encounter datasets and document collections in office work, in education and commerce, and even in the home that are unmanageable with the prevailing graphical user interface paradigm. The third important aspect of visualizations is that they are interactive, dynamic structures. The soul of computation has been breathed into the formerly inert forms of visual representations that have already played a major role in the history of science and thought. Lets consider each of these aspects in more detail.