Simplicity and Prediction
Concerns about piling on too many elements into models, which can easily arise when representing the cosmos, also arise in the world of neuroscience. Gully Burns is project leader in biomedical knowledge engineering with the University of Southern California’s Information Sciences Institute. He has spent a lot of time thinking about the organization and presentation of neuroscience data. Burns has seen visualizations that might be spectacularly ornate and arresting, but that very quality can short-circuit their main purpose and value. “Sometimes people tend to revel and get lost in the complexity,” he points out. Burns recalls a paper describing cerebral function that was formative in his own career. The authors were showing connections in the brain that were beautiful and complicated. Burns thinks what was lost in the image was a larger point about a hierarchical pattern; it was obscured by the complexity. The value of simplicity in design is certainly not a new idea. However, with the increasing use of interactive visualized models and simulations, being able to look past complexity can make it possible for underlying patterns in the data to emerge and give new kinds of displays more predictive power.