Where's Transparency With So Many Layers of Abstraction?

Some older scientists, for example, justify their use of opaque software by pointing to the infinite regress of computer representations. After all, they argue, it doesn’t really mean much to know how your simulation is programmed if all you are looking at is a high- level computer language. The “real guts” of the program is in assembly language and in all that lies beneath that, and no one wants to go to that level with today’s complex machines. In the 1980s, Professor Barry Nilo= insisted that his students learn the physics of display technology; today, such scruples seem of a di=erent era, practical impossibilities that lead scientists, young and old, to accept opacity. These days, the problem for the working scientist boils down to a question: What level and language will provide enough understanding for me to compare the simulation before me with what I know of nature.