I found Neil Ribe and Friedrich Steinle’s article “Exploratory Experimentation: Goethe, Land, and Color Theory” in the July 2002 issue of Physics Today (page 43) both stimulating and provoking. Like all good articles, it raised as many questions in my mind as it posed answers.Consider Isaac Newton’s “color circle,” shown in Ribe and Steinle’s figure 1. How did Newton ever get the idea of representing a continuum of colors on a circle? Surely painters had long known that mixing a bit of red with a bit of blue produces violet (or indigo). But what theoretical justification could Newton have posed for wrapping the diagram and joining it along a line (D in the figure) between red and violet? That step does not seem to be supported by his observations with light.His observations did suggest to Newton that color values appear on a continuum, rather than as shown in Johann Wolfgang von Goethe’s simplified picture. Newton’s empirical chart is not greatly different from the modern CIE chromaticity diagram, based on sensory stimuli, in which the greens and yellows occupy almost half the area.Newton’s color wheel suggests to me a kind of empiricism just as bold as Goethe’s. I agree with the authors’ characterization of “exploratory experimentation,” but I believe such investigation plays an even larger role in science than they suggest. All productive research is carried out within some sort of theoretical framework; the difference between the traditional model and the “exploratory” or “empirical” model is only in the degree to which experiments are guided by theory. Karl Popper’s idea that experiments should be wholly guided by theory is useful, but it does not embrace the notion of progress—except as advanced solely by theorists. Thomas Kuhn’s understanding of scientific revolutions, on the other hand, suggests that all revolutionary ideas are formulated outside the bounds of current theory. Many great discoveries, particularly in astronomy and cosmology, have come about through a stretching of the limits of theory.I can find no better example of the difference between the traditional model and the “exploratory” model than the article on magnetic confinement of fusion plasmas by Richard Hazeltine and Stewart Prager in the same issue of Physics Today (page 30). Perhaps someone long ago said, “Construct a magnetic bottle, and it will contain a magnetized plasma indefinitely,” but theory provides little guidance on exactly how the bottle should be constructed. Fantastic geometries such as those shown in the figures on pages 33 and 35 did not spring directly from theoretical reasoning. They took countless hours, over many decades, of tweaking the parameters, just as Newton, Goethe, and Edwin Land arrived at their ideas of the colors seen by the eye.© 2003 American Institute of Physics.