Genius_ The Life and Science of Richard Feynman - James Gleick [220]
Consider a Mayan astronomer, he suggested. (In Mexico he had grown interested in the deciphering of the great ancient codices, hieroglyphic manuscripts that employed long tables of bars and dots to set down an intricate knowledge of the movements of sun, moon, and planets. Codes, mathematics, and astronomy—eventually he delivered a lecture at Caltech on deciphering Mayan hieroglyphics. Afterward, Murray Gell-Mann “countered,” Feynman said, with a series of six lectures on the languages of the world.) The Maya had a theory of astronomy that enabled them to explain their observations and to make predictions long into the future. It was a theory in the utilitarian modern spirit: a set of rules, quite mechanical, which when followed produced accurate results. Yet it seemed to lack a kind of understanding. “They counted a certain number and subtracted some numbers, and so on,” he said. “There was no discussion of what the moon was. There was no discussion even of the idea that it went around.”
Now a “young man” approaches the astronomer with a new idea. What if there are balls of rock out there, far away, moving under the influence of forces just like the forces that pull rocks to the ground? Perhaps it would make possible a different way of calculating the motions of the heavenly bodies. (Feynman certainly had memories of a young man confronting his elders with new, half-formed physical intuitions.)
“Yes,” says the astronomer, “and how accurately can you predict eclipses?” He says, “I haven’t developed the thing very far yet.” Then says the astronomer, “Well, we can calculate eclipses more accurately than you can with your model, so you must not pay any attention to your idea because obviously the mathematical scheme is better.”
The notion that alternative theories could account plausibly for the same observations had slipped into a central position in the working philosophy of scientists. Philosophers called it empirical equivalence, when they began to catch up. The recent history of quantum mechanics had pivoted on the empirical equivalence of Heisenberg’s and Schrödinger’s versions. The empirical equivalence of very different-seeming theories could be demonstrated mathematically, as Dyson had shown for Feynman’s and Schwinger’s quantum electrodynamics. Scientists knew, usually without thinking about it, that empirically equivalent theories could have different consequences, mathematics and logic notwithstanding.
For Feynman, especially, the tension between alternative theories served as a creative force, an engine for generating new knowledge. Perhaps more than any living physicist, he had made a specialty of learning what models could be derived from which principles, and what models from each other. To Dyson’s astonishment, he had stood at a blackboard one day in 1948 and interrupted their heady discussions of quantum electrodynamics to show him something different. Sketching quickly, he derived the nineteenth-century Maxwell field equations—the classical understanding of electricity and magnetism—backward from the new quantum mechanics. Einstein had started with the Maxwell equations and then shifted the perspective of the observer to arrive at his theory of relativity; Feynman went the other way in a fit of ahistorical perversity. He began with a void, no fields or waves, no concept of relativity, not even a notion of light itself, just a single particle obeying quantum mechanics’ odd rules. Before Dyson’s eyes he traveled back mathematically from the new physics, with its riddles of uncertainty and immeasurability, to the comforting exactitude