Richard Feynman (1918-1988) is the case that extends synesthesia from the arts into the physical sciences, and his description of his own experience is important because it was offered almost as an aside rather than as an essay. In What Do You Care What Other People Think? (1988), Feynman wrote: "When I see equations, I see the letters in colors — I don't know why. As I'm talking, I see vague pictures of Bessel functions from Jahnke and Emde's book, with light-tan j's, slightly violet-bluish n's, and dark brown x's flying around. And I wonder what the hell it must look like to the students." This is grapheme-color synesthesia operating during mathematical reasoning, and what Feynman was doing with it is the structural point worth pausing on.
A non-synesthetic physicist tracking multiple variables through a long derivation has to maintain those variables as abstract tokens in working memory, which is hard cognitive labor — the variables look alike, they can get confused with each other, the labor of keeping them straight is a substantial fraction of what doing mathematics feels like from inside. A synesthetic physicist tracking the same variables has each variable tagged with a distinct color, automatically, without any voluntary effort. The j is light-tan. The n is violet-blue. The x is dark brown. The tracking is not something Feynman had to do. The tracking was a byproduct of the perceptual system he was already using to read the equations. What the synesthesia gave him was a free organizational substrate for mathematical reasoning — a cognitive affordance that ordinary perception does not provide and that non-synesthetes have to manufacture the hard way. It is not the only thing that made Feynman one of the most creative physicists of the twentieth century. It is, however, one of the things, and it shows up in the specific visual-spatial character of his famous diagrams.
Feynman diagrams are the pictorial notation Feynman invented in the late 1940s for representing particle interactions in quantum electrodynamics. The diagrams let physicists reason about particle interactions as geometric objects — lines, vertices, intersections — rather than as the algebraic machinery of perturbation theory expressed in equations alone. The diagrammatic method was so powerful that it was adopted almost immediately by the physics community, and every physics textbook since the 1950s that treats QED treats it at least partly through Feynman diagrams. The cognitive architecture that saw the letters of equations in color was the same architecture that saw particle interactions as geometric shapes in space. In both cases, the discovery specialty was externalizing an internal visual-spatial perception into a form the rest of the physics community could use. The integration community — Feynman's colleagues at Cornell and Caltech, the physics journals that published his work, the graduate students who learned to draw the diagrams and then taught them to their own students — turned Feynman's idiosyncratic internal perception into the standard visual vocabulary of modern particle physics. Feynman had the perception. The community inherited the notation.