Detail from Jan van Eyck’s Arnolfini Marriage, with chandelier. Painting: Erich Lessing/Art Resource, NY

In the late 1990s, British artist David Hockney put forth a controversial theory of how realism seeped into the art of certain Renaissance masters. He asserted that the natural-looking faces and near-photographic precision in 15th- and early 16th-century paintings by Jan van Eyck, Lorenzo Lotto, and others resulted from the secret use of camera-like devices. The artists, said Hockney, pointed concave mirrors or convex lenses at subjects positioned in sunlight in order to project full-color, upside-down images onto canvases that they’d placed in darkness. More than 400 years before film, Hockney posited, Renaissance painters were tracing out the precursors of photographs.

Hockney’s theory caught the attention of Charles Falco, a professor of optical sciences and condensed-matter physicist at the University of Arizona. In 2000, the physicist approached the artist, suggesting that he help Hockney prove his hypothesis using mathematical techniques. The results of their collaboration have appeared in Hockney’s 2001 book, Secret Knowledge: Rediscovering the Lost Techniques of Old Masters, and in Falco’s published papers, technical analyses of suspect proportion and perspective in selected paintings, using the science of geometric optics. (Hockney argues that artists, including Caravaggio and Ingres, continued to employ optical aids until photography grew readily available in the mid-1800s; Falco concentrates his analyses on the painters of the Renaissance.) On a late Friday afternoon in November, Falco attracted a crowd of nearly 70 students and professors to a lecture in Fulton Hall; latecomers stood two rows deep in the back aisles.

Falco had been invited to campus by Ellen Winner, a professor of psychology who has written on the rift between art historians and scientists caused by Hockney’s theory. While scientists and studio artists have generally found Hockney’s idea compelling (scientists “tend to see merit in the quantitative evidence,” Falco said after his talk, and artists “tend to grasp the utility of timesaving tools”), many art historians have been troubled by the dearth of documentary evidence supporting the projection thesis; if optical aids were used, they say, artists would have revealed this in their writings or paintings. Some in the field dispute the suggestion that highly skilled masters needed such devices to produce their work.

A tall, mild man with a salt-and-pepper mustache, Falco emphasized in his talk that neither he nor Hockney was accusing the artists of cheating; nor was he trying to downplay their art. Lenses, he said, “can’t draw a line.” “These were tools, just like paintbrushes,” he added, and artists used them for some aspects of some paintings, not to project a whole scene. For the next hour, he presented his evidence. “The proof,” he said, “is in the paintings.”

On a screen behind Falco, a slide of Lorenzo Lotto’s Husband and Wife (c. 1543) appeared, a portrait of a couple seated at a table that is draped with an elaborately patterned tapestry. The painting, which Falco called his “Rosetta stone,” bears several signs that the artist used a lens—or a mirror—in the depiction of the cloth. While the husband and wife appear stiff and wax-like, Falco pointed out, the tapestry looks almost like a photograph, its complex patterns appearing to curve and dip flawlessly into the material’s folds. The fabric’s flaws, too, seem photographic. Although the pattern at the front of the cloth is sharp and precise, it becomes blurry in the near distance. That’s not the way the human eye captures images, Falco said: As the eye courses over a scene, it automatically refocuses. Lotto, Falco argued, could not have visualized this image in this way with his eyes alone.

The painting yields quantitative evidence as well, Falco continued. In works by Leonardo da Vinci, Piero della Francesca, and other Renaissance artists who mastered the laws of perspective, the angles converge at a single vanishing point, creating a realistic illusion of depth. But in Husband and Wife, the vanishing point of the tapestry’s angles changes thrice. In order to project the entire tapestry, Falco hypothesizes, Lotto had to position his lens or mirror three times. Each shift resulted in a slightly different magnification and therefore another vanishing point. The subtle errors that resulted are extremely unlikely to occur when an artist employs imagined and plotted perspective, said Falco. And the way the errors play out in the painting conforms neatly to geometric optics equations. Falco calculated precisely how much the three vanishing points would ultimately diverge from one another, and then predicted mathematically how much the magnifications would have had to change with each shift in order to achieve that divergence. “When we compared our predictions to the measurements from the painting,” he said, “they agreed to within 0.5 percent accuracy.”

Falco next turned his analysis to the near-flawless perspective, symmetry, and detail of the complex chandelier in van Eyck’s Arnolfini Marriage (1434), a painting in which nearly everything else retains the flat character of the medieval style. Falco presented a digitally enlarged view of the chandelier (which, in the painting, measures no bigger than the size of a cupped hand), and then, through computer animation, rotated and superimposed all six of the chandelier’s arms on top of one another. Although there are some small variations in arm size and ornamentation—differences the physicist chalks up to the painter’s creative choices—Falco calculated that the main arcs of all six arms are the same to within 1.5 percent in length and 5 percent in width. The probability of achieving such perfection without optical aids, he argued, is miniscule.

Evidence of the use of optic devices also appears in Renaissance portraiture, Falco told the audience. He showed the eerily perfect match—pores, nose hairs, warts, and all—between a small 1431 black-and-white van Eyck drawing of Cardinal Niccolo Albergati and a full-color, much larger portrait the artist painted a year later. Van Eyck, Falco argued, likely projected the small drawing as a guide for the larger painting; when Falco enlarged and superimposed the drawing onto the painting, they were all but identical.

When the lights rose, hands shot up around the room. David McFadden, a professor of chemistry, asked why the artists hadn’t discovered and corrected their mistakes in perspective, though he didn’t doubt, given Falco’s evidence, that they’d used optics. “I can’t tell you why [artists] did things, I can only tell you what [they] did,” Falco answered.

“Where is the written evidence?” a white-haired gentleman asked. “If it happened, why didn’t anyone write about it?” First of all, Falco said, it wasn’t clear such documentary evidence didn’t exist—“check Science or Nature for an article by me in about two months,” he suggested cryptically. He went on to offer two reasons for the scarcity of material support: Most of the texts from that time were written in Latin, which can be scientifically ambiguous; “glass,” he said, might actually mean “mirror.” Second, “These [painters] were businessmen.” They were using aids to paint accurately and quickly, but few would have wanted to give away trade secrets.

Jeffery Howe, an art history professor who specializes in late 19th- and early 20th-century art, asked why Falco had never tested to see whether artists drawing freehand would arrive at the same deviations he’d identified. “I’d like to do that,” Falco responded, “but you’d have to raise an artist on a desert island and never show him a photograph.” Modern-day artists’ perspectives reflect the photographic images they’ve grown up with, he said.

In conversations with a reporter days later, it was apparent that not all listeners found Falco’s answers satisfying. “Just because lenses existed in the Renaissance,” said Michael Mulhern, a studio art professor who teaches drawing, “doesn’t mean [artists] used them. . . . If you’re going to change the complete history of art, you have to do it on more than probability.” Art historian Kenneth Craig noted the absence of concave mirrors and lenses in “the many images showing artists at work in their workshops,” in which can be seen “many of their other tools”—or in period instruction manuals for artists, including that of Italian polymath Leon Battista Alberti, De Pictura, written in 1435, “shortly after he met van Eyck.”

At 6:30, after two and a half hours, Ellen Winner tried to end the session, but students and professors rushed to the podium. Trailing a swarm of questioners (“Why hadn’t these techniques shown up in non-Western art?”; “Did van Eyck use a lens in all of his paintings?”), Falco responded quickly to a few as he walked toward the exit. In the hallway, an art history professor and a physicist were conversing heatedly, their raised voices echoing down the corridor. “I don’t think they’re going to solve this one tonight,” Falco said. “They’re speaking to each other in different languages.”

Read more by Cara Feinberg