6 ideas
Around 1430 in Flanders, painters such as Jan van Eyck shifted abruptly from stylized forms to highly detailed, optically convincing rendering of fabrics, metals and faces. The book argues that no gradual improvement in skill explains a jump this sudden. It attributes the change to painters adopting concave mirrors and lenses to project images of their subjects onto a surface.
Some objects are too hard to draw freehand with perfect accuracy, such as foreshortened patterned fabric, ornate chandeliers, armor reflections and folded collars. When these appear flawlessly rendered while other parts of the same picture are awkward, that mismatch points to projection. Looking closely at such details, rather than at overall composition, changes what counts as evidence of how a picture was made.
A lens or mirror projects a picture from a single fixed point with limited depth of field, and it tends to produce certain things: sharp local detail, consistent directional lighting, and left-handed sitters from mirror reversal. A painting built from several separate projections, refocused and pieced together, shows multiple vanishing points and scale shifts between areas, a telltale 'collage' of optical windows. A painter drawing by eye produces a different, more flexible space without these seams.
Hockney noticed that Ingres's small pencil portraits had the swift, confident contour lines he associated with tracing aids. He then used a camera lucida himself to draw sitters, and his own drawings took on the same quick, measured outlines. He used this hands-on replication by a working painter as evidence that the look of a drawing reveals the tool that made it.
Instead of relying mainly on documents, the argument pins hundreds of reproductions side by side in chronological order on one long wall, so that sudden changes in optical quality become visible across centuries and regions. Pattern recognition by a practitioner's eye is treated as primary evidence, and texts and scientific tests are used afterward to corroborate it. The method makes the case compelling to lookers but leaves it open to challenges about causation and missing documentation.
The book holds that using lenses and mirrors does not diminish the masters, because a projection only gives fleeting outlines and tonal cues. The painter still had to fix the image, coordinate multiple projections and supply color and meaning.