Why Copying the Masters Is Back (Done Right)

In 1900, a serious painting student would spend several months a year producing careful copies of Velázquez, Reynolds, Rubens, Bouguereau. The copying wasn't a side activity. It was the core of training. A first-year student might spend half their studio hours on master copies before being allowed to attempt original composition.

By 2000, this practice had nearly vanished from American and European university art programs. It was, in the language of the period, derivative. Original work was the only legitimate output of an art student.

It's coming back. Worth saying why.

What it actually is

Master copy isn't tracing. It isn't making a near-photograph. It's a disciplined reproduction of a specific work, at scale or close to it, using the materials and procedure the original painter used.

Solomon, in 1910, describes the procedure pretty cleanly: study the original for half an hour before you touch a pencil. Block in proportions. Lay in the underpainting in the order the original was built — usually monochrome, value-first. Build color in glazes and scumbles the same way. Compare. Correct. Repeat.

The whole thing is taught as a way to learn to see, not as a way to produce a saleable picture. Most working ateliers in 1900 had a stack of copies in the back room that the students had made and would never display.

What we noticed at the corpus level

We built recreation guides for 2,491 famous paintings. Across all of them, the same handful of techniques kept showing up: grisaille underpainting in 85% of works, glazing or scumbling in 78%, fat-over-lean in essentially all the oil paintings. These aren't obscure techniques and they aren't style-specific. They're the substrate.

A painter who hasn't done careful master copies tends not to internalize this. They've seen a lot of finished surfaces but they haven't reverse-engineered the lay-in. The reason copying transmits so well is that it forces the analytical work — what did the master actually do first, second, third? — that mere looking lets you skip.

Why it became unfashionable

The 20th-century rejection wasn't stupid. Three positions, each with real weight:

• The originality argument. If art's value is conceptual originality, copying is at best a warm-up.
• The pastiche risk. Painters trained to imitate masters tend to produce imitation masters. The late academic painting of 1880–1910 — much of which sits in Beaux-Arts museum basements — was technically polished but unable to escape its own training.
• The economics of attention. Four-year university programs have limited hours. Time spent copying is time not spent on original work.

The mistake wasn't the questions. It was the absolute answer. No master copy at all turned out to be the wrong amount, the way only master copy and nothing else would also be wrong.

What changed

Three things:

The ateliers reopened. Florence Academy in 1991. Grand Central in 2010. Watts. Angel Academy. Aristides. These institutions train classical painters using the 19th-century curriculum updated where appropriate but unmodified in its core sequence — and the painters coming out of them produce work the contemporary art world recognizes as serious. The pure-modernist position that traditional training prevents originality turned out to be falsifiable, and it got falsified.

The internet made the technical gap visible. A generation of painters trained without the classical curriculum can see, on Instagram and YouTube, what the atelier-trained painters can do. Whatever the philosophical arguments, the technical comparison is settled.

And — this part is funny — working painters never actually stopped copying. Sargent copied Velázquez extensively. Van Gogh did dozens of master copies, including the famous ones of Millet and Delacroix. Picasso made copies and variations on Velázquez and Manet throughout his career. The 20th-century official position against the practice was always at odds with what the actual working painters were doing.

Where Apprentice fits

We built our recreation guides specifically for this kind of practice. Each guide breaks down a specific master painting into its documentable techniques, phase by phase, with citations to the classical books that describe those techniques. The guide isn't a substitute for studying the original — it's scaffolding that gives a careful student a clear sense of what the master was doing at each step.

It's the missing third option. The first option is an in-person atelier — best in absolute terms but geographically and financially constrained. The second is a university MFA program — accessible but no longer teaching the classical tradition. The third is the atelier curriculum, applied to specific works, available outside the geographic and credentialing constraints of the original ateliers.

We think that third option has been quietly missing for a hundred years.

The collection is at /collections. The books we cite — Solomon's chapter on copying among them — are at /sources.