The great Bendor Grosvenor tries to explain the mess that the Rembrandt Research Project has made of the artist’s oeuvre. Here’s how it happened but you’ll have to—and ought to—click through to read Grosvenor’s very sensible explanations for why so many of the works that are still disputed have not be reclaimed by their institutions:

In the first half of the 20th century, Rembrandt was believed to have painted some 600-650 works. But from the 1970s onwards that number shrank rapidly to around 250.

What happened? Put simply, Rembrandt connoisseurship (that is, the ability to tell who painted what by close inspection) imploded. In 1968, the Rembrandt Research Project (RRP) was established with an admirable objective – to say definitively what was and was not a Rembrandt. But two key factors doomed the RRP’s approach. First, it tried to make attributions by committee, thus allowing indecision and groupthink to reign. It is easier and less risky to say “no” to a picture than to say “yes”. In such situations, the hardest-to-please scholars gain kudos for being “disciplined”, and influence others.

Second, connoisseurship itself fell out of fashion. “New art history” (which became dominant from the late 1970s onwards) believed that connoisseurship was a redundant, elitist practice, and was no longer taught as a key skill for art historians and curators. Social, economic and philosophical generalisation was the order of the day. As a result, the wide and informed debate that should have taken place every time a Rembrandt attribution was questioned didn’t happen. Few ever came to Rembrandt’s defence.

How to identify a real Rembrandt?  (FT.com)