“Don’t overwork it” is
terrible advice. Even the freshest of Impressionists reworked paintings.
![]() |
Vase of Flowers, started 1882,
Claude Monet
|
For earlier generations, a painting was complete when it had
a slick surface with plenty of detail. The mechanics of painting were carefully
hidden underneath the bling of the finish layer. Part of the ideal was that the
viewer should have no idea about the sheer hard graft involved in painting. Unfinished
paintings had no place in collections and were often destroyed on an artist’s
death.
Late 19th century painters inserted the process
of painting into the finished work. They used thick impasto, left parts of
their canvas bare, and kept outlines and drawing marks visible. That made the
painting a temporal record of development as much as a snapshot of a moment. These
ideas continue into the modern period (perhaps in some cases to overripeness).
![]() |
Edmondo and Therese Morbilli, started 1866, Edgar Degas
|
But were they ever that straightforward? Recently, researchers
Kimberly Jones and Ann Hoenigswald analyzed
Edgar Degas’ portrait of Edmondo and Thérèse Morbilli at the National
Gallery of Art in Washington. They were trying to determine what “finished”
meant to Degas.
Degas was a compulsive tweaker of his paintings, sometimes
repainting them even after they’d been shown and sold. Many of
his paintings appear to be unfinished, but are they? Jones and Hoenigswald discovered that this Morbilli portrait (he painted the couple more than once) was extensively reworked over a period of
decades. Passages that appear to be open ground were, in fact, painted
over sections with finished detail. The researchers' conclusion was that, even though they
could lay the order of his process bare, they could not determine his
intentions.
The myth is that Impressionists recorded things quickly,
easily and confidently. Claude Monet’s Vase
of Flowers appears to be a finished painting, but in his
correspondence, he mentioned his dissatisfaction with it. Analysis shows that he repeatedly returned to it, scraping paint off or
painting over dried sections. It sat in his studio for forty years, occasionally
being reworked, until he signed it in the 1920s. Did he truly think it was
finished, or was he, in his eighties, just sick of working on it?
Part of the shift in what constituted ‘finished’ might have
been driven by economics. During the Renaissance, most work
was done to commission. An artist didn’t have the luxury to continuously fiddle
with his work.
But part of this is also attitude, and it’s worth thinking
about in terms of our own work. Camille Picasso famously said, ‘To finish a
work is to kill it.’ The modern interpretation of this is the overused injunction
to “not overwork” a painting. This is a corruption of the process-baring aesthetic, and usually terrible advice. If you
don’t hit your limit, you’ll never learn how to negotiate past or around it.
Jones and Hoenigswald’s analysis shows us that, for Degas, it was possible to
bring a painting back from the brink repeatedly.
When part of the aesthetic is process—as it is now—only the artist can definitely say that a painting is finished.
How we arrive at completion is another matter. I prefer a less-detailed surface myself. I often get there by painting over, painting out, and scraping out fussy passages, and, yes, paintings sit around my studio for a long time sometimes. It’s nice to see these techniques validated by art historians.
When part of the aesthetic is process—as it is now—only the artist can definitely say that a painting is finished.
How we arrive at completion is another matter. I prefer a less-detailed surface myself. I often get there by painting over, painting out, and scraping out fussy passages, and, yes, paintings sit around my studio for a long time sometimes. It’s nice to see these techniques validated by art historians.


