{{ currentSlide }} / {{ totalSlides }}

On Cleaning

By Alain Goldrach - 30. September 2022
Alain Goldrach on the necessity and dangers of cleaning a painting.

Part conversation, part handbook, our In The Studio series visits conservators, framers and other craftsmen in our field to explore what it takes to look after an Old Master picture. For the second installment, we met with Alain Goldrach in his Upper East Side studio to discuss the necessity—and dangers—of cleaning a painting.

Originally from Paris, Alain is a protégé of John Brealey—widely considered one of the most important conservators of his generation. In his own words, Brearly trained him “like a racehorse”. From the Met, Alain went on to run the paintings conservation department at the MFA Boston for fifteen years. He has worked independently for a variety of institutions and private collections in the last five decades, as it turns out, with a lot of French paintings from the 18th century to Impressionism. Our conversation shows how the cleaning of paintings is as much an art as a science.


Click to listen
In The Studio with Alain Goldrach

What does it mean when somebody says a picture has a good clean in it?

A good clean means you know that the picture would benefit from cleaning; below a dirty layer, there’s a picture in good condition.

Is this dirty layer just discolored varnish?

It can be airborne dirt, smoke, soot from the fireplace, and it can be the discolored varnish—all these things can affect the look of the painting.

How do you clean a picture in a way that will not take away the original pigment?

Judging from the amount of pictures that have been overcleaned—many of which you can see in museums—clearly, cleaning is dangerous, and often abrasive. Cleaning is, in effect, the only non-reversible thing in the conservation of painting. Once cleaned, there is no going back.

You have to have a knowledge of the materials so that you use solvents that will not affect the original. Because if you remove everything that is extraneous to the paint layer, you will end up with a picture which will not necessarily read as a work by the artist you’re working on.

For the most part, pictures are treated the same way in terms of solvents and practices. What matters is the attitude. Paint discolors, colors can shift—green to brown, blue to black, though white remains very white. You have to account for all that which makes the picture a real illusion in terms of space, form, volume, mood, and atmosphere. This becomes a problem of interpretation and subjectivity.

Sometimes we see people walking around art fairs or auction previews with a UV light. What can somebody learn about a picture with the use of such a light?

What you see with a UV light is primarily the discolored varnish and the retouches. The haze of varnish is usually the discolored varnish, and any dark touches that fluoresce are usually the retouches in paint.

At times, there are things that fluoresce that were painted by the artist—it’s not that straightforward. If you look at Turner, for instance, you might find hundreds of original touches that fluoresce. If you look at Delacroix and Gericault, it’s the same thing. There’s a particular group of painters, including Courbet, Goya, Delacroix, and Gericault, that are very difficult to clean.

An expert examines a Caravaggesque Master painting under UV light at Sotheby’s London, Dec 2019 © Alamy / Guy Bell
Why is that?

Because of the insane technique that they used! They were inventors, experimenting all the time. Delacroix liked to use a ground, often red or cream, and then varnish it. Then he painted his picture in oil—some of it loaded with diluent, some of it loaded with medium—to achieve various effects. And if you take a solvent to something like this and undermine the varnish below first, you can take absolutely everything off the painting. Courbet used bitumen, which never dries—pigments sink into that like a pebble. And since the bitumen is soluble, everything can come off. With Goya, it’s typically all about the varnishes and glazes. And this is where experience comes in.

For instance, I cleaned a great Courbet at the Met, Woman with a Parrot (1866). It was a very difficult picture—a nightmare—and every day my heart fluttered because I could not tell what was going on. But then at the end, you end up with a picture like that one, one that is particularly beautiful and convincing—in terms of form, space, and movement. That gave me nothing but the greatest pride, and it gives me a thrill every time I visit it.

Woman with a Parrot, 1866, by Gustave Courbet at the Metropolitan Museum of Art, New York © Alfonso Vicente /Alamy

Sometimes cleaners use test patches. What are you testing?

You’re testing the solubility of the varnish. Its removal, when you end up with a little clean square, is always spectacular but never terribly instructive.

Do you always do it before you begin working on a picture?

Not necessarily. I do things in different stages. First I remove the dirt—both airborne and water soluble. Then I varnish the picture to see it saturated. I only  start cleaning after I do a little test, in the blue and red tones, mostly to figure out what is soluble and what tonalities you’re looking at.

There used to be a tradition where someone would clean the picture and leave a little square dirty. This was to show what they had done. It’s the reverse of doing a cleaning test in the corner, even though that’s not what cleaning is about.

What solvents do you use to clean?

You want the solvent to be the most effective and the least destructive. Most people use acetone as a basic solvent. It removes varnishes and natural resins very nicely. But acetone also removes oil paint, so you have to make it weak enough so it doesn’t remove the paint but strong enough to remove the varnish. You can also add mineral spirit and all sorts of other additives to prevent blanching.

How do you clean paintings that are not varnished?

You don’t clean them with solvent. Instead, you can use different dry and wet methods, including soap and water. Spit is the best, though if you have an enormous picture, that would obviously not be very convenient. But for a small picture, it does everything it needs. It has enzymes, and it’s not very wet.

Doesn’t it have bacteria?

So what? Everything has bacteria. Anyway, you can also use soap—I use a surgical soap called Green Soap. But soap is difficult, because it leaves soap scum.

Does the material support of the painting affect how you clean a picture?

Not really. The support mostly affects the structure of a painting—cracks, flaking, fragility and so forth.

And what kind of tools do you use?

Most of the time, you use swabs and friction to remove the varnish. But you cannot do that in places where the paint is loose, obviously. In that case, you can use rolling motions, setting the paint down, and all sorts of tricks to prevent the fibers from catching the loose paint.


For instance, I cleaned Roger van der Weyden’s Portrait of Francesco dEste  at the Met, which is a marvelous picture painted directly on the white background. It’s a very fine portrait, so it’s not something you want to put a big swab to, wash up and cover with varnish on top. The difficulty was that the picture was ridden with very fine craquelure. It was not so much of a problem in the white background, but on the face you could see how the cusp of these little cracks had been overcleaned in the past. So in this case, the retouching consisted of closing these cracks with watercolor. That easily took two months of work with a very fine brush.

Rogier van der Weyden, Francesco d’Este, ca. 1460, oil on panel, 29.8 x 20.3 cm., The Metropolitan Museum of Art, New York, The Friedsam Collection, Bequest of Michael Friedsam, 1931. Inv. no. 32.100.43

You touched upon the subject of over cleaning. How common is it precisely?

It happens every day, and it’s happened every day over the centuries. Take the example of a large collection of Italian primitives, which are not known for their fragilities, right? The Jarves Collection at Yale is the most famous one for this group of pictures, and their pictures were destroyed by the combination of an insane director and a completely incompetent restorer . They not only took the retouches off, but they also dug all the way to the panel with scalpels and drills—they wanted to recover what was beyond the original.

Do you think they were influenced by the trend in conservation back then?

This was 50 years ago. In America, this came out of several schools. There was a frenzy for wax lining, plastic varnishes, and acrylic paint, and very little interest in a subjective approach to cleaning. But it’s not always the case that things that are not right necessarily have to go. We simply don’t know enough.

Piero di Cosimo, Virgin and Child with Saints Vincent Ferrer and Jerome, ca. 1508, oil on panel, 208.9 × 205.1 cm, Yale University Art Gallery, University Purchase from James Jackson Jarves, Inv. no. 1871.73

So you don’t necessarily remove all of the retouches from the past when you’re cleaning?

The retouches are not a problem. It’s a question of the level of cleaning for different tones so that they make sense—so that suddenly, the space reopens and that the form comes back, and you acquire volume. The discoloration of the varnish affects many, many things aside from the color. It also affects the reading of the picture, the luminosity, and the atmosphere. So what you do is to try to get some sense of what was there.

Can you give an example of something not being right, but perhaps worth keeping?

Well, what am I going to say? Salvator Mundi, the, new Leonardo da Vinci. There are some beautiful parts in it. I mean, you shouldn’t throw it away!

The Capponi portrait at the Frick had the codpiece painted over it a long time ago. That was not based on science, but prudery!

But to clean a picture of that importance just to reveal the codpiece—maybe one should ask, should we do so? There’s something very nice about visiting a museum that has not cleaned everything. That way, you’ll still see a testament to the past, to people’s tastes, old surfaces, patina. All these things go by the wayside once an entire collection is cleaned.

So when does it become truly necessary to clean a picture?

Well, it depends. If you’re going to keep it on your wall, I would say don’t clean it. If you’re going to sell it, sometimes you should, sometimes you shouldn’t. But sometimes, cleaning is part of the business of selling pictures.

Let’s say, that you have a real problem in the sense that you’re looking at a picture which is heavily repainted—through various techniques like infrared, X-ray, you may be able to determine that there is overpaint all around. So you’d say this is definitely a picture we must clean because X-ray has proven that there is a picture below.

Like the cleaning of the Vermeer in the Dresden Gemäldegalerie that has revealed the cupid in the background.

Right, exactly.

Before cleaning
Johannes Vermeer, Girl Reading a Letter at the Open Window, 1657-59 Condition after restoration © Gemäldegalerie Alte Meister, Staatliche Kunstsammlungen Dresden, Photo: Wolfgang Kreische

And how can a collector identify the pictures that are worth cleaning?

How do you acquire your sense of quality? It’s like music. If you’ve never heard music, how would you know if it’s good or bad? I think it’s all part of an education, interest, and knowledge, which you can acquire through books or visits. The best thing to do would be to organize trips to various museums and maintain a steady diet of looking and comparing.❖

Alain Goldrach is the owner of Alain Goldrach Restoration ltd., New York.
Editor: Mary Wang
You may also like