Click here to subscribe to M.A.D.

The Winterthur Furniture Forum: What Is Original?

by Lita Solis-Cohen

The enormous advances in science have changed our understanding of the surfaces on American furniture. The March 2 and 3 furniture forum at Winterthur asked the question "What is original?" and put forth some answers.

In the past some collectors were far ahead of the current thinking. In her unpublished and undated memoir, Mrs. J. Insley Blair, who did her collecting in the 1920's, 1930's, and 1940's, wrote: "The 'Amateur' collectors, the really wise ones who bought to sell, wanted the condition as nearly original as possible, and did practically no restoring, so that their pieces should be unquestioned when the sale finally took place. This was good judgment, as each collector has his own ideas as to how pieces should be restored and what finish is most desirable. For me, personally, some of the finest pieces and most important, from the point of rarity, have been utterly ruined by doing too much to them. This is especially true of the finish—the 'Grand Rapids' look for which nothing can usually be done. In a few instances I have had luck with some done over in the 1880's and 1890's which fortunately at that time had not been scraped down to the wood. Left with the old finish still under the French polish and red stain, these came back to their former state with careful handling."

Morrison Heckscher, the Lawrence A. Fleischman chairman of the American Wing at the Metropolitan Museum of Art in New York City, read this excerpt during a panel discussion at the end of the first day of lectures at the forum. About 275 people lucky enough to get a ticket learned about cutting-edge scientific methods and critical thinking.

Speaking to an impressive audience of appraisers, dealers, conservators, and students of American furniture, the speakers, most of them conservators, explored the methods used to decipher layers of finish on American furniture. Using video-assisted demonstrations and PowerPoint illustrated talks, the speakers also showed how surfaces inherently degraded over time and then gave some options for their treatment today.

It is always difficult to play to a diverse audience, to strike the right balance between new scholarship and technical innovations, to hold the interest of the well-informed while teaching those who are coming to the subject for the first time. The conference may have been too elementary for some professionals in the audience, but it certainly brought the layman up-to-date in a hurry, imparting information that would be tedious to get from a book.

Wendy Cooper, Winterthur's senior furniture curator, who with Gregory Landrey, director of Winterthur's conservation program, put this forum together, began by pointing out the dilemmas. Using a slide of a John Singleton Copley portrait that showed a shiny surface on a mahogany table, Cooper noted that centuries of dirt from fires, mildew, etc. have certainly changed the surface of that table today.

"Should we restore it to its original look?" she asked, adding that it has probably been repolished many times in its history and only by some rare chance had it been left alone. Cooper referred to John Kirk's chapter in his 1975 book The Impecunious Collector's Guide to American Antiques, "Buy it ratty and leave it alone." She said she has watched the pendulum swing over the years, from shining it up until crud became king.

She acknowledged that Winterthur paid a king's ransom ($1,688,000) for a Peter Stretch clock at the Jeffords sale at Sotheby's in October 2004. It is a 9' tall clock with original crud. The rare wavy mahogany is barely visible except inside the door. She said that at the end of the conference she would ask the audience to vote on what the museum should do. Should the virgin surface be left alone or lightly cleaned? Repair the crack or leave it as is? Or should the museum continue to study it?

To prepare the audience for an intelligent response, a parade of scholars discussed every aspect of the surfaces of wooden objects.

Michael Podmaniczky, senior furniture conservator at Winterthur, discussed "Topography of a Surface: Before the Finish." He showed how the finish enhanced the wood, how different hand planes presented different surfaces, and how carving can lose vitality over time. He talked about what is original, authentic, period, and genuine and what only appears to be genuine.

Robert D. Mussey Jr., a private conservator in Boston and author of The Furniture Masterworks of John and Thomas Seymour (the recent book on the Seymour shop in Federal Boston), spoke about the technical aspect of finishes. He defined an original finish as "the triumph of hope over documented reality." Using early cabinetmakers' recipe books, he talked about the arts of varnishing, dyeing, and staining. He said the earliest cabinetmakers' price books had no prices for finishing. In 1803 Thomas Sheraton, in his The Cabinet Dictionary, talked about oiling and polishing.

Mussey spoke of smoothing the surface with pumice stone, water, or oil. He said brick dust in water or oil applied with a rag gives wood color and fills the grain, providing a smooth surface. He spoke of old methods of scouring the wooden surface with rushes or glass paper, a forerunner of sandpaper.

Defining staining as the application of a colorant without a binder, Mussey said it required a clear finish on top. He said most dyeing, which penetrates the wood, was used for coloring veneers. Binders such as linseed oil and sometimes walnut oil or gum were mixed with pigment as a semitransparent finish applied to birch, maple, and poplar to make them look like walnut and mahogany. The Dunlaps made a mahogany stain out of logwood (campeche wood) that was shipped to New England in vast quantities.

Mussey then discussed oil polishing. It is quick and inexpensive. Linseed oil, red lead, and turpentine soaked into the wood, and many coats created a glossy finish. Sometimes brick dust and oil were used to polish wood to a high gloss. Oil polishing, however, didn't hold up well. Beeswax was the second-most common finish. The wax was mixed with spirits of turpentine. But it doesn't wear well either. It gets sticky and washes off. Wood was often covered with varnish, that is, a spirit varnish of resin mixed with alcohol that when applied sits on the surface of the wood, is transparent, and protective. "You can see into the wood cells, which gives it an iridescent appearance," Mussey said.

A bit after 1800, French polish was introduced. The first written record was in the 1820's, but it was common by the 1830's. After building up many coats of spirit varnish (resins in alcohol), it was rubbed to a high gloss. In 1825 a French book called it "English polishing." There was endless experimentation with plant and tree resins, clear and colored fossil resins, like amber and copal, dissolved in hot linseed oil or tung oil from an Oriental tree, with great variation in color.

In the Federal period hard, colorless spirit varnish was developed for light woods like satinwood and maple. After 1840, finishers cooked the resin and dissolved it in hot linseed oil thinned with turpentine to produce a piano finish, which Mussey said is hard to get off when it alligators. Mussey uses paint stripper to remove it.

At the other end of the spectrum, he said he sometimes can wash off crud with water and find an old finish. He bemoaned the practice of scraping off the old finish down to the wood and putting on a new finish, as was done repeatedly in the late 19th century through the 1970's. Focusing on what was done in the past, Mussey warned that today whatever is done should be conservative. Do not rush in to take off material before you know what is there. Another important point that was stressed was the number of tools we have today to help us know what is there and the well-tested methods to remove just one layer at a time.

Mark Anderson was supposed to demonstrate the making and applying of dyes, stains, and varnishes, but he was ill. Instead, Christine Thomson, a furniture conservator in private practice in Salem, Massachusetts, presented a video-assisted in-person demonstration of gilding and burnishing. She demonstrated oil gilding and water gilding. On the workbench of the late John Bivins, a recent gift to Winterthur by his widow Anne McPherson, Thomson laid out the tools of her trade: a leather pad, squirrel fur brushes, gold leaf, and an agate stone in the shape of a long dog's tooth for burnishing water gilding.

For oil gilding, she used oil varnish (or size) mixed with a little bit of yellow ochre paint. Then, with a brush she followed an anthemion design she had incised on a black surface. She then laid on the gold leaf. Oil gilding cannot be burnished but was given a coat of varnish, she said. This is how decoration was put on Classical furniture.

Water gilding is more complicated. More materials were used, and there are more steps. Mixing hide glue with gilder's gesso (a white material), she applied it to the wooden surface to be gilded. She then covered the surface to be gilded with a colored clay ground, called a "bole." Clay mixed with glue and with blue, green, black, gray, red, or ochre pigment was used. The gold leaf was applied and burnished, and then a protective material was applied. Sometimes it was shellac with a color toner added.

As a bonus, Thomson also showed how to make a tortoiseshell surface, often the ground for japanning. First she painted with vermilion and then brushed on black. Using gold powder, not gold leaf, she gilded the raised japanned design, and then she painted leaf designs directly on the background.

Gregory Landrey's topic, "Understanding the Surface—from Day 1 to Year 2006," was an attempt to get the audience to understand and respect coatings and surfaces. He said the preferred surface on 18th- and 19th-century mahogany furniture was one that repelled water and had a shine. It was typically not achieved with oil or wax but with a resinous coating that has changed with time. In the early 1950's and 1960's and into the 1970's, furniture at Winterthur was given a coating of linseed oil, vinegar, and turpentine, now known as "Winterthur dressing," which produced a sticky, blackened, complex surface over time. Landrey pointed out that hygiene cleaning and coating removal are very different. The latter is another level of work.

Using slides, he demonstrated how light energy has an enormous impact on wood, changing its color irreversibly. He quoted the late Benno Forman who called patina, "a thin film that forms over the eye of the collector."

Using the Stretch clock as an example, Landrey asked if we were looking at an original surface. He can just make out the rare and exceptional figure in the mahogany on the interior of the door. A photograph of the surface in ultraviolet light shows that the finish is discontinuous. There is dirt, grime, and polish. He wondered if it is fundamentally wrong to leave it as it is, but then he wondered about preservation. He felt that the clock's surface needs more study. Winterthur has done nothing to it yet and will probably soon decide what to do. It is on display as the centerpiece in the galleries.

Landrey found a painted chair from Albany, New York, an easier object to conserve. Spectrographic analysis showed a cross-section with three layers of paint (blue, red, and gold leaf) as well as resin, oil shellac, and acrylic. Removing the three coats on top of the original coating was a big improvement. The chair was easier to read.

Landrey then addressed the surface history of the Wichmann-Fendelman schrank with sponged decoration, a promised gift to Winterthur from Helaine and Burton Fendelman, New York collectors. It is now on view in the galleries. Made between 1790 and 1810, it is attributed to John Beiber, who worked in Berks County, Pennsylvania. It is an example of fully developed decorative painting.

At one time it was covered with oak graining. That was scraped off prior to its exhibition in 1974. The most recent analysis of the painted surface shows that this overpainting (that was stripped off) had protected the original painted surface. Its removal was done sensitively, but the original varnish was swept away, removing the clouded cataract lens. Some of the molding has been retouched; some inpainting has been detected. It looks garish today because the varnish is gone. "Its appearance may be more original than any other painted object," said Landrey. It looks like it did when the decorator finished it.

To show that there are different answers to each piece of furniture, Landrey put up a slide of a Findlay card table with a finish that completely obscured the decorative layer. It was removed without affecting the Findlay varnish. On the other hand, the Dominy high chest, made in Long Island circa 1800, has no figured wood of note. The piece is about form. Although the transparency of the finish is lost, Winterthur decided to do nothing to it, to leave it alone. The message of the conference was, minimize change, make change reversible, and don't freeze into inaction.

A panel discussion with Heckscher, Mussey, Baltimore dealer Milly McGehee, and San Francisco collector Susan Doherty fielded a lot of questions.

Heckscher showed a slide of a Philadelphia high chest with a bird finial that when bought at Freeman's in Philadelphia in 1975 for $92,000 had an untouched surface. Against Heckscher's suggestions, Berry Tracy, then head of the American Wing, sent it to Sack Restoration, where a picture was taken documenting the surface being scraped down to the wood. A new finish was put on, and the fire gilt brasses were cleaned. Last year, the Met bought the Thomas Townsend chest-on-chest at Christie's and, after studying the surface, left it alone, thereby raising public awareness of old surfaces.

Susan Doherty, who collects Federal furniture, said she looks on inlay and veneer as sculpture on a flat surface. "Why would you obscure it when you could take off a layer without stripping it?" she asked. She suggested that different periods require different solutions.

Dealer Liz Feld of New York City's Hirschl & Adler Galleries, in the audience, suggested that Classical furniture needs to be brought back to a shiny surface, and dirty gilding needs to be cleaned.

The consensus of the panel was "when in doubt, don't." Tastes change. Leave an early finish on unless it obscures the intent of the maker. There is no single answer. Personal preferences differ. Some enjoy seeing years of patina; others find it unattractive.

Dealer Milly McGehee pointed out that conservation is market driven. "Once you take off a finish, you can't put it back," she said. "The decision is a partnership among collectors, curators, and conservators who want to preserve the original intent and patina and keep the option of changing their minds later. The object has to meet the aesthetic goals without changing the history of the piece."

The second day of the forum began with Nancy Goyne Evans's talk on "'Well painted with different colors as the buyer chooses': Historical Evidence of Original Surfaces on Vernacular Seating Furniture." She showed a slide of two rows of Windsor chairs drying on a rooftop, an enlarged detail taken from an early 19th-century bird's-eye view of New York. This careful scholar claims that green was the perfect color for Windsor chairs because most of them were used outdoors. Green was the color of choice until after the Revolution.

Showing engraved images of horse mills, paint pots, and animal bladders filled with prepared paint, Evans went on to document the use of green paint for Windsors in paintings and letters. She said it could be olive, sage, or pea green. An ad in the Pennsylvania Packet in 1785 offered knocked down Windsors as venture cargo, painted green, blue, red, white, and black. A color chart, also reproduced in the new (2005) volume III of her remarkable study Windsor-Chair Making in America: From Craft Shop to Consumer, shows green is the most popular color and yellow, next. After 1800, bamboo Windsors were painted cane color.

Thomas Jefferson ordered black-painted Windsors with gilt rings for Monticello and Poplar Forest. Evans found period ads for white-painted chairs, some with real mahogany arms. Brown chairs were painted with umber and sienna, and some were grain painted. Gray and cream chairs appealed sporadically. Prussian blue, developed in Berlin early in the 18th century, is uncommon and fugitive. It turned olive in time. By 1790-1810, chairs were painted with two-tone paint to simulate upholstery. She showed slides of Windsor chairs with painted imitation rush seats.

Furniture conservator Christine Thomson came back to talk about "Assessing Appearances: How Japanned and Gilt Surfaces Change Over Time." She said pure gold does not tarnish, but gold mixed with other metals gives the effect of solid gold and can change color. Some coatings on hardware that look like gold are not gold. Varnish mixed with turmeric makes brass look gold. A toner, called a "bole," was often used over the gesso; sometimes it was a yellow primer. (In America, it was often gray.) The gold leaf is applied by floating water on the surface and placing the gold leaf on it. Then it was sealed with varnish; nevertheless, it often flakes and falls off.

Stenciling is always oil gilding. A bronze powder (mixed with oil in the 19th century and with lacquer and acrylic in the 20th century) was used to touch-up gilt frames. It turns color with time. It is easier to remove dirt from water gilding than from oil gilding. It was done successfully on the Classical sofa illustrated on the cover of Wendy Cooper's 1993 book Classical Taste in America, 1800-1840. As for japanning, deciding which layers are original requires ultraviolet light. Later layers fluoresce in ultraviolet light. Varnish can be removed with new materials, and there is hope of recovering lost images under coats of varnish.

Susan L. Buck, a private conservator in Williamsburg, Virginia, is a whiz at surface analysis. She teaches a course at Winterthur on cross-section microscopy and how to analyze surfaces in historic houses and on furniture. She said analyzing surfaces is like doing a three-dimensional jigsaw puzzle with some of the pieces missing.

Buck showed how she arrived at the decision to remove obscuring layers of varnish and paint from the Hannah Barnard court cupboard at the Henry Ford Museum. Using enzymes and isopropanol gel, both of which are controllable, she began the slow, meticulous process of removing layers, checking the cross-section at every level, and gradually recovered the polychrome design. After removing the obscuring layers, she found the polychrome design related to the "SW" chest at Deerfield. It showed that the court cupboard was part of an important Connecticut River Valley shop tradition.

Prussian blue, a synthetic pigment first made in Berlin in 1703, was in America by 1715. After finding evidence of blue on the columns, Buck reconstituted the blue columns with a modern paint, a gouache, that is reversible.

The samples needed for cross-section microscopy are the size of a pinhead. Because each pigment appears separated, it can be identified by microscopy, and overpainted layers can be identified. Infrared microscopy showed the layers of Prussian blue and white lead on the columns of the Hannah Barnard chest.

Catherine Matsen, assistant scientist in the scientific research and analysis laboratory at Winterthur, spoke about her research identifying interior paint colors at the Corbit Sharp house in Odessa, Delaware. Using scanning electron microscopy (SEM), she mapped the elemental components and identified 19th- and 20th-century paints. Using energy dispersive spectroscopy (EDS), she was able to map the complex mixture of materials. Then using Fourier transform infrared microspectroscopy (FTIM) she could identify pigments such as Prussian blue, white lead, and chalk. Using X-ray fluorescence (XRF), a nondestructive method, she was able to identify inorganic components that adhere to the wood, such as chromium oxide green that was not available until 1862.

Finally, using the hand-held XRF instrument (the size of a large flashlight), she looked at the Peter Stretch clock and was able to obtain the elemental data of inorganic materials. "We are one of ten museums with a lab equipped with the instruments to do this sort of study," she said after her talk.

Those who took the workshops given on Wednesday and repeated on Saturday (before and after the central two-day forum) learned more. For example, at one of them Dr. Jennifer Mass, senior scientist at Winterthur's research lab, and Greg Landrey covered analysis and treatment options for furniture finishes. Dr. Mass discussed resins for furniture finishes and then demonstrated the hand-held XRF spectrometer. Landrey looked at the entire object with UV light and demonstrated viewing one of the cross-section samples under the microscope. It looks easy, but it really takes an experienced specialist to use the instruments and analyze the data.

(At other of the hour-and-a-half workshops, which were offered twice on Wednesday and twice again on Saturday, Richard Wolbers, associate professor and paintings conservator at the University of Delaware and Winterthur Art Conservation Program, did a workshop on cleaning systems and decision making. He discussed paintings and furniture and showed how enzymes can be used to remove "Winterthur dressing," which grabs dirt and turns black. Christine Thomson did a workshop on gilt surfaces. Michael Podmaniczky's workshop was on surfaces before the finish. Wendy Cooper held a workshop on Pennsylvania painted case furniture.)

After lunch on the second day of the forum, Peter Deen, a conservator and painter of decorated surfaces in Nottingham, Pennsylvania, gave a talk on "Painted Surfaces and Beyond: Learning to Look with Eyes Wide Open." Deen has been a paint conservator and decorator for 35 years. He showed his audience how to detect his work. He signs it with an "A" and/or scratches his name in several places. He worries that we have frightened off young collectors and said we "must do something positive to allow them the feel of good investment until their collecting disease kicks in." There's a large body of work that scares him, and unless you are educated to the illusion of age and surface, collecting can be a hazardous journey.

In order to look with eyes wide open, you must understand the painted surface and what causes it to change. What can the layman do? Get a magnifying glass, the best you can afford. Look for telltale signs of age, like crazing or fine cracks in the paint. Look for dimples on the tops of blanket chests where kids wearing hobnails on the bottoms of their shoes stood on it. Look for crescent moon dings on the sides where their shoes kicked the chest when they were using it as a bench, and notice where it has exfoliated.

A shot of sunlight overheats the paint; rain or moisture breaks the bind; and paint flakes off on the hard grain. Flyspecks are just about impossible to remove. Fake flyspecks will come off with a solvent. If the chest has cracks, there should not be original paint inside the crack. On the back of the chest the oxidation should be the same top to bottom. If the feet have a different oxidation discoloration, they are not original.

The unicorn chest in the Fraktur Room at Winterthur has new feet. They were obvious when viewed with the video-assist technology that Deen used for his live demonstration.

Deen then compared a real Mahantongo chest, painted green with birds and flowers, with one he made and signed in several places with his signature "A." He said a look through a loupe would show his paint had not crazed. And he said that if tested, his pigments are not those used in the 19th century. He also uses modern binders. He did use old longleaf yellow pine as the wood; the crack is shrinkage; but the stenciled rosettes are round and show no signs of shrinkage, as they do on the old chest. Moreover, the compass stars on the top are perfect circles showing no shrinkage.

Deen then went on to demonstrate vinegar graining. Using Prussian blue pigment mixed with vinegar and a little honey, he finger-painted it on plastic board, simulating sponging, and he used a scrunched-up plastic bag to give it a twist. He produced tiger maple grain using his hand on a board, and he made a worm design with a condom filled with sloshing water.

Deen was part of the panel discussion that followed. Also on the panel were Susan Buck, dealer Sumpter Priddy III, Jeannine Disviscour of the Maryland Historical Society, and Joan Johnson, a collector from Meadowbrook, Pennsylvania. The consensus was that there are different answers for different pieces. It is OK to remove "Winterthur dressing," but don't take a piece back so it looks brand new.

Thanks to digital photo editing software, you can make a digital representation of what a piece may have looked like. There are different standards for different styles. Generally, don't touch 18th-century furniture, but Classical furniture can be refinished and polished. The 17th-century Hannah Barnard court cupboard was incomprehensible before treatment. The treatment, however, stopped before removing the first varnish.

Conservators of paintings have gone way too far using materials that cut into the surface and then revarnishing. Painted furniture could suffer in the same way. Furniture restorers must be sensitive to the top layer of modern filth. Clean it but do not go too far was the message. Every object must tell the whole story. With the increase in knowledge, there is one group that says don't touch it, another that says they want to see the wood or the paint bright, and another that prefers moderation and goes slowly, exploring all the evidence first.

One question asked was, when does conservation become restoration? The answer: when there is structural change, when varnish is removed, and when cleaned to a point of inpainting. Peter Deen said he wants to read a piece as an aesthetic whole, and his business is largely inpainting and not always with reversible paint.

Michael Podmaniczky, in the audience, said the party line is that restoration is part of holistic preservation. There are cases when deterioration needs to be arrested. "If a chunk is out, put it back," he said, noting that much of our current conservation is determined by the marketplace. People stretch to buy things with old surfaces and don't have enough money left to do anything to it.

As for the Stretch clock, Winterthur is waiting to determine what story needs to be told for the current generation and future ones. They are asking for as much input as possible. In the meantime, they are dusting it with a soft brush and keeping all their options open. They continue to study it.

A vote by a show of hands reflected this thinking. When asked what to do about it, a quarter thought nothing should be done, half thought a light cleaning and filling the crack in the base was in order, and a quarter thought more study was needed before any decisions were made.

The fully subscribed forum ended with applause for all. Wendy Cooper suggested that several one-day mini-forums on this subject of original finish may be offered in the fall. The workshops, which were limited to eight to 12 people, were over-subscribed and will be part of the repeat offering. The 85 people on the waiting list for the March forum will get first crack at the tickets. Dates will be announced.

"What I hope people came away with was that very few pieces survive without anything having been done to them," said Cooper when the forum was over. The term original finish has not been used correctly in dealers' ads and auction house catalogs. Until you can describe in detail what has been done to a piece, it is more accurate to say that a piece with patina has an old surface.

Part of the fun of a forum is exchanging ideas with participants over lunch and coffee breaks. One topic of conversations was the collections put together in the 1960's and 1970's when removing old surfaces and refinishing was common practice, and their fate when they come on the market. We have already seen the Meyer sale in 1996 and the Blair sale in 2006 create market spikes. In both cases, pieces with old surfaces were embraced. Now that we have the tools for identifying various layers or finish and for removing one layer of dirt or additional modern wax or varnish or dressing at a time, it may well have an impact on the market, and dealers and auctioneers may have to catalog the surface.