5 INTERPRETING THE EVIDENCE OF VARNISHED AND UNVARNISHED PAINTINGS

The evidence of paintings themselves—whether a painting was never varnished or has an old varnish layer—must be interpreted with great care. Proving intent on the part of an artist can be extremely difficult.

One of the few cases in which we can say with confidence that an artist intended a painting to be varnished is when one sees an artist's retouching on top of a varnish layer. In these rare instances (see the discussions of Hassam and Chase below, and other American Impressionist painters in Mayer and Myers 1993), one can say with certainty, based on the evidence of the painting itself, that a painting had varnish on it while it was in the artist's hands.

The survival of a large number of unvarnished paintings by an artist might suggest that the painter intended them to remain unvarnished. The case of John Steuart Curry (1897–1946) (not an Impressionist painter, but a painter whose techniques are well documented) is an example of how such a supposition could be incorrect. The majority of paintings that remained in the possession of Curry's widow after his death were unvarnished, but a study of letters from Curry to his dealers make it clear that Curry intended that his paintings be varnished, and he left that job to his dealers—to be carried out when paintings were sent to New York for exhibition (Mayer and Myers 2002, 27). Many painters, including Curry, believed that paintings should not be varnished until about a year after they were completed. Hassam's correspondence about his flag paintings (discussed below) shows that Hassam, too, believed that an artist should wait a period of time before varnishing. So when one sees an unvarnished painting, even one that has survived unvarnished until the present day, one must admit the possibility that the artist may have intended that it be varnished in the future, but no one ever got around to it.

It is also important to remember that when one sees a painting that has a varnish layer, even a very old varnish layer, the painting might have been varnished against the artist's wishes by a dealer or owner. One might ask why so many paintings have been varnished in spite of the fact that artists may have not wanted them varnished. Since varnishing was often done by a dealer, framer, or other third party, it might have been done simply as a matter of course, without considering the artist's intent. Because unvarnished paintings would have looked unusual at the time, a dealer might have thought that varnish made a painting look more “normal” and therefore more salable. Another justification for varnishing has been that varnishing has a protective function; Ranger used this argument when he advocated the varnishing of every painting, including paintings by Monet that Ranger knew Monet intended to remain unvarnished (Bell 1914). Another reason that unvarnished paintings might have been varnished is that the increased contrast of values and intensification of colors that occur upon varnishing might have been seen by some as an improvement. A dealer might believe that brighter colors would attract a customer, or a conservator might believe that making a noticeable change in a painting's appearance would justify the cost of treatment. As discussed in connection with Willard Metcalf and Theodore Robinson below, this sort of apparent “improvement” is often at war with the proper relationship of the parts of a painting and the appearance of the painting as a whole.

Painters may also change their practice over the course of their careers, and evidence from one period of an artist's career does not necessarily apply to other periods. Impressionist (and non-Impressionist) painters may also have made decisions about varnishing on a case-by-case basis—varnishing some paintings but not others. Charles Hawthorne (1872–1930) wrote that a dark painting might require varnish while a light-colored one may not (Hawthorne 1938[1999]). Other painters can be documented to have had a case-by-case approach to varnishing individual paintings (Mayer and Myers 1993, 135; see Mayer 1998, 27, 29, for examples from earlier periods).