On connoisseurship

February 29 2012

Image of On connoisseurship

Picture: Philip Mould Ltd. Fig.1: Attributed to John Greenhill (c. 1644-1676, 'Portrait of John Locke', c.1672-6, Graphite on vellum, 5 1/4 in. high (oval) Private collection, U.S.A.

I recently wrote an article on connoisseurship in the US-based magazine, Fine Art Connoisseur. It's been a while since the issue was out, so here, with editor Peter Trippi's permission, is the full article for any readers who may be interested.

On the Importance of Connoisseurship

When the celebrated English philosopher John Locke sat to Godfrey Kneller for his portrait in 1704, he made a special request. He asked “Sir Godfrey to write on the backside of mine, John Locke 1704 ... this is necessary to be done,” he continued, “as else the pictures of private persons are lost in two or three generations and so the picture loses of its value, it being not known whom it was made to represent.’” 1

Sadly for Locke, not everyone has followed his advice. About a year ago, Philip Mould and I found a fine portrait drawing of him (Fig. 1) in a sale at Christie’s secondary saleroom in London. It was catalogued as Portrait of a Gentleman, and, proving that Locke was right to worry about his portrait’s future value, was bought for just £386 — a fraction of its true worth. It relates to a painting by John Greenhill in the National Portrait Gallery, London (Fig. 2).

Picture: National Portrait Gallery. Fig. 2, John Greenhill (c. 1644-1676), 'Portrait of John Locke', c.1672-76, Oil on canvas, 22 1/8 in. high (oval)

As art dealers, we scour the world’s auction catalogues daily for paintings that are in some way wrongly identified. In any week, our finds might range from a misidentified Tudor icon to a misattributed 18th-century landscape, not to mention some optimism-induced mistakes. Here, I want to discuss why paintings become detached from their identities or attributions — becoming in the process art history’s orphans — and how one can go about finding them again.

Contributing Factors

The most frequent cause of misattribution is simply the lack of a good label. This is particularly the case with portraits, even those that have been in the possession of the same family for many years. As Locke warned, it takes only a change in generation for important knowledge to be lost: after all, how many of us can easily recall the names of our own great-grandparents?

Even portraits of significant historical figures can become mis-identified. Intriguingly, new image and facial-recognition software can now identify portraits of popular figures. For example, upload a portrait of George Washington by Gilbert Stuart onto tineye.com and you’ll get a number of matches of varying different versions by or after Stuart. It is a fairly rudimentary technology — it would not have worked for that John Locke — but presumably, in the future, one will be able to identify any number of portraits. For now, it still helps to have that mental Rolodex of obscure historical portraits flipping through your mind.

Picture: Philip Mould Ltd. Figs. 3, before conservation, Anthony Van Dyck (1599-1641), 'Study for the Head of St. Joseph', c.1630, Oil on paper laid onto panel, 14 1/2 x 10 1/2 in.

A painting’s condition is also crucial, and is another major cause of misattribution. If a picture is obscured by centuries of thick and discolored varnish, smoke, and dirt, then it is obviously difficult to assess its quality. We recently bought a small profile sketch of an old man (Fig. 3) that had been catalogued at Christie’s South Kensington as by an anonymous artist. Such were the layers of disfiguring grime that the sketch seemed to be monochromatic, devoid of any color or life. But cleaning revealed a finely painted work by Sir Anthony Van Dyck (Fig. 4). It is Van Dyck’s study for the head of St. Joseph in his composition The Holy Family [Manchester Art Gallery], painted around 1630. 2

Picture: Philip Mould Ltd. Fig.4, Van Dyck's 'Study for the Head of St Joseph' after conservation. 

The hardest cases are those that cannot be directly linked to another existing work, unlike our portrait of Locke. Most lost paintings appear without any supporting evidence in the form of documentation, provenance, or past attribution. Such pictures call for a straightforward assessment of aesthetic quality — simply by looking at the painting, is it possible to tell who painted it? Can a physical observation of the canvas, the way the paint is applied, the composition, and the characterization of a sitter’s face alert the viewer to the presence of a portrait by a particular artist?

The ability to tell almost instinctively who painted a picture is defined (as readers of this magazine surely know) as connoisseurship. The word is derived from the Latin cognoscere, to get to know. The theory is that the repeated study of an artist’s work allows one to become so familiar with his or her style and technique that they can be easily recognized, just as we may recognize the author of a letter not from the signature at the end, but from the handwriting at the beginning.

With some artists, connoisseurial judgments are simpler than with others. For example, the Dutch “Golden Age” artist Pieter Jansz. Saenredam (1597-1665) had a reasonably distinctive style and confined himself mainly to painting church interiors. He had only two pupils, neither of whom went on to become a major artist in his own right. He has been called “a low risk case,” 3 a fact that may explain why a painting catalogued as “Studio of Saenredam” at Bonhams London in December 2010 with an estimate of £20,000-£30,000 sold for £1,476,000 after fierce bidding between some of the world’s top Old Master dealers. Their opinion was that the painting displayed all the hallmarks of a work by Saenredam, and the lack of any credible alternative candidate as the artist gave each of them the confidence to bid strongly. 4

Van Dyck, on the other hand, is an artist who would certainly be called a “high risk case.” Van Dyck’s peripatetic career and his restless inventiveness meant that his style and technique constantly evolved. A portrait painted by Van Dyck in Genoa in 1626 can look radically different from one painted in Antwerp just four or five years later. The connoisseur of Van Dyck’s paintings must navigate a complex range of attributional nuances. On the one hand there are straightforward copies, and on the other there are works by the master himself. But between these two extremes, the dividing lines become very blurred. First, his work and technique were widely copied. Then we have to contend with Van Dyck’s studio, in which he employed talented artists trained to paint exactly as he did. They not only made convincing repetitions of his own portraits, but worked with him on numerous pictures, especially large religious and classical scenes. The potential for confusion with an artist like Van Dyck is therefore endless. Telling the difference between a portrait painted by Van Dyck and one that was painted by both Van Dyck and a studio assistant can be challenging, to say the least.

The Demise of Close Looking

How, then, can connoisseurship be learned? The answer is only by repeatedly looking at pictures, and lots of them. To the uninitiated (and tineye.com), one portrait of George Washington by or after Gilbert Stuart can look like another. Recently, we bought, at a leading auction house in New York, a copy of Stuart’s portrait of Washington by Thomas Sully. But when the package arrived at our gallery in London, we found a different (and much better) portrait of Washington from Stuart’s studio. Thanks to a mix-up, the auction house had sold a different painting than that illustrated in the catalogue. To save them embarrassment, we sent it back. If the art handlers and specialists at the auction house had received even the most basic training in Washington’s iconography, the mistake would likely not have been made.

But the one place connoisseurship cannot be learned is in the classroom. And unfortunately, the classroom is where most art historians and students seem exclusively to study these days. In part, this is because connoisseurship itself has become a controversial concept. From about the late 1970s onwards, art history as a discipline saw a reaction against not only connoisseurship, and by extension the whole question of making attributions based on visual evidence, but against the study of artworks in their own right. In essence, the study of the object, be it a painting or a sculpture, became less important than the study of its context. Some art historians went so far as to declare the very notion of authorship irrelevant, their thesis chiming with the growing trend among historians to turn away from the study of the individual (not to mention the rise of literary criticism).

As a result, both art history and history as disciplines increasingly focused on identifying other elements that determined historical and art-historical “outcomes,” be they economic, social, or gender-based, in a headlong quest for generalization. And since connoisseurship inevitably involves a detailed biographical study of an individual artist, connoisseurship as a skill became less valued. The shift of emphasis in both history and art history is best reflected in their respective historiographies — modern historians wrote fewer biographies, and art historians wrote fewer catalogue raisonnés.

Today the result of this shift away from what one might call a traditional history of art is that not enough art historians are, if I may say so, connoisseurs. I know of at least two scholars compiling monographs on big-name artists, books that will be billed as “definitive” guides when published. The trouble is, neither of these scholars can tell the difference between a work by their chosen artist and a hole in the wall. Here’s another example: a friend of mine once sat through a lecture devoted to the contextual analysis of a single painting, which, the lecturer said, helped explain key facts about a particular artist’s life and work. But the work in question was by an entirely different artist.

Of course, if few art historians are connoisseurs, then it follows that even fewer art history students (the curators, scholars, and auction house specialists of the future) are connoisseurs — which in turn helps to explain why opportunities arise to find miscatalogued paintings at auction, and even why, sometimes, museums deaccession important works by mistake.

Moving Forward

Happily, there are encouraging signs that the pendulum is swinging back in favor of connoisseurship, at least among some art historians. Two recent exhibitions at the National Gallery in London have included sections on connoisseurship. The most recent was a display on the gallery’s first director, Sir Charles Eastlake (1793-1865). For Eastlake, sound connoisseurship was an essential skill. He would regularly travel across Europe, buying expensive Old Masters for his gallery that were often unidentified. He had to be able to tell quickly the difference between Giovanni and Gentile Bellini, without photographs, or Google.

In fact, modern technology is an increasingly useful tool for today’s ambitious connoisseur. High-resolution digital photography allows the close comparison of works many thousands of miles apart. New methods of scientific analysis are also an aid to connoisseurship. It is now possible to determine what type of pigment or canvas a particular artist favored. Before, connoisseurs relied too heavily on their “instinct” to attribute paintings, giving the practice a bad name. In 1939 the noted art historian Max Friedlander wrote, “The way in which an intuitive verdict is reached can, from the nature of things, only be described inadequately. A picture is shown to me. I glance at it, and declare it to be a work by Memling, without having proceeded to an examination of its full complexity of artistic form.” Unsurprisingly, only about half of Friedlander’s attributions have stood the test of time. 5 Now, science can help us be far more accurate.

But to really get to grips with an artist’s oeuvre, there is no substitute for intimate first-hand experience of his or her work — and not just the well known clean and shiny examples in leading museums (a mistake too many academics make). My advice to any budding connoisseur is to get out there and look at as many paintings by your favored artists as possible, in whatever condition you can find them. Oh, and take a flashlight.


1. David Piper, Catalogue of Seventeenth-Century Portraits in the National Portrait Gallery, 1625-1714, Cambridge University Press, 1963, p. 209.

2. Originally estimated at £5,000-£8,000, this picture was bought by Philip Mould Ltd for £121,250 and is now valued at £350,000. It is [was] temporarily on display at the Rubenshuis museum in Antwerp.

3. David Carrier, quoting Gary Schwartz discussing Saenredam in “In Praise of Connoisseurship,” in The Journal of Aesthetics and Art Criticism, 1:2 (Spring 2003), p. 166.

4. Bonhams Old Master Paintings, London, December 8, 2010, sale 17863, lot 61: The North Transept and Choir Chapel of the Sint Janskerk, Utrecht, oil on panel, 50.6 x 40.7 cm. Price includes buyer’s premium.

5. Cited in Ernst van de Wetering, “Connoisseurship and Rembrandt’s Paintings: New Directions in the Rembrandt Research Project, Part II,” in The Burlington Magazine, CL:1259 (February 2008), p. 90.

Notice to "Internet Explorer" Users

You are seeing this notice because you are using Internet Explorer 6.0 (or older version). IE6 is now a deprecated browser which this website no longer supports. To view the Art History News website, you can easily do so by downloading one of the following, freely available browsers:

Once you have upgraded your browser, you can return to this page using the new application, whereupon this notice will have been replaced by the full website and its content.