Gustave Caillebotte: Mystery in the Making

Le Pont de l'Europe, Gustave Caillebotte 1886

Every time I walk down the street, the most instinctive, primal part of my brain asks two questions of each person I pass:

1: "Is this person a potential threat?"

2: "Is this person a potential mate?"

The first question may just be a throwback to growing up in Dublin, where every approaching kid in the lane by the railroad tracks was a potential threat to this skinny art kid. It was always tense and nervy. The second is, well, just human nature. It's a fleeting nod to my simian side, and I don't have to act on it to recognize its existence. I'm a believer that we are animals first, and humans second. But maybe that's just me, I don't know.

Anyhoo: If the answer to both questions is "no," then I feel nothing and go about my day. But if the answer is "yes" to either one (and if I'm being honest), then there's a momentary charge in the air. Even the mutt in the painting seems to feel it. The point is; I believe that Caillebotte created his painting, "Le Pont de l'Europe," to deal specifically with this instant of primal recognition. The mystery that's endured speculation is; to which question is he answering "yes?"

The dress code may be different, but "Le Pont De l'Europe" depicts that same scene from the railroad tracks of my childhood, except that here it's a flaneur from nineteenth century Paris instead of some scumbag from Dublin. 

Le Pont de l'Europe from Le Gare St. Lazare, 1868 [source]

The area around the railway station, Le Gare St. Lazare, was pretty seedy around the time. The protagonist (let's assume it's Caillebotte for now) is seen walking a couple of steps ahead of a woman dressed in black. Women dressed in this manner, and hanging out by the railway station in nineteenth century France, were assumed to be prostitutes. If we assume that this woman is walking alone, and that old Gus has simply passed her by on the street, then we can be pretty sure in this assumption as to her trade. Women rarely, if ever, would walk unaccompanied in so rough a neighborhood. 

And if he is walking with her, why is he acting cagey and walking so far ahead? No well bred man would precede his companion. Either way, and despite the fact that she is looking at him, Caillebotte seems oblivious to her presence and is instead gazing at the man on the bridge. 

Dressed in the blue and bowler hat of a laborer, perhaps he is simply lost in thoughts of work and life. Caillebotte the engineer and lifelong bachelor, might naturally have an affinity for this. But perhaps this man in blue - idle during what is after all the middle of the work day, judging by the shadows - is on the prowl for someone just like Caillebotte. The painting has been interpreted as two men cruising each other.

I like how Caillebotte changed the architecture of the bridge between paintings

As if to prove my point, Caillebotte created this second painting around the same time. It's a wonderful cinematic moment that seems to capture the scene a few seconds after the first painting. Caillebotte has ditched his female companion (for it was never her that he was interested in), and has stopped by the man in blue. Is this just the engineer in him that's been distracted by the marvels of steel and industry on the horizon, or is he up for a little what-not in the hoo-ha with our man in blue? 

Well, "who cares," is the obvious answer. But it's fun to speculate, and the air of mystery around the painting is much of its enduring charm. What's especially interesting is how he has used linear perspective to help construct his narrative. 

He tampered with the reality of the actual scene, straightening buildings on the background that are in actuality oblique to the bridge, and widened the left of the bridge so that "the universe faithfully transcribed proves an illusion, and accuracy covers a lie." [Kirk Varnedoe]

"The superfluity of lines, which all seem to converge on the profile of the man, raises the question of what exactly Caillebotte was seeking to achieve through such unusual emphasis. Everything would seem to suggest that it identifies this man as the protagonist of a narrative encompassing the entire work." [Hirmer]

Various interpretations exist. Some say that this rigidly constructed scene "depicts a city that would have interested the Futurists, a place of dynamic intersections and personal anonymity, where human relationships count for nothing." [source]

"They seem to be quite alone," says Nicolas Sainte Fare Garnot, curator of the Jacquemart-Andre Museum. "Every person is lost in a very wide world."

In an earlier painted sketch (above), the architecture of the bridge dominates and overwhelms the scene. Though we see the same man, it's hard for our attention not to be sucked to the back of the painting, towards Place de l'Europe.

This sketch seems to back up the "critique of modernization" interpretation, with its overbearing architecture dwarfing a faceless man.

In another sketch, he's introduced a couple to stop our gaze from being distracted by the roaring perspective, and to draw our attention back to the guy on the right. Caillebotte paints the couple's heads at the same level, suggesting a whispered conversation regarding the other man's intentions. Perhaps this is the moment before a robbery. 

He's instantly created narrative, and established that human interaction (or the lack of it) is precisely what the painting's about.

Comparison between sketch (left) and final painting (right)

Comparing the compositions of the sketch and the final painting (above), the woman in the sketch (left) looks like a basketball player. She's massive. With her long dress, she appears to be even taller than him. This clearly wouldn't do for a gentleman to be dating a giraffe in 19th Century Paris.

But Caillebotte obviously wanted to keep those two heads close together and aligned with the horizon line in the final painting. Why? Maybe it's simply a compositional choice he made about the connection between their heads being more important than the distance between their steps.

In order to do that, and to remain true to linear perspective, he had to position the woman a few steps behind our protagonist so that she didn't appear to be as tall as him. 

Maybe the painting is simply about a shared moment of intimacy between them. In the close up, he appears to be leaning back as if to hear what she's saying. She seems to look at him adoringly. We're also aware that there are other figures on the bridge. The tight bunching of heads along the horizon line emphasizes the isolation of the figure in the shadows.

Suddenly, we've got it all wrong. Caillebotte is not the man walking with the woman; he's the man standing by the bridge. Isolated by the shadows and alone with head in hands, it's Caillebotte the sad loner.

"Modern life doesn't create close relations between human beings," Garnot says. "You are [in] complete loneliness in these new buildings, new avenues, new boulevards. There's something quite sad about that."

We'll never know the intended meaning of the painting. Caillebotte was known in his time as a collector of paintings and not as a painter, his work only becoming popular as late as 1950. He started painting relatively late, at age 27, and only lived to 45. A catalog from his Musée d'Orsay show states that "since the artist is dead, no amount of documentation will ever recapture a complete reckoning of his view of the world, or understanding of him as an individual."

Exquisite Deception: The Trompe l'Oeil of William Harnett

William Harnett, The Old Violin

“A painting has been added to the Art Gallery, which has created a furor.

Visitors will need no guide post, they will find it by following the crowd.”

— Cincinnati newspapers, 1886

In the days before David Blaine, this was as close to public spectacles of mesmerizing sorcery as people could get. If anything, the newspapers understated the effect of Harnett's painting.

Paul Staiti, in his essay entitled "Con Artists: Harnett, Haberle, and Their American Accomplices," says that "Harnett's painting The Old Violin so agitated and attracted crowds at the thirteenth Cincinnati Industrial Exposition in 1886 that a policeman was detailed to stand beside the picture and prevent viewers from trying to take down its fiddle and bow."

Besides his technical wizardry, William Harnett was a recognized master of composition. Each element in The Old Violin was positioned to achieve a precise balance. An interesting exercise is to block out any element, such as the door handle, with our hand. It's immediately apparent that each element was deliberately composed so as to achieve balance and order, and that not one item is superfluous to the overall composition.

John F. Peto, Old Violin

Thanks to a widely distributed chromolithograph, The Old Violin would fast become an icon of American art, inspiring a group of illusionist painters (including John F. Peto) with his crisp, linear style.

William Harnett, Music and Good Luck

The shallow space created by filling the entire canvas with an impenetrable door was essential to the trompe l'oeil effect. Harnett presented the objects directly before the viewer, preventing the eye from moving beyond and into the work, in direct refutation of Alberti's "window pane." Alberti would famously bid us stare through the picture plane, at an illusory world of space behind the canvas. His picture plane is an imaginary window in the foreground, with all the action taking place behind this plane. This was the remarkable and revolutionary idea that sparked Renaissance painters to adopt linear perspective.


Harnett reverses this heirarchy of space. Instead, he has all the action appear to happen in front of the picture plane. Through technical sleight-of-hand, he pushes the picture plane itself backwards, and then projects objects forward so that they appear to occupy our space.

detail of torn shreds of sheet music, Alan Carroll

Harnett, the undisputed master of illusionism, made the violin and sheet music the central images of his painting. The music sheet is a favorite visual trope that recurs regularly in Harnett's paintings. I had the opportunity to try my hand at painting music sheets myself recently (above), and understand the challenges that it entails.

John F. Peto, For the Track

Other artists would try to copy Harnett's compositional arrangements, with varying success. John F. Peto's For the Track makes an interesting comparison. Compositionally inferior, it was also technically the weaker painting. Peto's crude brushstrokes make us constantly aware of the presence of the artist, ruining any illusory effect.

John F. Peto, For the Track (detail)

Unlike Harnett's structured, almost spare arrangement, Peto's cluttered canvas, with objects vying for our attention, confuses the viewer and appears chaotic. 

Details from Harnett's "The Old Violin"

Typical of the best examples of trompe l'oeil, there are virtually no signs of brushwork. Harnett developed his own custom tools and techniques to help create his details -  the hinges, news clipping, twine, and stamp. In the details from the National Gallery site (above), we can see from left to right:

• To create the newspaper clipping, Harnett first painted narrow blocks of thinned black paint over a white ground. With the dark paint still wet, he used a tiny pointed instrument to trace lines through it, revealing the white underlayer and creating a "typeface" that looked real to the viewer, but was in fact illegible.
• The hinges were produced in a fairly simple, two-step process. Harnett first added either sand or coarsely ground pigment to brown paint to make a rough underlayer. For the rusty highlights, he pulled a dry brush, touched with orange paint, over the brown surface.
• Harnett used bright highlights on the metal ring and a soft circular shadow on the door to lift the ring off the canvas.
• The thin twine loop holding the violin is true testament to Harnett's painstaking methods. It is only five inches long, and the artist tweaked hundreds of fine lines in the loop's wet paint with a needlelike tool, producing the coarse-textured appearance.
• For the cancelled stamp, Harnett first painted a light square with a serrated border, giving each tooth its own delicate highlights and shadow. He then applied a thin layer of brown paint, scraping it with a blunt stylus to make the crossed flags. Next, Harnett used a minute pointed tool to make the almost microscopic engraving lines. He painted the cancellation mark with black and then smudged the entire stamp with his finger. His fingerprint is still visible.

"William Harnett transformed still-life painting in America when he tipped the picture plane on end, hanging objects on a rough-hewn door rather than placing them on the customary table top. Harnett's pictorial innovations—the vertical orientation, his choice of tactile objects, and his painstaking trompe l'oeil techniques—made him the most famous still-life painter in the last quarter of the nineteenth century," and arguably the inventor of the style called American Trompe l'Oeil.

Deer and Wildfowl, Adolphe Braun, 1865

Of course, as with any invention, he didn't just pluck the idea from the ether. There were precedents. In the 1880s Harnett spent six years in Europe, studying old master still-life painting. In Munich, he was exposed to the photographs by Adolphe Braun. Braun's flat, textured pictures of dead game, hunting implements, and horns suspended against a textured wall would have a profound effect on Harnett's work.

While most famous for his vertical still lifes of wild game, musical instruments, and paper ephemera, William Harnett is credited with introducing another, more controversial, subject to still life: money. 

Other artists had included currency in their paintings, but Harnett was the first to focus solely on paper bills and coins, making them look so real that in 1886 he was arrested for counterfeiting. New York law officers seized Five Dollar Bill from the saloon where it hung and demanded that Harnett hand over other "counterfeit" paintings. After viewing the painting, the judge advised that "the development and exercise of a talent so capable of mischief should not be encouraged." Harnett never painted money again.

Imitation, John Haberle (detail)

Harnett stopped making trompe l'oeil images of money, but John Haberle ignored warnings to "stop painting greenbacks" and made it his specialty. He neatly side-stepped accusations of forgery by simply titling his painting "Imitation." 

Five Rules of Trompe l'Oeil

1. The first and most important rule for painters of trompe l'oeil is that the object represented must be depicted actual size.

2. The shallower the space depicted, the more realistic the illusion. The most effective trompe l'oeil has always been a simple sheet of paper pinned to a board. As soon as the artist plays with deeper space (e.g. the space occupied by a musical instrument), the trick is not as effective. Our brains know that if it were a real musical instrument, our stereoscopic binocular vision would provide us with depth cues, also called parallax. Since this is just a painting, these cues are not present and we know we are being conned.

3. Hide those brushstrokes! Historically, trompe l'oeil was critically derided as being overly technical. Said to be merely virtuoso displays, they were dismissed as shallow egotism. But that was exactly the point! If it's to pass as reality, then it must be as smooth as a baby's butt.

4. Don't paint humans. Famous examples to the contrary by Mantegna and Veronese notwithstanding, objects that would in reality move about, should not be included in traditional trompe l'oeil. There's nothing like a smiling or waving human, frozen for eternity in a dry-mouthed grimmace, to ruin the effect of trompe l'oeil.

5. Don't cut the outline of objects with the frame. Objects in trompe l'oeil paintings must have their entire outline visible within the frame.

There are probably more rules, but I can't think of them right now.

Why is trompe l'oeil so popular?

The attraction of trompe l'oeil is the thrill of that magical moment where reality is suspended. The longer the viewer can be held in that suspended state, the greater the pleasure upon realizing the deception.

John Ruskin wrote of trompe l'oeil, "The mind derives its pleasure, not from the contemplation of a truth, but from the discovery of falsehood. . . . The degree of pleasure depends on the degree of difference and the perfection of the resemblance, not on the nature of the thing resembled."

Ruskin was right, says Staiti, that the pleasure of experiencing a painting by Harnett or Haberle is in "an expanded moment of passage from suspecting a picture is a deception to knowing a picture is a deception." The critical point of trompe l'oeil, he writes, is that "it is irony, not truth or beauty, that is triumphant. Paradox rules."

Canaletto and Superman

Westminster Bridge, by Canaletto (1746)

Canaletto is known to have used Camera Ottica to mechanically assist in creating his cityscapes of Venice, but closer study reveals that that wasn't the end of the story. He manipulated the perspective in his scenes, sometimes changing the vantage point and giving a more flattened and distant view. Pushing the stationpoint back gives the viewer a more panoramic and flattened picture plane.

I call it Superman Perspective. It’s the flattened way things would look if you could see detail from a thousand miles away. In his “Westminster Bridge” painting, we get the sense that Canaletto must have used a high-powered telescope to see his subject. There’s very little evidence of linear perspective in the arches under the bridge. The closer you are, the more distortion will be apparent. In reality, you’d have to swing your head around to take in this whole scene, and the bridge would appear to balloon towards you in the center. Canaletto has made the artistic choice to straighten the bridge ‘unnaturally’.

It’s more impersonal, but it depends upon your goal. Maybe, in the end, the telescopic view speaks volumes about Canaletto’s personality. Perhaps he wanted to see the world from a distance, to rise above the cacophony of human interactions. For someone like Caravaggio, on the other hand, human drama was clearly the very stuff of life. He'd have us sit right at the table in Supper at Emmaus, involving us directly in the conversation.

These cubes drawn in two-point perspective (above) illustrate my point. In A, on the left, the viewer is positioned a thousand miles away from the cube, viewing it as if with Superman’s supervision. The red orthogonals (lines drawn along receding parallel lines to the vanishing point) of the cube stretch out to some infinitely distant theoretical point on either side when presumably they would converge into Vanishing Points 1 and 2. The further away from the object we are, the further away the vanishing points are.

They closer we get to the object (or more accurately the picture plane), the closer to the centre line those Vanishing Points come. In B, on the right, we are so close to the cube that it’s sides have become completely distorted and are towering over us. You can see that the vanishing points (VP1 and VP2) are so close to the center line that they are visible within the frame. This is known in architectural draughting as “accelerated perspective,” because it gives an exaggerated sense of spatial depth.

In both these photos of the same vase from the Vanderbilt mansion in Rhode Island, the horizon line is level with the lion’s head. However, there’s a difference: we can tell that the camera was closer to the vase in the photo on the right. How do we know? The orthogonals on the right photo recede to vanishing points closer to the center line. This creates the kind of distortion we saw in the cube illustration.

In a large mural of a niche, for example, we would want to paint the vase more as it appears on the left, regardless of the stationpoint of the viewer in actual space. Why? Too much distortion in a mural is ugly (that's just my opinion).  I'm not talking about anamorphism here (where the stationpoint strictly determines the construction of the paintings perspective), I'm talking about wall sized paintings in the style of the French landscape panoramas of Zuber et Cie. I'd argue that large murals are meant to be viewed from a distance, so our linear perspective should reflect that.