Portraits of a Marriage

John Haber
in New York City

Models of taste in the arts . . . must be composed in a language both dead and scholarly, . . . scholarly, so that it will have a grammar that is not subject to the whims of fashion but has its own unalterable rule.
     — Kant

Edwin Hall: The Arnolfini Betrothal: Medieval Marriage
and the Enigma of Jan van Eyck's Double Portrait

Craig Harbison: Jan van Eyck: The Play of Realism

Linda Seidel: Jan van Eyck's Arnolfini Portrait

Even within his own time, Jan van Eyck was the stuff of legend. Within a century, the legend might well have belonged to the most distant and allusive origins of art.


How does an artist become a legend? Part of it is coming from outside a culture, as a sign or as a savior. Sure enough, modern accounts of van Eyck so often begin in Renaissance Italy, with how he appeared to Giorgio Vasari. Borrowing terms from tales of the painter Appeles a millennium before, Vasari's Lives credits van Eyck with the invention of his craft. Jan van Eyck's Arnolfini and His Wife (National Gallery, London, c. 1434)He was the first to paint in oils, an alchemist, a mathematician. In his technique, he had mastered a science.

Modern accounts still evoke the mythic origins of art: like Zeuxis in antiquity, who lured birds to painted grapes, van Eyck's illusions fool the eye because they are at one with nature. "United with its frame," his panels offer something more than "the modern idea of . . . imaginary space." Like a "tangible piece of luminous matter," they "confront us with a reconstruction rather than a mere representation of the visible world."

Listen to those words again: "a reconstruction rather than a mere representation." The mythic has stepped ambivalently into this world, but not quite. A reconstruction is a redoubled construction, a part of nature, not just a copy. Conversely, it is a second, belated construction—a memorial and an artifice.

But whose memorial and whose artifice? This essay is going to ask how those questions apply to any interpretation of art. So here is the source of those words, and then here is my plan. All set?

The players . . .

The quotes are from another frame entirely, Erwin Panofsky's 1953 Early Netherlandish Painting. The fruit of a historian's long and distinguished career, it has framed debates about artistic meaning in the Renaissance—and ever since.

Art historians still turn to the final pages of Panofsky's chapter on van Eyck—based on a longer 1934 article in Burlington Magazine. Panofsky discusses a double portrait, now in London's National Gallery of Art, of Giovanni Arnolfini and his young bride, Jeanne (or Giovanna) Cenami. When I was just a boy, the New York Times art writer, John Canady, even drew on the wondrous ingenuity of Panofsky's interpretation to advertise a popular introduction to art. If that is what art historians do, one was asked to surmise, they must have access to magic.

A book about van Eyck thus walks straight into questions about artistic origins, boundaries, meaning, documentation, and truth. Three books have dared that within just a couple of years. Edwin Hall and Linda Seidel write just about the double portrait—Hall sternly and pretty badly, Seidel very well indeed and from a postmodern point of view. Meanwhile, Craig Harbison surveys van Eyck's entire output as portraiture.

. . . And the questions

All three take on Panofsky's famous essay and a changing art history. In the discipline today, artistic truth makes sense only as part of a broader social history. Each book therefore has to find a place between traditional art history and contemporary critical theory. I shall argue that each has to build fences and then sit on them, and I do not mean comfortably.

If van Eyck is all about murky origins, my strategy is to begin at the beginning. I want to fill in first some background about van Eyck, Panofsky's interpretation, and the changing practice of art history. Then I quickly contrast how the three books handle it all, before tackling each one. I find that an interesting debate over quite another artist, van Gogh, ran up against that fence I mentioned.

Finally, I ask if I can clamber down off the fence. Can I sort out once and for all portraiture's historical meaning and art history's history? The answer is no: art and art history always take interpretation, but interpretation takes art, and it has a history. In fact, I spot Panofsky up there on the fence long before anyone knew it existed. Clever guy.


Like oil paint in Italy, van Eyck's portrait made him an outsider. Working for the Duke of Burgundy, the northerner crossed borders all the time. Here his subject was another emissary, Arnolfini, a prosperous Italian merchant then in Bruges. So what is his story, and then what did Panofksy say and a newer art history reply?

The bare facts

A richly dressed young couple, hands joined and facing forward, stand in a small chamber, apparently a bedroom flooded with warmest sunlight. Their gentle gestures, frank bearing, and comfortable surroundings hold out a vision of intimacy. The symmetry of the composition echoes their frontal bearing. Its diagonals caress the Arnolfinis' outstretched arms.

The bond is like remembering someone dear, thanks to van Eyck's mastery of perspective and the surfaces of material things. The erect poses, along with their artist's extraordinary command of light and shadow, give the bond solemnity. So do the man's sandals, cast aside as commanded to Moses in Exodus 3:5: "put off thy shoes from off thy feet, for the place whereon thou standest is holy ground." Somehow, this painter has learned to free painting from old rituals. At the same time, he gives the new naturalism the air of a sacrament.

He died in 1441, among the first generation to work in a new manner and the new oil medium, although he may have illuminated manuscripts as well not long after the Limbourg brothers, only a few years before the Master of Catherine of Cleves and well before the Claude Master. His greatest successes came after settling in Bruges, in modern-day Belgium, where he first entered the service of the Duke of Burgundy. Still, not bad. The medium and illusion obsess artists even today.

His scenes, invariably interiors, appear illuminated by real sunlight, peopled by ordinary men and women, and furnished with accessories of mundane living. Some are portraits. In others, donors seem miraculously to enter the world of the Holy Family—or is it the other way around?

Actually van Eyck did not originate a style, not even if one imagines the Renaissance as a decisive break. Textbooks about northern Europe generally give that honor to Robert Campin, who also may have been the first to sign a painting. The actual development of oils in Italy had to take account of an entirely distinct naturalism in Venice anyhow.

To bring him down further from a point of origin, Jan admits that he learned his art from his older brother. (Or at least that is what he claims, in the inscription on one altarpiece.) Still, no two art historians seem to agree on attributing even one painting to Hubert van Eyck, rather than to Jan van Eyck or his greatest follower, Petrus Christus—much less other unknowns in his workshop. Think of the way cosmologists cannot agree on the time since the Big Bang. (Hey, I take seriously metaphors in science as in art.) Somehow, Hubert only adds to associations of Jan with the mysterious origins of nature and of art.

The critical heritage

Similarly, Panofsky's interpretation of the Arnolfini portrait has lost none of its authority. It is still the source. Okay, so it falls nearly 200 pages into his book. Okay, so it gets tucked at the end of a chapter, after a full chapter on symbolism and artistic meaning. This is the real thing.

Panofsky argues that the painting does not just portray something: it does something. The portrait exceeds representation, as if trumping it in a desperate game.

I return to the rules of the game later, but first its object. By recording a marriage, Panofsky argues, van Eyck witnessed it in the legal sense as well as the visual one. He renders the marriage valid as well as representing its validity. His witnessing combines the act of painting with what is painted.

Panofsky found signs of matrimony lurking everywhere. The couple joins hands, for one thing, and Giovanni raises his right hand as in a vow. The back wall holds some legal necessities, including a third party's mirrored presence as ceremonial witness. He is presumably the artist's only self-portrait, and I can verify that he resembles well enough the blur on a saint's armor in another painting. The artist plays his part in the world he recreates, and that world does not shy away from accidents of nature.

Just above, van Eyck's signature adopts the elegant Latin script appropriate to a formal document. The private ceremony gives every care to its public record. Such a ceremony would suit the Arnolfinis, Italians with needs poorly served by a northern church.

Unique among sacraments, by canon law marriage drew its validity from the consent of the participants, not from a priest who might be presiding. Painting, one may infer, is sanctified in very much the same way. Panofsky actually called his readings iconology or iconography—literally, the study or writing of signs charged with belief. Here that means detecting signs of matrimony. Icons are symbols within the art, but they also allow the art itself to remain an icon today, in a more secular age.

The new art history

The term iconography recasts images as a kind of sacred writing, to be read only with knowledge of a long-past culture and its rituals. It committed three generations to a special art-historical discipline. For them, every artistic influence and historical fact could have force.

For the same reasons, however, iconography also anticipated other approaches, in which interpretation resembles a chain of texts. Meanings result from the arbitrary rules of powerful forces, not universal truths. In no more than twenty years now, such meanings have come to life in a provocative wave of scholarship, sometimes called the "new" art history.

To catalog the "new" art history—or the surviving old one—would be a mammoth job. "New" includes postmodern theory, political limits on interpretation, art in a video age, and doubts about the old idea of authenticity or genius. On the "old" side, people like to imagine realism, historical accuracy, connoisseurship, and true interpretation.

The "new" history is sometimes said to have begun with T. J. Clark's early essays on Impressionism, culminating in The Painting of Modern Life, or with feminist criticism, such as Linda Nochlin's Women, Art, and Power and Griselda Pollock's Vision & Difference. Svetlana Alpers's look at representation in The Art of Describing has also served many as a model. (I have source notes on all these at the end, as well as a postscript sixteen years later on how this page came to be.)

Early examples include Roland Barthes's Image, Music, Text and Julia Kristeva's psychoanalytic readings of female imagery. Absorption and Theatricality, by Michael Fried, has a place here too often overlooked. It marked an art critic's leap 150 years into the past, and it inverted the very valuations that he had used to defend high-modernist formalism against Minimalists. For the young Fried, heavily influenced by Clement Greenberg, Minimalism spoiled Modernism by turning the gallery into theater. Now, however, he saw that theatricality, starting with J. B. Greuze and the Enlightenment, opens onto a code of meaning in past art that Modernism only repressed. Something was in the wind.

A few problems left over

The twin legacies of iconography—good old historical research and a newer textual criticism—continue to face off. For most people, an academic discipline is easy to ignore. Still, if either side is right, then who is in control matters, and not just to them. It also determines what any imagery conveys to anyone. On it hinges as well the social status of artists and women today.

Another source of conflict is sheer style, and surely style ought to matter in discussing art. The "new" historians veer from cold outrage, most often directed at high culture and the wealth behind it, to deceptively meandering personal confessions about life and the movies. The former is necessarily evoked by the felt weight of a racist, sexist past; the latter is commanded by the subjectivity of meaning. Either must seem plain undignified to historians trained in Panofsky's magisterial studies.

The dilemmas existed before in art. In fact, van Eyck could well be said to have invented them. The symbols that attend the Arnolfini wedding developed in altarpieces for mass devotion, maybe the ultimate popular art form. In mapping them onto a domestic interior, within a panel destined for a member of a foreign elite, the painter helped make art institutions the puzzle that they remain today.

What makes the dilemmas so pertinent, I think, is the institutional success of late modern art. It is natural for art today to prove itself to universities and museums. It is natural for theorists to be critics of contemporary art. A historian's puzzles are just those of artists, who are never just struggling with the past like a dead weight. They also bring it alive again to experience its crises and truths again—and with a difference.


The three books I shall review helped me see the questions more sharply—or maybe I should say "fuzzily." I had better juxtapose the three before giving Hall, Harbison, and Seidel each in turn some care.

The three books

Edwin Hall, in a whole book devoted to the one painting, argues that Panofsky's interpretation violates plain facts. Panofsky had used the term "clandestine marriage" to describe the Arnolfini wedding in a home. Hall counters that an Italian dignitary surely would not act clandestinely, as shamefully as an eloping couple. His stature would demand an ostentatious ceremony.

Hall summarily eliminates symbols as simply natural presences. A real wedding bed was sure to lie in an upstairs chamber; this bed just piles on more of a wealthy man's property. Dogs are dogs, and reflections in the ornate mirror occur naturally. Arnolfini, Hall concludes, follows up on a wedding rather than conducts it—and with excellent reasons. The Italian is out to show what a good catch he is.

Craig Harbison, not unlike Hall, reconstructs the character of those portrayed. His book finds portraiture lurking in van Eyck's entire career. In so many paintings, clearly secular men, warts and all, pose in a religious context. It could be a wedding, an adoration of the Madonna, whatever.

Harbison's book coheres badly. The short chapters, sometimes several to a painting, tend to start from scratch, as if for separate scholarly publication. Too often, I felt that I was reading discrete vignettes by a needlessly obscure specialist. On rereading, however, Harbison's theme emerges more clearly: each sitter has commissioned his painting to express himself. Each offers up his secular nature and his private devotion, as best as human failings permit.

Harbison includes just one chapter on the Arnolfini painting, and he has no quarrel with Panofsky's method. Yet he erodes its validity subtly in his own way. Panofsky had subordinated this world and the painted one alike to a supersensible one, a culture that gives meaning to both. Harbison wants to free people, like the lovely couple in London, locked in religious narratives.

My last writer, Linda Seidel, is clear and wide-ranging, easily the best of the three at hand. Yet she patiently eats away at Panofsky from another direction. She chooses not to insist on the historian's priority as judge of the facts. Rather, she draws on criticism and contemporary philosophy, especially on post-structuralism, to ask what an act of interpretation—any interpretation—involves.

Seidel covers so much ground that her focus is very hard to remember. One feels that she is using deconstruction to hedge her bets about her discipline. Every new argument deepens the painting, because any perspective can vary and enrich a viewer's potential experience. A thesis gradually emerges, but it never quite acknowledges dissonant inferences along the way.

Each writer is begging for criticism. Seidel, one might easily say, lacks a coherent point of view. Or at least she lacks the courage to acknowledge the limits of coherence. Harbison indulges in downright trivia. Hall struck me as simplistic, not to say just plain obnoxious.

What intrigues me more, however, is what they share. All three are caught between the art historian's meticulous practice and the demands of trendy theory. No doubt professors of art history and their students today churn out radical or deconstructive "readings" with all the heavy, polished machinery of academia. Yet these books attest to an uneasy vitality in a doggedly conservative field, an unrest that could equally well be glamorized as a crisis or even a death.

Different ideas, right?

For each of the three, the conundrum ultimately leads to a traditional interpretation, in which the art stands in for someone else's intentions. In Hall's blistering account, the painting must be truthful to hard facts, to its historicity. It must be truthful, that is, to the remembered presence of a great artist. To Harbison, the painting must be true to its subject, the remembered presence of the person portrayed. For Seidel, the painting must be truthful to its multiplicity of meaning now as then—to the remembered presence of the viewer.

Perhaps surprisingly, each discusses art as identical to its content and characters. More even than original intentions perhaps, a painting stands for someone's very humanity. Can this really have been the century of formalism? Can it really have been van Eyck's own oil medium that led to art for the sake of form, sensation, and materials?

Each in a different way suggests limits to historical assurance. Each puzzles out the claims to truth of iconography. Yet each also falls prey to the same stern demand for a truth in painting. For each, as Seidel puts it courteously and fatalistically, Panofsky has "forever fixed the context within which the work would be received." Yet each sidesteps direct discussion of the older man's methodology.

First, however, the differences. What kinds of truths enter into each?

Hall: the second-best bed

I have to admit it. I hated Edwin Hall's book. The play on clandestine typifies its insensitivity to nuance. Panofsky plainly meant private, surely only needful to a powerful Italian concerned to uphold his religious traditions and thus the high honor of his heritage. At best, an association with impulsive teenagers obscures matters.

Think what a modern, risqué usage of clandestine overlooks: Panofsky was describing a historically sanctioned practice. If it was also a controversial practice, then van Eyck knowingly walked into the thick of controversy. With testimony comes a kind of authority, and the Church could certainly aspire to witness every marriage, but nothing more. Not long after, Rogier van der Weyden even painted that aspiration. In his Seven Sacraments, the scene of marriage pointedly recasts the van Eyck portrait as church ritual.

Ironically after the historical slippage, Hall soon goes on to wield historical accuracy like a club, as if to settle a score. After softening the reader up with his defense of Arnolfini's dignity, he picks up the pace. He is out to knock all those symbols down to size.

Hall never confronts directly the notion of iconography. He never has to: he sticks to the plain facts. If something could have been present naturally, it hardly bears explaining. Or does it?

In an earlier chapter, Panofsky brilliantly termed van Eyck an archaeologist. The painter, he points out, reconstructed rather than represented a world. As one can easily forget, van Eyck's architectural details belong to the style of churches of a century or more earlier. With his own brand of historical respect, van Eyck meticulously constructed a space, and who is to say what belongs there? Only, I think, someone able to guess at a complex interplay of context and character.

Hall's fact and fiction

For Panofsky the painter's triumph was to invest new facts with old meanings. Surely van Eyck's faith was a faith in this world, a trust that would eventually lead to Hall's more modern notion of realism, but it was also a creation, a fiction. The facts might not be so plain after all. Bedroom furniture might well not appear in a downstairs sitting room, as Hall notes, but one's associations with bed would still remain.

Then too, maybe the Arnolfinis' bedroom furniture is more real than the sitting room in which it seems to appear. To divide it up more fairly, imagine a contemporary interior. A yuppie might ask to pose near the window, so that his portrait gives due pride to the room's southern exposure. He might also ask that his favorite objet d'art appear by his shoulder, and there would be no use mistaking it for the keepsake "really" over by the window.

If one is to let the facts speak for themselves, they may not have much to say. An artist—or even an art historian—may communicate much more by deceit. By much the same token, Hall would forbid deciphering Rembrandt's characters from their portraits. After all, everyone has eyes, so what else is there to say about them?

Hall's strength is to insist on one social function of both art and marriage, the furtherance of tangible property across families and generations. His weakness is to deny the very means by which art and culture wield their power. Indeed, in an obvious sense, art from its beginning has functioned simultaneously and ambiguously as representation and as what J. L. Austin called a performative. A cave painting may have had a role in rituals at which one can only guess, and yet the hunt would have continued without them. A court portrait legitimized power, and yet the sitter was a head of state before it. A purchase today certifies the taste and wealth of a collector, and yet the collector is claiming the taste and wealth to have made the purchase.

Basically, Hall believes that fiction and fact are logically exclusive. As a result, he can no longer properly comprehend either one. Strangely, in denying that van Eyck served here as visible presence at a wedding, he goes out of his way to trust the artist's historical fiction. While giving in to the contemporary critical concern with power and male privilege, he clings all the harder to the historian's traditional criteria for truth.

Harbison: members of the wedding

If the artist's moment in time fades so easily into illusion, who can these people be who even now seem so vital? Panofsky first wrote amid an art-historical fashion for identifying sitters. When he brilliantly speculated about van Eyck's older brother Hubert, he similarly evoked a lost world. He also showed his determination to sort out the works of the masters. The iconographer proved he could be the equal of an older generation of historians, connoisseurs such as Bernard Berenson.

Yet Panofsky did not look primarily to archival research. Rather, painting itself became the document that one must learn to unfold. Out of dead artistic styles, the painter helped to create a vocabulary of signs, so that he could articulate his place in a religious and secular world.

Like Panofsky but more single-mindedly, Craig Harbison asks what sorts of people these could have been. "One purpose of this book can be stated simply: to evoke the human dimension." Such an innocent-sounding formula may sound more cuddly than Hall's outlook, but it too stands for stories of power:

This artist's religious images, even rather anonymous looking ones . . . reflect a patron's wishes, either by promoting his or her actual position and sensibilities, or by appropriating sensibilities admired in someone else. Issues of gender, social class or clerical perception are embedded.

Despite the concession to propriety in "his or her," Harbison knows quite well that power lies with the man. Arnolfini "is given the controlling gesture: he calls forth his wife's fecundity." In other paintings of van Eyck's, clerics such as Canon van del Pael or Chancellor Rolin may kneel before the Virgin Mary, but they barely repress a scowl. They just insist that the cityscape behind them records parish property and official sanctions.

Does Harbison's bow to class and gender issues already seem too familiar? His ingenuity lies in showing how hard men like Arnolfini had to work at their privileges. A market for propagandistic images in fact arose because the world was being turned upside down. The clerics were reacting to the Great Schism, while Arnolfini's plea for offspring could have shielded him from a nasty lawsuit some time later: "Barrenness was often given as the reason for the husband's unfaithfulness."

These men celebrated at once ritual and property because they no longer knew in which their future lay. As men used to crossing borders on official missions, both van Eyck and Arnolfini must have had to face such negotiations in value all too often. The wedding couple had something definite on their minds:

They did not select an ecclesiastical setting for their portrait; and I do not think they allowed a simple or single-minded religious attitude to rule their image, or transform their lives into a theological exercise. For this couple, and for growing numbers like them, life must have been about enjoying the world, about coping with its demands while furthering one's ambitions and taking what steps seemed necessary to ensure personal salvation.

Harbison documents the ambitions, fears, and courtly tastes of character after character. "This very pretension . . . reveals an underlying doubt." Even a detour to discuss Christ's foreskin reminds one of the growing humanism's new edge.

Harbison and some all-too-human purposes

However, if the painting stands as a rearguard maneuver, it no longer can be said to depict so simply its people and its time. Propaganda, my dictionary tells me, has a lot to do with lying. The closer one looks to the paintings for the figures they portray and the stories they tell, the less compatible those two goals appear. Paradoxically, the paintings can defend traditional hierarchies only by "mingling worlds that we might feel should be distinguished: sacred and secular, public and private, male and female, clerical and lay."

I had the feeling that Harbison was protesting the plainness of his enterprise a little too much. Words like simply and bluntly appear disconcertingly often. Is that why he distinguishes his interpretive discipline from "some writers," who "seem to have been wary of contaminating the art with minute biographical data"? Unlike Hall or Panofsky, he never peers within the famous mirror.

His distance from scholarly minutiae would be greater did he not take a block of loosely connected chapters to describe a single sitter. He would be a lot more fun, too. I have to admit: he is curiously boring for a humanist, perhaps why such a worthy book languished in print for some years without an American publisher. Conversely, his sophistication obliges frequent disclaimers of his own simplicity, not unlike the treacherous "his or her" that I quoted before:

While I would not advocate an approach determined by personal whimsy or self-projection, the analysis of an art as personal as is van Eyck's must make some allowance, within the appropriate historical framework, for the personal—in the sense of the intangible and of the imaginative.

The complexity of the times, as he sees it, demands an imaginative projection, even "allowance" for some human perplexity.

And who are Harbison's people?

I had to wonder, however, whether van Eyck would have been as thoroughly confused. As Seidel observes, "market matters, social practice, and political relationships were all intimately entangled with Church customs." More important still, they all throve on the tangle. Did these sitters really feel so torn between God and Mammon? Did newlyweds back then really need the threat of lawsuits to want children?

Long-standing criticism of greedy officials would one day trigger the Reformation. That same new humanism helped to create van Eyck's representations. Yet the Church's success, as a proud secular authority with traditional values, also drove the Renaissance. Any association of a burgeoning art with a culture of pessimism sounds suspiciously modern.

In an earlier article, Harbison discussed the painting's original frame, which may well have included sexually charged lines from Ovid. On another panel, possibly even a companion panel to the Arnolfini portrait, van Eyck may have painted a nude woman—and she could be the same woman who stands so demurely for her wedding portrait. (I am grateful here to Seidel, whose book introduced me to speculation about a more worldly Jan van Eyck.) A copy of the nude survives; it also peeks out, centuries later, from within a Dutch painting of a gallery interior.

Erotic poetry or art might well have underscored promises of love and children. Harbison preferred in his article to see an ironic commentary on the solemn vows. He does not repeat, or even mention, his argument now, perhaps out of deference to scholarly caution. Yet his modernist association of art with irony remains an unexamined critical assumption.

So who in the end are these people? What sorts of individuals can speak so directly to the living while anxiously bearing the contradictions of their age? At what point do the contradictions undermine traditional faith in the unity of character? When do the sordid traces of materiality and text—mirror, surface, and frame—destroy traditional beliefs in painterly representation?

Harbison is caught between his audiences—popular and scholarly, cultural and modernist, art historical and philosophical. And he never can quite reconcile or integrate his approaches. Realism "was meant to have a negotiating or exchange value, allowing a dialogue to take place between artist, patron, work, world, and, finally, the spectator today." But exactly who is each speaker—and how real? All one may ever know for sure about Giovanni Arnolfini, delicately holding up his enormous headgear, is that he first made his fortune in hats.

Seidel: stories of another master

Charmingly, Linda Seidel banks on the speakers she knows best—herself and other historians. She constructs her book as series of lively but loosely connected "stories of an icon," or tales of the painting's varied modern reception. And all her stories come down to a dialogue with a rather different sort of icon, Erwin Panofsky.

In an epilogue she notes how closely she identifies with him and, indirectly, with another exile, Arnolfini. Panofsky's 1934 paper was his first in his beautifully nuanced English after leaving Nazi Germany. He held thereafter in print to his adopted language, even while returning home to support Jews and academic independence. He also self-consciously assumed the mantle of Max J. Friedländer, a Jew and the author of his own definitive Early Netherlandish Painting. Seidel's husband was likewise a refugee from Nazism.

Panofsky's felt isolation, she argues, led him to overstate the Arnolfinis'. Other citizens of Lucca, Italy, did live in Bruges and could have borne witness in a traditional church ceremony. Moreover, if marriage required only the consent of bride and groom, a painting need not record and enforce the promises. Besides, there was still another witness, a mysterious fourth figure who stands alongside the artist's reflection in the mirror.

For Seidel as for Panofsky, the portrait serves as a witness—but not a legally necessary witness. The painting supplements the claims of a public ceremony. It tells us what marriage, love, painting, and even witnessing can mean. And it holds out that meaning for someone far from the source, like Arnolfini or Panofsky so far from home.

Whatever the artist may once have documented, then, the only necessary eyes today are the viewer's own. But what do they see? They see above all the social functioning of a marriage. Seidel shows how images work to authenticate a woman's virtue, her dowry, and her acceptance of a husband's wishes. She finds, for example, other paintings with eager witnesses to a pregnant woman—religious scenes of the Virgin. As a mother, Giovanna will have a tough act to follow.

Then too, note that Giovanni alone raises his hand to swear fidelity. A woman's role was inactive, even "in the most scrupulously transacted marriages":

[Women] were represented by their fathers throughout, even misrepresented by them, until their husbands took over. As "bartered" brides, women were tokens of significant financial exchange between the men who governed their lives.

Giovanna meant money, like the color of her dress, at least if one can accept Seidel's association between green cloth and the money-changers' guild. (Yes, outside of Panofsky's evocation of an actual room, the layering of hidden symbols can get rather showy.)

Somehow, I am right back where I started: the portrait represents something rather than enacts it, something more tangible than sacramental. So is Hall right after all? (Even his aversion to hidden symbols must look awfully welcome just now.) In fact, Seidel has already moved quite a distance from Hall's interpretation.

So far at least, her conclusions may look like Hall's, but they depend on broader and, happily, more slippery evidence. As for Panofsky, the scene plays out against a background of religious images and a world of meaningful objects. The portrait, she argues in proper critical jargon, "interweaves the discourses of social practice and painted imagery." Further, van Eyck worked "for a private patron . . . in relation to a familiar, 'public' body of imagery."

She is interested, she says, not in the symbols, but in "how they became activated." Once one starts to tell tales, it is hard to stop. "Stories inevitably structure their own endings," and they may not come to an end in fifteenth-century Bruges:

Detailed familiarity with that time and place is essential, to my mind, for comprehension of the ways in which Jan's art functioned and would have been understood; still, the painting responds to other inquiries and has accrued other histories, each one of which resituates it in, as it is resituated by, our own experience.

Seidel's loaded experience

She follows two inquiries especially. First, she finds nuances in the relation of these young lovers. The man takes possession, but only by entering the richly imagined woman's world. The domestic interior even includes a cute little dog, not to mention the cast-off shoes. I imagine Giovanna demanding he take off those muddy shoes before he ruined perfectly good furniture.

Giovanni does not relinquish all his pride of place. He still bars access to a man's world, unseen beyond the walls at left. The bride and groom meet as opposites rather than as equals, much as they join hands but not the same hand. (Panofsky explained the frontal stance as a matter of sheer painterly symmetry, appropriate to the work's gravity.) They "have come from opposite ends of the space: he from outside, she from within, to meet, for all eternity it seems, at the center of the panel."

Seidel's second line of inquiry concerns the witnesses, including the painting, painter, and viewer. The mirror, evoking the original frame, accentuates the work's materiality and unity, much like the convex mirror in a diptych of the Virgin and a donor by Hans Memling. The portrait "resembles an official record . . . more closely than even Panofsky proposed."

At the same time, painting reflects on its act of witness. The mirror defines a vanishing point and places the viewer within it:

We are frozen in our tracks by the man and women who have been positioned by the artist uniquely for us to observe. The couple address their actions and their glances toward us. . . . We sense that we cannot pass easily from their midst without disrupting their performance.

Encompassing and surpassing a written document, the portrait proudly states the greater truth of images and the authority of art. If the entirety of the work consisted as well of erotic verses and a nude, it playfully reasserted the man's prerogative in marriage. The dog at the couple's feet might serve as a symbol of the sexual free play pertinent to the joys of marriage. (Remember Harbison's suspicions of infidelity?)

Seidel's seeming digressions turn out to meet two centuries after the Arnolfini wedding, in the hands of the greatest master of Baroque Spain. In Las Meninas, Diego Velázquez stands poised at his easel, while a mirror on the rear wall reflects the royal couple. Artist on canvas, couple in the mirror, the viewer who knows where: the enigmatic relations between them appear to reverse the terms of van Eyck's double portrait.

In other works—influenced by Titian, the great Renaissance Venetian—Velázquez uses fables from Ovid to assert the claims of his art. Once again, images within images link masculine pleasures to perplexities of looking. Once again, they pose the status of painting among the arts:

Disorder is the inherent underside of seamless presentation. Ovid's texts . . . force us to contemplate the dark side of amorous relations between the sexes, the sometimes unspeakable advantages men take with the women they marry and those whom they more transiently encounter.

And once again, painting both perpetuates old privileges and exceeds them. It reaches back to the very first fables of art, like Panofsky and then Seidel reaching across generations.

But what is Seidel's history talking about?

Has anyone followed me this far? I had my work cut out for me reducing her virtuosity to a story line. Seidel takes on a lot of intellectual history. Despite it all, her book is art writing at its best. She somehow manages to stay sharp, relaxed, and entertaining. Even my criticism hopes mostly to pin down what I have enjoyed so much.

Why then is it so hard to remember what these chapters are about? She never intended a seamless argument: she means to keep her "stories of an icon" distinct, and that suits me fine. The trouble is not that the seams are showing—but that she really could care less what the seams look like. The different stories neither quite confront one another nor quite give up their claims to complete, tidy interpretations.

Consider that innocent-looking dog and those filthy, cast-off shoes. They function in turn as Panofsky's symbols of fidelity and mutual respect, as signs that the man has entered a woman's world, and as reminders of his sexual opportunism. I buy them all; I have myself studied how men enter the world of Jan Vermeer's women or Shakespeare's Juliet. (In the first, I have found traces of van Eyck's women as well.) And yet the book's movement toward a critical happy ending left me uneasy.

Are any of these meanings unsettled? Can their conjunction help pin down a progressive artist's involvement in the patriarchy? They sneak in and out way too comfortably, like a gloss out of Oscar Wilde: "And certainly once a man begins to neglect his domestic duties he becomes painfully effeminate, don't you think?"

Or look at how the chain of references to Ovid unfolds. Perhaps van Eyck cited Ovid in a lost frame; Velázquez cites van Eyck's placement of mirror and artist within the frame; Titian's classical nudes and allegories also influenced Velázquez; therefore, Titian's myths extend to van Eyck. Hmm. Post Eyck, propter Eyck?

Over and over, Seidel restates her allegiance to the "new" art history, in which "meaning is neither found nor given, but . . . takes shape arbitrarily." It explains some less than legitimate moves between fiction as act of the imagination and fiction as lying. It accounts for the book's epilogue on herself, her husband, and the Nazis.

But I wonder. I missed the passion of Marxists and feminists, for whom class and gender issues reveal nagging anxieties. I also missed the challenge of deconstruction, which allows conflicting meanings to tug at the heart of meaning. Even that epilogue comes off as a little too conscious an effort to be postmodern.

She is out to pull the painting's frame—its physical frame and its social context—within it, because she cannot yet abandon the painting's complex unity of intention. She adds speakers, rather than give up the identification of meaning with speech. She assigns the dog and shoes to successive owners—a couple, a woman, a man—rather than question the notions of ownership and truth. A new historian might note that the New Testament quotes the command to Moses, in Acts 7:33, but to accuse the people of turning against the prophets. And the preacher, Stephen, is then martyred for, in effect, an act of appropriation.

I guess her heart is with the "new" history, but her head is with the traditionalists. Early on, she notes that interpretations have long been "archival" or "hermeneutic." Like Hall, hordes of scholars have dug up the facts; or else, like Harbison, they have revealed dead voices behind art. She does not wish, she says, to choose between them or to abandon them.


My three writers all testify to a scholar's unease. They point to what others must face every day. Older historians wish to make grand theories serve craft—their own care and the painter's art of communicating meaning. Younger critics bring their invaluable immersion in philosophy and its creative jargons.

Historians shrink from theory. It might put what they do in question. Or at least it might make their assumption of historical facts look naive. Theorists hesitate to pin down a historical context. It might make talk of philosophical foundations look naive.

Both sides remain committed to interpretation. And that means they have to judge the truth of painting with the eyes of someone who may no longer be around to deny it. It could be the painter, the patron, or some later owner—even centuries later. Pretty tough situation.

Derrida on art history's rigidity

Jacques Derrida anticipated these issues, in a book indeed called The Truth in Painting. Starting in lectures in the early 1970s, the French philosopher noted how "rendering" (in both French and English) means both returning to its owner and representing. Panofsky's study implies much the same senses. After all, representation puts the sitter in control of his bedroom.

In the historian's appeal to the facts, Derrida too often detects recourse to a human stand-in for a painting's meanings. The painting acts suspiciously as a "natural" home for the elusive objects that the painter has so disconcertingly strewn about, like Arnolfini's muddy shoes. The shoes may make a pair, and they may belong to Arnolfini, but the painting does not say. Just as Vincent van Gogh does not reveal the owner of the shoes whose life his painterly gestures may seem to uncover, van Eyck has the right to remain silent.

Derrida cites a letter of Cézanne that promises la verité en peinture—to paint truthfully, or maybe to paint the truth, or to speak truthfully about his art, or to paint in fidelity to the medium. In this, Derrida sees the painter's dedication to opposing truths. Meaning, communication, and historical context threaten to collide. So too do the difficult promises that these unleash.

A promise, he notes, is what a philosopher, J. L. Austin, had called a performative—a statement that does something rather than pronounces fact. Every painting wishes to be an act much like the Arnolfinis' wedding vow, a promise of certain meaning within a fixed historical frame that no art can ever have. A work of art resembles a game or a hypothesis.

For an example, Derrida looks at Meyer Schapiro, a vigorous defender of modern art who was writing about van Gogh's painting of two shoes. Schapiro was criticizing a sentimental essay from Martin Heidegger, someone else with a past tied up with Nazi Germany. Derrida thinks the art historian used facts rigidly to settle scores. He says that art historians make a game out of restoring a painting to its owner—which might be the painter, the viewer, or the subject. He could easily be speaking about Hall, if not indeed about all three books under review. He could be speaking for art today that tries mirroring the Holocaust.

But I wonder. If entrapment between truths and between fictions is inevitable, perhaps it is the most fruitful place to be. As I hope I have suggested, I do not wish to be without Seidel's fictions, much less van Eyck's.

There is no one true witness to the Arnolfini wedding. But is it so easy—or even worth the trouble—to abandon the search for witnesses? If a painting is a reconstruction, its artist a visionary, its images a turning back of vision, and all its witnesses a fiction, it attests to the human need to remember. Like a photo album or a maiden name, it does not recover the past, no more than it can put the past at a safe distance.

Derrida also overstates Schapiro's rigidity. (Perhaps he means to: for him, criticism is a kind of appreciation.) Schapiro looked for traces of Vincent van Gogh and Jackson Pollock in their subject matter, shoes and drips. However, he does not stop at identifying art with the artist. In mostly later essays, he also looked for biographical evidence of interpreters such as Berenson in their work. Once characters start coming in from the wings, the play is just getting rolling.

Recovering disbelief

I do not mean to revive a simple belief in many voices; I have already criticized such faith in Seidel. Mikhail Bakhtin movingly evoked the "dialogic imagination," or contrapuntal voices of the characters in a novel. And Craig Owens, a late critic of contemporary art, tried to answer Derrida and Schapiro by passing art's authority on to its viewers. No doubt, but they only continue the game of looking for authoritative speakers in a different universe of fictions.

I mean a kind of productive disbelief. An audience, like that of Owens or Seidel, may be no more or less real than the artist identified by Schapiro or Hall—or than the subjects identified by Harbison or Bakhtin. To hold meaning, each must derive from and stand in relationship to the work and its history. Some critics, artists, and subjects are just better than others.

Viewer, artist, and sitter come back to resume their dialogue, no matter how often they are declared a fiction. To call their balancing act a fiction is to compliment it anyway, like the fictional balance in Vermeer's Woman Weighing Pearls, where the density of light, the obvious symbolism, the reticent emotion, and the elusiveness of meaning all remind me so much of van Eyck.

When the painter places words into a frame or a mirror into a frame, it equally evokes the powerful instability of the "I," what linguists call a "shifter." That strategy has separate, meaningful contexts in van Eyck's time and in ours. As Schapiro once documented, the illuminated manuscripts that preceded van Eyck throughout Europe, such as Le Livre de la Chasse, first made the viewer's I so dangerous an eye. The crisis in art history is still very much the place to be.

My doubling of historical approaches is like van Eyck's doubling of witnesses. The added person in the mirror, distinct from the artist's image, is and is not the viewer now.

Panofsky: the rules of the game

Remarkably, Panofsky was there before. One does him an injustice by imagining that his iconography stood so simply for a single speaker. Let me trace his interpretation's power in more detail. I cannot dream of copying Panofsky's eloquence and insight, but I can hope to follow his moves in the game that he created. As they shuttle amazingly between all the players, they take one back to van Eyck's "atmosphere of mystery."

First comes that magical technique of reconstructing nature minutely, like "infinitesimal calculus." In Panofsky's next move, the very excess over ordinary technique returns van Eyck to representation. Because the painter has no need to select from reality, he attains "description" rather than "interpretation."

What he describes, Panofsky argues, is a wedding. One recognizes formal ritual in the painting's symmetry. One sees it in the couple's joined hands or Giovanni's right hand, raised in a gesture of avowal. The painting's diagonals carry the eye along their outstretched arms.

Now, however, painterly interpretation returns with a vengeance. Symbols accumulate to create a unity of meaning, much as minutiae create a unity of vision. Marriage is sanctified, as a ritual, but also by its promises of fidelity and fruitfulness. The dog is a reminder of faithfulness; so is the praying woman carved above the bedstead. The bride's hand resting on her womb and the bed go together.

The single marriage candle, burning in rich daylight, is another part of the ritual, while apples by the window and a gargoyle above the joined hands evoke the human sin that the ritual must overcome. Remember the twentieth-century parody in which Giovanna's appearance of pregnancy turned this into a shotgun wedding? The Arnolfinis did not need recourse to verses on a physical frame to accept sex. Nor, however, could van Eyck let mere human desires altogether unsettle a couple's prospects.

Just when things start to line up too neatly, Panofsky turns the tricks against themselves. The symbols include signs of witness, and surely every marriage needs one. The painting now turns back again to representation, to vision, but now with the artist's vision part of the work. Here is where mirror unites the artist's image, eye, signature, and authority in a promise.

With that move, however, the fixed point of view defined by the mirror can no longer hold, just as the actual viewer can no longer match the person reflected. As promissory note, the painting now opens outward toward the world of objects. The "atmosphere of mystery and solemnity . . . takes tangible form."

Panofsky finally is left with a "concord," but without the fixed borders that formal unity would demand:

In the London Arnolfini portrait, then, Jan van Eyck not only achieved a concord of form, space, light and color which even he was never to surpass, but also demonstrated how the principle of disguised symbolism could abolish borderline between "portraiture" and "narrative," between "profane" and "sacred" art.

Iconography and art could have abolished borderlines—or be living on them.

In and out of the picture

Like the three recent books, Panofsky's interpretation grows increasingly contradictory and dizzying the more one respects it. I would like them to share historical care, fluidity of vision, and deeply felt outrage at the class and gender biases they uncover. On the whole, as I say, Seidel comes by far the closest, simply because the post-structuralism that she emulates is most useful in situations when there is no sensible answer.

I went in search of a radical break in art history. As I hope one may see, I found instead a discipline climbing all over itself to sit on a fence. Of course, that would not be possible if the boundaries were really so evident as one often supposes, if a painting could be placed securely in its frame once and for all. In practice, I felt the claims of each side contaminating the other.

The metaphor suggests clumsiness, contradiction, and hypocrisy. I have tried to analyze them all, but not only that. I am arguing that a fence might well be the nicest place to sit after all, and not just for now. Maybe everyone is climbing up to catch a better view!

To paraphrase Derrida again later in the same book, the trick is not exactly to reinsert art into its frame in search of a complete, larger work. Nor is it to keep pretending that the frame is not there. Seidel, or Harbison on good days, tries the former; Hall goes along with the latter. There will always, however, be more.

The physical frame belongs in and out of the work's content. So, however, do the whole original hanging, a broader historical context, and the seemingly trivial decorative elements within the art. Each disturb and enhance meaning; each create new people to imagine them. They are always there, alongside things, neither fully within or without.

I identify with the artist and the feminist critic, but only by turning the artist into a fictive viewer and the feminist into a man. And what does that say about who I am?

A postscript: how this review came to be

In January 2012, reviewing Andrew Moore, I tried a word not only in self-defense, but also in defense of criticism. Words need not stand in the way of art, but rather open new loves, new doubts, and new understandings. And of course I meant a specific model of criticism, defended here often before—one that refuses to disentangle theory and practice, description and interpretation. It can share enthusiasms and questions, but it may not offer the last word. That, in the end, is up to you.

As a follow-up, my home page linked to this review. It is about one of my favorite paintings—and about a whole history of interpretation. I wrote it in early 1996, when this was site was not four years old and had at most a couple of dozen reviews. (By 2012 it had well over a thousand.) It tried to embody that model of what criticism can be. Yet an excerpt had never appeared in my blog before.

That is because I did not have a blog and had not even heard of one. I think I can safely boast that this was the first site devoted to art criticism. Still, I thought of it as a "webzine" and, like most of the site even now, it was in static html. (Well, technically I soon migrated to xhtml, I added the blog in 2002, and it in turn migrated to WordPress some years later.) The home page was then a game I had written in javascript, and one had to win merely to enter. The idea that someone might actually want to read this—or that I should be seeking out rather than discouraging readers—never occurred to me, and if it had I might be a bit more successful now!

An artist friend's teacher had just finished a book on the painting, in the National Gallery in London, and I had just seen another. I found a third recent book almost immediately, and I had been nurturing thoughts on quite a few famous interpreters for a long time. They allowed me to write about the book that first turned me on to art history, by Erwin Panofsky, as well as about changes since then thanks to the "new" art history and deconstruction. Even within his own time, as I started off above, Jan van Eyck was the stuff of legend. He did not really invent the Renaissance or oil painting, but already a portrait is about presences and reflections in a painted mirror.

"United with its frame," as I quote Panofsky, van Eyck's panels offer something more than "the modern idea of . . . imaginary space." Like a "tangible piece of luminous matter," they "confront us with a reconstruction rather than a mere representation of the visible world." In fact, Panofsky argued, they also sanctify the visible world, by witnessing and completing a marriage. A book about van Eyck, as I noted, thus walks straight into questions about artistic origins, boundaries, meaning, documentation, and truth. In the discipline today, as I say, artistic truth makes sense only as part of a broader social history, and I try to take that into account, too. Yet art like this can still take your breath away, and I meant to show you why.

I know that this runs long, to over ten thousand words. I like that this site allows me to go into more depth than ordinary magazines, but I am talking about maybe five times that usual length. I doubt that I shall ever be this ambitious again, but at least I tried once. If you did not get to the end, fine with me, but I I hope you had a taste to see what I am about (and I got to edit a few things, like the Biblical references and more recent source notes, after sixteen years). Next time I could cut readers a break by "getting real," not to mention getting back to contemporary art. And with luck I will still have in mind an ideal that does not fit so easily into blogging.

BACK to John's arts home page



Edwin Hall's The Arnolfini Betrothal: Medieval Marriage and the Enigma of Jan van Eyck's double portrait was published in 1994 by the University of California Press. Craig Harbison's Jan van Eyck: The Play of Realism was originally published in 1991 by Reaktion Books and issued in American in 1995 by the University of Washington Press. Linda Seidel's Jan van Eyck's Arnolfini Portrait: Stories of an Icon was published in 1993 by Cambridge University Press. Of course, the van Eyck double portrait is in The National Gallery in London.

A review this long must serve as an excuse, a project, and a debt. It let me spend a delightfully long time with a painter. I could also pursue a project: to examine widespread claims of a crisis, death, or revolution in art history. It remains to discharge the debt.

As usual, I took my place somewhere between journalism and scholarship. I have omitted footnotes and tried hard to make references self-explanatory or needless. I want to invite readers even a fraction as much as Jan van Eyck had rewarded my attention. Still, I have to acknowledge that debt, in addition to those I have cited in context. Although I have not relied on it, those baffled by me and my sources may prefer a less contentious introduction to iconography, online in All About Johannes Vermeer.

Someone looking simply for an overview of feminist criticism might try a survey text, Women, Art, and Society, by Whitney Chadwick, or two readers of feminist criticism by Norma Broude and Mary Garrard. Similarly, the English "new" historians were collected in The New Art History (A. L. Rees and F. Borzello, eds.), and Twelve Views of Manet's Bar (Bradford Collins, ed.) introduces the issues very well indeed for us poor Americans. The last has a very fine selected bibliography. Laurie Schneider Adams offers a decent introduction to Psychoanalysis and Art, from Freud to Lacan and after, with a good enough bibliography too.

To see the lack of demarcations in art histories, one can turn to a whipping boy for the movement, Mark Roskill's What Is Art History? Even it begins by acknowledging that its focus on the individual artist amounts to a Western construct. Or consider David Summers's Michelangelo and the Language of Art. The book does amount to old-fashioned cataloguing of sources as a clue to meaning, but the pervasive influence of semiotic criticism, such as Norman Bryson's many books or David Carrier's Artwriting, is evident from its title. From the other side of the fence, Art History's History, by Vernon Hyde Minor, interestingly places Panofsky's very writing on the Arnolfini portrait in the camp of semiotics and the new.

In laying out different fictional individuals behind each of my three books at hand, I had in mind the exploration of Patterns of Intention by another truly establishment figure, Michael Baxandall. In a curious reductio ad absurdum, he takes as his example the construction of the Firth of Forth. It must be the auteur theory of bridge-building.

In describing each book under review as clinging to older notions of interpretation, I was influenced by David Bordwell's Making Meaning, on the recent course of film theory, as well as by its obvious debt to Susan Sontag's Against Interpretation. In defending the utility of intention after all, so long as it relates to the work, my wording was refined by encounter with Bradford Collins's essay in the Manet volume. My qualified optimism about interpretation also may reflect Hans Belting in The End of the History of Art.

With Velázquez, Seidel leads me right into a minor industry centered on the relation of viewer and painter to mirror images within a painting. Michel Foucault spawned it all in the introduction to The Order of Things, and a reply by Svetlana Alpers (an article in the journal October) has also helped me, while Leo Steinberg late disrupted the painting's perspective and point of view even more (again in October). So have the extensions of the mirror theme to Manet (by critics noted above), Picasso (especially by Steinberg in the journal Critical Inquiry), and the movies (by Laura Mulvey in her book Visual and Other Pleasures). These extensions show the great importance of Jacque Lacan's claim of an infant "mirror stage," as well as the role of the Other's glance in Sartre and Beauvoir.

Finally, I have invoked an intricate chain that takes off from van Gogh's shoe paintings. In On the Origin of the Work of Art, by Martin Heidegger, the great German philosopher begged for correction—and got it in spades, in two papers by Meyer Schapiro, both in Collected Papers, Volume 4. I have also alluded to the essay on Berenson collected in Schapiro's same volume.

Besides his own clever dislocation of the van Gogh dispute, the book of Jacques Derrida mentioned in my review has another essay on which I much relied. "The Parergon," a reading of Kant's third critique, discusses the status of the frame. The Critique of Judgment, by the way, supplies my epigraph from Kant.

I mention Craig Owens's getting in a few extra kicks at Schapiro. "Representation, Appropriation, Power" is in Beyond Recognition. As I note, its idealization of the painting's audience takes one too far—and too unknowingly—from Derrida, but at least back onto that fence.


Browse or Search by artist or critic Browse by period in art's histories Browse by postmodern ideas Check out what's NEW Some of my own favorites Museums, galleries, and other resources online Who is Haberarts? Return HOME