Printer Friendly

The Arnolfini double portrait: a simple solution.

The range of responses to Jan van Eyck's Double portrait in the National Gallery is an indication of the painter's stupendous achievement (Fig. 1). The mystery has only grown with the passing centuries. This picture seems both too alien to grasp and at the same time entirely straightforward--encouraging scholars of every variety to register their own different interpretations in print. The dominant account has been that of Erwin Panofsky, who published his first treatment of the picture as long ago as 1934. (1) No matter how certain scholars have become that Panofsky was mistaken, however, his reading is the one every subsequent author must address. It will be necessary therefore to rehearse Panofsky's arguments, as well as those of some of his critics. It is by way of this revisionist history that I arrived at what I will present here: a new and simple solution to the function and meaning of the work.


Panofsky argued that the picture showed a clandestine marriage ceremony, witnessed, he claimed, by the painter himself, shown in the reflection in the mirror. With the addition of his signature, Panofsky concluded, Jan van Eyck endowed his image with the power of a legal document. Arguably, only a masterful scholar could have convinced so many people to accept such an unlikely scenario. His reading was to play the leading role for more than hall a century, so compelling was his erudition and so elegant his prose. One part historical research, one part manifesto, this was the essay in which Panofsky launched his influential but misleading concept of 'disguised symbolism' (whereby an ordinary object painted in a naturalistic way functions as the sign for an idea that--because the symbolism is unknown to modern viewers is hidden). This essay was also the popular test case of Panofsky's ambitious method of 'iconology' that was to dominate the discipline until recent times. Due to its persuasiveness and prestige, then, nearly all subsequent scholars, and the informed public at large, still follow Panofsky and refer to the picture as the Arnolfini wedding. (2)

Panofsky's contention that Van Eyck literally painted a marriage certificate was rooted in two early accounts of the picture. In 1568, Marcus van Vaernewyck described the double portrait as 'a very small panel' in which was painted 'a marriage of a man and a woman who are married by Faith'. (3) In 1604, Karel van Mander, sometimes called the Vasari of the North, perpetuated the misunderstanding in his own commentary, having drawn from Van Vaernewijck. However, it seems unlikely that either of them ever saw the work. Van Mander's interpretation was based on the assumption that the couple's right hands were clasped (since this was required in a marriage ceremony) and that a personification of Faith joined them together. (4)

It was only a few steps from there to the marriage theory set forth by Panofsky, who judged the misunderstandings to be the result of poor Latin, arguing that these earlier sources intended to say the couple was married per fidem, a legal term indicating a private marriage. 'According to canon law', Panofsky wrote, 'marriage was concluded by taking an oath, and this oath (fides) implied two actions: that of joining hands (fides manualis) and, on the part of the groom, that of raising his forearm (fides levata, a gesture still retained by out legal procedure).' (5) His learned coinage of the phrase fides levata--a convincing but altogether fictional Latin term (6)--would contribute to the overwhelming success of Panofsky's account.

Important information concerning the picture may be gleaned from inventories. These were often, if not always, made with the original to hand. (7) The inventory of Margaret of Austria's collection at Malines, taken in her presence, dates from 17 July 1516, and records 'a large picture which is called Hernoul le Fin with his wife in a chamber, which was given to Madame by Don Diego, whose arms are on the cover of the said picture; done by the painter Johannes.' (8) Another inventory of the same collection, made between 9 July 1523 and 17 April 1524, includes this entry: 'another very exquisite picture, which closes with two shutters, where there are painted a man and a woman, standing, touching hands, done by the hand of Johannes, the arms and device of the late Don Diego on the said two shutters, the name of the personage being Arnoult Fin.' This 'Don Diego' is Don Diego de Guevara, a Spanish nobleman who grew up and lived in the Low Countries; he died in 1520. (9)

The Double portrait passed by descent to Mary of Hungary, who had moved to Spain in 1556. Her inventory from around 1558 includes the following: 'a large panel, with two doors with which it closes, and in it a man and a woman who take each other's hands, with a mirror in which the said man and woman are shown, and on the doors the arms of Don Diego de Guevara; done by Juanes de Hec, in the year 1434.' (10) The panel then passed to the King of Spain, who absorbed Mary of Hungary's collection on her death.

A visitor from Leipzig in 1599, Jacob Quelviz, saw the picture in the Spanish royal collection and described it as follows: 'an image where a young man and a young woman are joining hands as if they are promising future marriage: there is much writing and also this: Promissas fallito quid enim promittere laedit/Pollicitis diues quilibet esse potest.' (11) The Latin quotation is from Ovid's Ars amatoria, and this document represents the earliest known reference to an inscription including these lines on the now lost frame. Nevertheless, the possibility that the couplet, as well as the frame itself, had always been part of Van Eyck's picture should hot be ruled out.

An inventory from about 1700 of paintings belonging to King Charles II of Spain gives the only other reference to the inscription. This entry demonstrates the way in which a text written on a frame can actually increase confusion regarding a work's subject-matter: 'a picture on panel with two doors that close with its wooden frame gilded with unburnished gold, some verses from Ovid written on the frame of the picture, which is of a pregnant German woman dressed in green giving her hand to a youth and it appears that they are getting married by night and the verses declare how they are deceiving each other and the doors are of wood painted with marbling, valued at 16 doubloons.' (12)

An inventory from 1794 offers little in terms of the subject of the picture, but reiterates a common misconception from Vasari concerning the painter's technical originality: 'one vara high by three quarters of a vara wide, a man and a woman holding hands, Juan de Encinas, inventor of off painting, 6000 reals.' (13) During the Peninsular War, Van Eyck's painting was transferred to England, before entering the collection of the National Gallery in 1842. (14)

The frame with the Ovidian inscription was lost at some time between 1700 and 1842. Most extant portraits by Jan van Eyck include some kind of identifying inscription on the frame, often featuring the sitter's name, painted illusionistically as if chiselled into stone or carved into wood (Fig. 2, 3 and 4). In the case of the Double portrait, the inventories eventually stopped mentioning the names of the persons depicted--either because the sitters had passed out of memory and their inscribed names were not considered worth mentioning, or because there was no inscription by the time a given inventory was made. Perhaps the names were cleverly encoded, as on Van Eyck's Portrait of Jan de Leeuw, in which the date is represented in the form of a chronogram (Fig. 3).


The first two inventories do, however, state the identity of the man depicted. The name 'Hernoul le Fin', mentioned in Margaret of Austria's inventory, depended either on an inscription or on documentation from Don Diego. (15) Presumably a discerning former owner would have left any inscription on the frame, especially one painted by a master of Van Eyck's prestige. Edwin Hall claims that the different vernacular spellings of the surname demonstrate that the name's appearance in the inventories depends ultimately on oral--rather than written--history (as we have seen, the 1523-24 inventory calls him 'Arnoult Fin' and that of 1516 'Hernoul le Fin'). (16) However, Hall fails to mention that the painter's name is translated into the vernacular in some inventories as well ('done by Juanes de Hec, in the year 1434'; 'Juan de Encinas, inventor of oil painting'), despite his signature's prominent position in the middle of the painting itself ('Johannes de Eyck fuit hic 1434'). Furthermore, the 1516 inventory reads: 'a large picture which is called Hernoul le Fin with his wife in a chamber'. If it was 'called' something, then there was likely in be an inscription on the frame; assigning titles was not common in this period and seem most likely in have been derived from words inscribed by the painter himself.

We learn from Quelviz around 1600 that 'there was much writing and also this'--the couplet from Ovid's Ars amatoria. While the possibility has recently been rejected out of hand, I would like to suggest that Van Eyck himself inscribed the frame in this way, following his usual practice. (17) The inventory of c. 1700 referred to above describes 'its wooden frame gilded with unburnished gold', which corresponds in the way Van Eyck also decorated his supposed Self-portrait (Fig. 4) and notes 'some verses from Ovid written on the frame of the picture ... the doors are of wood painted with marbling." It is likely that a gilded frame, as well as shutters painted to simulate marble, belonged in the work from the beginning, since such features are entirely characteristic of panels painted by Jan van Eyck (see Figs. 2, 3, 4, 5, 13). Indeed, their inclusion was common artistic practice in this period in general: only rarely do we find a northern renaissance portrait that did hot originally have a cover of some kind. (18) Moreover, the undecorated reverse of the Arnolfini panel suggests that it was created as the centrepiece of a triptych. Van Eyck's extant single-panel portraits are all decorated on the reverse, whereas the central panels of his surviving triptychs are not. Furthermore, Angelica Dulberg has noted that quotations from Ovid appear frequently on portraits dating from the fifteenth and sixteenth centuries. (19) Sayings by this Roman author were popular because of the way they juxtaposed life's pleasures with death and this-worldliness with the transience of existence. (20)


'The very same conduct which Ovid ironically recommends', wrote C.S. Lewis, 'could be recommended seriously by the courtly tradition ... The contrast inevitably raises in our minds a question as to how far the whole tone of medieval love poetry can be explained by the formula, "Ovid Misunderstood". (21) Perhaps this is equally true of the mysterious Ovidian text that used to decorate the frame. Or was there another reason for the tone of the inscription? The text may have been much longer, since one of the inventories specifically states that 'there is much writing and also this ...' One can only imagine what else might have been gleaned from the frame.

In 1857, Crowe and Cavalcascelle recognized the London Double portrait as the one described in Margaret of Austria's inventory, and translated the Flemish/French corruption of the name 'Hernoul le Fin' back into its original Italian, Arnolfini. Pinpointing which Arnolfini has proved to be more of a challenge. Until very recently, it had been universally assumed, following Weale in 1861, that the couple were Giovanni di Arrigo Arnolfini and his wife Jeanne Cenami. This Arnolfini was the most prominent member of an important Lucchese family resident in Bruges in the fifteenth century, but--as Lorne Camp bell has now proved--the portrait cannot be of Giovanni di Arrigo, since he only married in 1447, thirteen years after it was painted and six years after Jan van Eyck's death. (22)

There were no less than five Arnolfini in Bruges at the time who could have commissioned the painting. Campbell identifies the man as Giovanni di Nicolao Amoltini, the elder of the two Giovanni Arnoltinis living in Bruges during Van Eyck's lifetime. While there is no trace of Giovanni di Arrigo in the Bruges archives until 1435, (23) Giovanni di Nicolao had lived in Bruges since 1419 or earlier, and would have had every opportunity in become acquainted with Jan van Eyck well before 1434. Giovanni di Nicolao was married in 1426 to Costanza Trenta, who like him came from a prominent family from Lucca. (24) They had been betrothed on 23 January 1426, when Costanza was thirteen, (25) which means she would have been twenty one when the painting was executed. It finally looked as if Campbell had hit upon the perfect match. But then a further archival discovery was made: a document proved that Costanza was dead by 1433, the year before Van Eyck dated his picture.

Costanza's mother, Bartolomea, was the daughter of Giovanni di Amerigo Cavalcanti, a Florentine of considerable stature. In 1416, Bartolomea's sister was married to Lorenzo de'Medici, the brother of Cosimo il Vecchio. It is thanks to this family connection that we learn of the fate of her daughter. On 26 February 1433, Bartolomea writes from Lucca to Lorenzo de'Medici to congratulate them on the birth of their son, and in the course of a discussion of her children, she mentions that her daughter Costanza is no longer alive: 'solamente ne viveno due, Paulo e Johi. Morio la Costanza e Lionardo. Paulo si trova in Avignone ... Johanni e a lucha ...' (26) Reaching what appeared to be another impasse, Campbell finally concluded that: 'If Hernoul le Fin is rightly interpreted as Arnolfini, then van Eyck's couple may be tentatively identified as Giovanni di Nicolao Arnolfini and his putative second wife.' (27)

I believe Lorne Campbell is correct that this is Giovanni di Nicolao Arnolfini. Moreover, I agree with him that this is no picture of a betrothal ceremony, nor indeed of a wedding. But is it a straightforward portrait of two wealthy people, as he contends? More crucially, is he correct in assuming that this is a second wife, of whom we have no record, and with whom there were no recorded children? What happens if we look at the portrait in an altogether different way? What if it is of the deceased Costanza? It is my contention that Van Eyck's picture is a posthumous representation of Costanza, the only wife of Giovanni di Nicolao of whose existence we find any evidence.

If we first of all consider various details in the picture that have particularly attracted the attention of scholars, as well as some that have been previously overlooked, we find they make far more sense within the context of a posthumous portrait of Costanza and her living husband.

A consistent focal point of much of the literature on the work is the man's raised hand. Is he shown taking a marriage or betrothal oath, as Panofsky, Hall and others would have it, or is he merely greeting the two gentlemen who enter the room, as Campbell contends? This brings us back to the grain of truth on which Panofsky based his interpretation. The term fides does indeed exist in the context or oath-taking ceremonies. As Hall reminds his readers:

In other respects the basic meaning of fides remained unchanged: to 'faithfulness,' 'honesty,' and 'promise', which are among the primary connotations of the word, Roman law had added the idea of an honest keeping of a promise or the obligations consequent to an agreement--essentially what Augustine meant when he termed fides one of three 'goods' of Christian marriage. Fides, in a somewhat broader medieval usage, meant not only a solemn promise to do something, but--by extension--an oath associated with such a promise or agreement. (28)

In other words, the oath gesture Arnolfini makes may be a reference to an oath already taken, or perhaps an oath that serves to renew a promise made at some point in the past.

However, Hall's conclusion that the two are shown at their betrothal ceremony should be questioned for several reasons. Klapisch-Zuber has shown that in Tuscan families in this period, the betrothal process usually did not include the bride-to-be but only the men of the family. (29) Furthermore, the rites of marriage and/or betrothal were hot as straightforward as either Panofsky or Hall claim; they did not follow a strict, predictable or recognisable protocol. (30) Finally, it seems worth noting that in the majority of images of weddings of this period, the bride's hair is worn long, in the manner of a virgin, (31) not as the woman wears it in Van Eyck's Double portrait. In the Seven Sacraments Altarpiece by Rogier van der Weyden and his workshop, the sacrament of marriage is represented by a woman taking her marriage vows, with long hair similar to that found in images of the Virgin Mary (Fig. 17). In the rite of baptism, a pair of matrons have their hair pinned up, under a veil, in much the same fashion as Signora Arnolfini (Fig. 18). In my view, what is depicted is a reference to the couple's already established union rather than a ceremony of betrothal or marriage.


If we turn to the rich cultural context evoked by Johan Huizinga, we find that in this period the taking of an oath could have more than just legal connotations. Huizinga portrays life at the Burgundian court--precisely the world in which Jan van Eyck and the subjects of his painting lived--where it was customary to take oaths taken for almost any reason. In his words:

The most famous solemn vow of the fifteenth century, the Voeux de Faisan, was taken in 1454 in Lille during a court festival given by Philip the Good in preparation for the crusade. What it still reveals of all this is not much more than a beautiful courtly form. Not that the custom of taking a spontaneous vow during an emergency or moment of strong emotion had lost any of its power. This custom has such deep psychological roots that it is bound neither to education nor faith. The knightly vow as cultural form, however, as a custom elevated to an embellishment of life, reaches its last phase in the splendid extravagances of the Burgundian court ... Others take cautiously conditioned vows that testify both to serious intent and to self-satisfaction with a beautiful pretense. On some occasions the vows are addressed to the 'much beloved' who is but a pale remnant of herself. (32)

The oath depicted in Van Eyck's painting therefore need not have had either a religious inspiration or a legal cause. There exists a whole dimension of courtly, secular life that has been left out of much of the literature on early Netherlandish art, and that may inform the meanings of all sorts of pictures.

Scale discrepancies between the chandelier, the mirror, and even the figures, in relation to the space they inhabit (mistakes on the part of this particular artist seem inconceivable), along with the absence of a fireplace, make it impossible to see the picture as the portrait of a room that really existed. (33) This was Jan van Eyck's typical practice: he built spaces that--although entirely believable--were in fact imaginary. Christopher Wilson has shown that the architecture of Van Eyck's interiors is habitually made up of the elements of several different buildings, and would most likely not stay up if built. (34)

The Arnolfini have dressed up for this occasion, as sitters were wont to do when having their portraits painted. In that society, one did not wear clothing such as this at home. (35) As Charles de la Ronciere has observed: 'But the moment one left the private realm to be seen in the outside world, the gamurra [informal dress] ceased to be appropriate. At such times the chests were opened and the richest fabrics brought out, for the clothing worn in public was a matter of individual and social distinction.' (36) The exhibition of one's wife as if putting a treasure on display was considered proper; (37) and, indeed, this is part of what Van Eyck's painting shows.

In mid-fifteenth-century Florence, the husband typically spent about a third to two-thirds of the dowry on clothes for the new wife and furnishings--often well above the cost of the trousseau. A woman's wedding clothes were frequently the property of her husband, sometimes provided to make a show of her as his property; they were well adorned, sometimes with family jewels that were not always given to her permanently. The fact that the widow was on occasion required to leave her clothes behind in the event of remarriage is often mentioned in wills. Furs were a common wedding gift from the husband, and were subsequently worn on important occasions. (38) The enormous train on the lady's dress resembles what a royal personage, such as Van Eyck's St Catherine of Alexandria (who according to legend was a princess, and is portrayed as one) in the right wing or the Dresden Triptych (Fig. 5), would wear.

In the long history of responses to the London picture, there has been a general tendency among those who view it for the first time to assume the lady in it is pregnant. Those who know Van Eyck's whole oeuvre well tend to believe otherwise--that it was simply the fashion to wear this style of clothing and that her attire should hot preclude her from being seen as a virgin bride. Van Eyck's St Catherine (Fig. 5) again comes to our aid here, since she betrays the same bulge around the belly. But whether literally pregnant or not, the fashion for accentuating the womb itself relates to women's duty to bear children and to a physique that makes this possible.

The colours worn by the Arnolfini are likely to carry significance--something hitherto ignored by scholarship. Signora Arnolfini wears vivid green wool with the whitest white ermine. Her underdress is blue, and is also trimmed with white fur. Her husband, on the other hand, wears sombre tones of deep purple and black. Huizinga had a great deal to say about the colour symbolism of clothing in his Autumn of the Middle Ages, not least in an eloquent passage relevant to the colours worn by Arnolfini and his wife:

It is remarkable that black and violet are more popular for clothing than green and blue, while yellow and brown are almost entirely missing. Black, and above all black velvet, undoubtedly represents the proud, somber splendor that the time loved, with its arrogant distance from the gay wealth of color found everywhere. Philip the Good, after having passed the days of his youth, always wore black and had his entourage and horses in the same color. The rare presence of blue and green should not, incidentally, be entirely regarded as a direct expression of the sense of color. More than the other colors, blue and green held symbolic significance and these meanings were so specific that they nearly rendered both colors unsuitable for regular clothing. Both were the colors of love: green symbolized the state of being in love, blue faithfulness. Or, better put, these two were in a very special way the colors of love, but the other colors could also serve in the symbolism of love. Deschamps says of a group of suitors: 'Some dress themselves for her in green/the other blue, another in white./Another in vermillion like blood,/And he who desires her most/Because of his great sorrow dresses in black.' But green was especially the color of young hopeful Minne: "You will have to dress in green/It is the livery of those in love.' (39)

The work was painted before black clothing became de rigueur at the Burgundian court. Philip the Good, Duke of Burgundy, wore black and created the fashion for it (Fig. 19). (40) But 1434 was the time of his father, and in paintings by Jan van Eyck, or indeed any paintings from around that date, one is bard pressed to find other examples of men dressed with such extreme sobriety. Our interpretation of Arnolfini's clothing may draw once again on Huizinga:


The dressing of sentiment in the garb of a suggestive form reaches its highest development in mourning. There were unlimited possibilities for a splendid exaggeration of sorrow, the counterpart of the hyperbolic expressions of joy during the grandiose court festivities ... Court mourning during Burgundian times can only be understood if viewed in relation to elegy. The displays of mourning demonstrated in beautiful form how totally powerless the affected individual is in the face of suffering. The higher the rank the more heroic the display of pain. The Queen of France had to stay an entire year in the room where she was told of the death of her husband. (41)

Like birth and marriage, death had its own set of rules. The corpse was moved from the private bedchamber to the tomb in a public procession not unlike that at a wedding, with the family marching in hierarchical order. This was the last public spectacle associated with the deceased. Also public were the signs of mourning. (42) One's marriage was an aspect of life especially put on display al the moment of death, as either--in Dominique Bathelemy's words--a 'revelation of life's most essential relationship at a time when falsehood was out of the question, or a final opportunity to shape an image of the ideal.' (43) A certain Baldwin II, a count who died in 1169, may have had many extramarital lovers, yet 'nevertheless, he suffered greatly when his legitimate wife ... died in childbirth. In every way he seemed distressed and inconsolable.' (44)

Arnolfini came from one of the most prominent Italian families at the Burgundian court. Most of the Italians who came to live and work in Bruges and other important trading cities in the north were of the merchant class, and some of them--like the Arnolfini--possessed significant wealth. While living abroad, they probably came under the influence of their adopted society in many ways, but certain local customs must presumably have accompanied them from home. Arnolfini's choice of a fellow Lucchese as his wife suggests as much, and it is therefore appropriate to consult Tuscan sources for many aspects of their lives, especially in connection with marriage and family, whereas Burgundian culture is more likely to be relevant for their daily life together in Bruges.

Individual members of prominent fifteenth-century Tuscan families often kept family histories known as Ricordanze (memoirs). Rather than relying on memory or hearsay, the head of the family might cite notarised contracts, account books, and affidavits as supporting evidence. They were careful to leave accurate information about themselves and their children: age, name of godfather, hour and day of birth, and so on. Alberti was the first to recommend this practice. Another act of remembrance was the commissioning of detailed, lifelike, and convincing portraits of their ancestors. (45) Yet the commemoration of women--dead or alive--was rare, and may contribute to our confusion regarding the true purpose of this image.

Klapisch-Zuber explained the limited role of women within family heritage during this period:

Women, then, were hot permanent elements in the lineage. Memory of them was short. An important woman, a benefactress for her kin, for example, would eventually be known under her own name and brought to people's attention; but the family chronicler or the amateur genealogist would feel obliged to explain why, since the process fit so poorly within their definition of kinship ... The family chroniclers keep the memory, of an alliance with a certain lineage, but forget, a few generations after the marriage, the given name of the woman on whom the alliance was built. (46)

This suggests that the purpose of the Double portrait may in part have been to display an advantageous marriage to posterity, whether or not the bride was still living al the time of the portrait, and whether or not her name was included in the artist's inscription.

On the other hand, this is not to suggest that within the profane display of courtly culture, there was no place for the sacred. In describing the beau ideal of manhood, many medieval authors celebrate a hybrid of secular and chivalric qualities. When aiming for this ideal, goodness does not mean aestheticism, nor does knighthood mean adultery. (47) As Lewis wrote: 'The Ovidian tradition, operated upon by the medieval taste for humorous blasphemy, is apparently quite sufficient to produce a love religion, and even in a sense a Christianized love religion, without any aid from the new seriousness of romantic passion.' (48) In short, 'Love is, in saeculo, what God is, in eternity.' (49)

The Arnolfini of Lucca, who spent their adult lives al the court of Burgundy, would have been very well aware of all this. As Huizinga reminds us, 'the history of culture has just as much to do with dreams of beauty and the illusions of a noble life as with population figures and statistics.' (50) A kind of tunnel vision may have affected interpretations of Van Eyck's Double portrait. Scholars have debated whether it is a portrait laden with Christian symbolism or something devoid of symbolism altogether, but rarely, if ever, have interpretations taken into account the courtly context of Bruges in 1434. 'The strict cultivation of the beautiful life in the form of a heroic ideal is the characteristic that ties French knightly culture after the twelfth century to the Renaissance.' (51) The forms of life assumed by the nobility were avidly imitated by those members of the third estate who could afford to do so. (52)

Later painters borrowed from the composition of the Double portrait, both for sacred and for secular subjects. These borrowings include an altarpiece wing panel in the Prado by a follower of Campin, and the Kansas City Holy Family by Petrus Christus (Fig. 6). In contrast to these images, manuscript illuminations by Loyset Liedet that depend on Van Fyek's portrait--both in terms of their composition and their details, are secular. (53) A painting on loan to Bonn from Bad Godesberg of 1470 was very closely modelled on the London panel--it shows a man and woman holding hands in an interior with a mirror on the wall behind them (Fig. 7). (54) This dependence is well known, but it has not been mentioned in this context that the picture in question has corpses painted on the reverse (Fig. 8). Other versions of the Eyckian original feature similarly overt references to death. Corpses are depicted on the reverse of a panel of c. 1470-80 by an anonymous Ulm painter (Figs. 10-11). (55) One striking example of a living sitter juxtaposed with a corpse is a diptych painted by the Basel Master of 1487, in which a young man in the left wing is paired with a decayed female cadaver in the right (Fig. 12).

[FIGURES 6-8, 10-12 OMITTED]

Double portraits (or images of an idealised couple) with a pair of cadaverous counterparts on the reverse began to appear in the mid 1450s. Essentially, these were an admonition to lead a virtuous life. The contrast between the peak of one's physical self (at a time such as marriage) and the deterioration of one's body after death contributed to this message. (57) These portraits are linked visually and symbolically to tombs representing the dead in the early stages of decomposition, (58) so-called transi tombs, which predate the Double portrait by a few decades and continued to be common for centuries to come. (59) The most common inscription on such tombs is a form of: 'I was like you, and you will be like me.' (60)

In his book on tomb sculpture, none other than Panofsky discussed an English tomb brass depicting a married couple (Fig. 9). Reporting that they are shown at the moment of their wedding, he finds this treatment both curious and provincial:


English and English-influenced tomb slabs and brasses, for example, show a married couple, both recumbent in the process of taking the matrimonial oath, occasionally in such a manner that the head of the lady reposes on a pillow while that of her husband does not ... (61)

Panofsky does not compare the tomb brass to the painting by Van Eyck, despite the clear formal relationship. In fact, the hands of husband and wife were often linked in English funerary monuments. (62) Needless to say, a tomb marker was not a marriage document: like Van Eyck's Double portrait, such an image merely referred to matrimony as a key moment in the life of those depicted.

However, Panofsky does link an epitaph in Tournai Cathedral of about 1438 (Fig. 14) with Van Eyck's Virgin of Canon van der Paele (Fig. 13). (63) Although it is often described as an altarpiece, this painting is far more likely to have served as an epitaph--a potential function of this and other pictures by Van Eyck that needs further study. As in the Virgin of Chancellor Rolin (Fig. 15), to which one might compare another contemporary epitaph (Fig. 16), and--I would argue--the Double portrait, here too the artist allows us to behold the earthly presence of a living individual together with an apparition that appears to that individual in his mind's eye, as if were. Many epitaphs made at this time in the Netherlands have the same fundamental structure. They show what the depicted person wishes he could see tangibly before him, but they also reveal the fact that it is only an illusion.


Seen in these terms, Van Eyck's Double portrait climaxes in the vision of the deceased lady. It is completely orchestrated around that basic idea, from the difference in the treatment of light--she is bathed in a kind of ethereal whiteness, while he stands in comparative shadow--to the inventory of objects that surround the two.

Panofsky argued that the dog in the foreground at the lady's feet, 'seen on so many tombs of ladies, was an accepted emblem of marital faith', while Hall and Campbell, in contrast, each found the Arnolfini dog to be devoid of symbolic significance. (64) Yet the dog is surely another link to death imagery. In ancient funerary monuments and medieval tomb effigies, of which the tomb of Margaret of Austria (Fig. 21)--the owner of the Double portrait--is a characteristic example, such dogs served to express 'the mutual affection of husband and wife in a happy marriage.' (65) A case in point is the Count and Countess of Henneberg Tomb by Peter Vischer (Fig. 20). The meaning of the dog may or may not be linked to fidelity--the real point here is that a common type of female tomb effigy in this period includes a full-length portrait of the deceased with a small dog at her feet. It may be significant that a dog appears in many effigies au vif--as in life--but not in effigies en transi. Rather than being a sign of fidelity, perhaps the dog's role is to accompany the dead in eternity, like the angels that also sometimes appear with tomb effigies (for instance, on the tomb of Philip the Bold at the Chartreuse de Champmol). (66)


What of the bed, which Campbell reads simply as relatively expensive furniture? (67) Beds depicted in rooms of state, birth scenes, and even Virgin and Child images may in some cases have had a neutral function--the expression of the social standing of the inhabitants of the house in question. But surely the portrayal of a man and woman clasping hands in such a room signifies something more.

Netherlandish images of this period depicting hung beds in use typically show a birth, as in the Birth of St John the Baptist from the Turin-Milan Hours (Fig. 22), or a death, as in the Office of the Dead miniature in the Spinola Hours (Fig. 23). In the ritual practices of the period, when visitors were received into the home, they were brought to the principal chamber of the house where the new mother, the dying, or the recently deceased were to be found lying in state on a hung bed. (68) In his remarkably thorough study of the Double portrait, Campbell draws attention to a contemporary source that sheds light on the possible use of the room depicted, observing:


It is difficult to find out about contemporary conventions and etiquette, but one invaluable source is Les Honneurs de la Cour, written between 1484 and 1491 by Alienor de Poitiers, the widowed Viscountess of Veurne ... She was particularly interested in the conventions observed when ladies of various ranks were lying in. (69)

Alienor de Poitiers observes that:

... a noblewoman below the rank of countess normally placed only one carpet in front of her bed when she was giving birth; only a lady of the highest rank was permitted to carpet the entire floor of her lying in chamber. (70)

Given that Costanza Arnolfini died childless around the age of twenty, it is not unreasonable to wonder whether she did not die in childbirth. This might explain why the floor is decorated in the way it is. As Campbell further notes: 'Bare wooden boards such as these are not often found in representations of interiors, where floors are usually tiled or inlaid with semi-precious stones.' (71) The carpet beside the bed indicates a chamber decorated for Costanza's lying in; its bare floor indicates a measure of austerity appropriate to her station.

From Alienor we also learn that two silver candlesticks were kept on the dresser in the lying in chamber of a lady; in the candlesticks 'there must be two large wax candles, to be lighted when someone comes to the chamber', day or night. (72) Similarly, Huizinga reports that two large candles in silver holders burned continuously in the lying in room of Isabella of Bourbon, where the shutters were kept closed for fourteen days. (73) 'Next to mourning', Huizinga writes, 'confinement during childbirth offered ample opportunities for serious pomp and hierarchical distinctions of ostentation.' (74)

The brass chandelier has long been considered particularly mysterious. Why light a candle in the daytime (Fig. 25)? And why only one? There are several religious paintings of the period with candles seemingly unnecessarily lit; pictures with a lit candle in the daytime are also often those in which a lying in chamber is depicted, such as the Virgin and Child now attributed to Jaques Daret in the National Gallery, or countless scenes of the Nativity with Joseph holding a lit candle.


An observant viewer of the Double portrait will discern the remains of a burnt-out candle in the front right sconce. This was recently pointed out by Campbell, but has hot as a rule been noticed by scholars. (75) If one takes the view that the portrait of Signora Arnolfini is posthumous, then the lit candle as opposed to the burned-out one refers to the couple, the lit candle on the side of the living, the extinguished one on the side of the deceased. I do not believe it is sufficient to call the single lit candle 'nothing more than good bourgeois thriftiness.' (76)

Petrus Christus's Holy Family (Fig. 6) includes a chair that is clearly part of the bed construction, as seems also to be the case in the London panel. The chair in the background of the Double portrait is decorated with carved figures. Rather than the more typical lions found in Christus's image, it is adorned with a figure of a very different sort. St Margaret, the patron saint of pregnant women, is usually shown having emerged unscathed from the belly of a dragon, standing over its carcass, triumphant.' (77) Here instead she is shown standing, or possibly kneeling, behind her attribute, with her hands clasped in prayer--an extraordinarily rare pose for this saint (Fig. 25).

A more menacing indication of the painting's true subject may be round below the mirror, on the wooden bench (Fig. 25). To quote Campbell once more:

On the arm is a carving of a monster with a grimacing human face, lion's ears and hoofs instead of hands: he wears a hat and a bib and is seated back to back with what is either a carved lion or another, similar monster. Benches were often decorated with carved lions; the grotesque is perhaps less usual. (78)

Within the context of an apparently normal, everyday setting, Van Eyck provided his contemporary viewers with enough information to allow them to recognise the conditions of Arnolfini's loss. The grimacing face of the monster hovers just above the touching hands of the Arnolfini, symbolically threatening their union. All of these unusual features follow a clear logic when the painting is viewed in light of the hypothesis that this is a painting in memory of Costanza Arnolfini after her death.

The convex mirror (Fig. 25), concerning which more must have been written than about almost any other mirror in the history of art, has ten roundels depicting the Passion of Christ--or so we are often told. To be more precise, there are eight scenes of the Passion plus two scenes showing Christ's life after his death. Beginning at the bottom and moving clockwise, we have the Agony in the Garden, the Arrest of Christ, Christ before Pilate, the Flagellation, Christ Carrying the Cross, the Crucifixion, the Descent from the Cross, the Entombment, the Harrowing of Hell and the Resurrection. (79) In a division that echoes that of the candles, the scenes of Christ living are on the left--Giovanni di Nicolao's side--and the scenes of his death and resurrection are on the right--Costanza's domain.

In art of the period, mirrors like that of the Arnolfini are also sometimes identified as mirrors of death--a further element in the repertoire of memento mori. The idea here is that mirrors tell the truth about life--in the mirror we are nothing but this transient being (Fig. 24). A small double-sided triptych by Memling includes a skull on one panel, another panel with the figure of Death, and a third with an image of Vanity holding a mirror, a small dog at her feet. (80) Here we find a neat summary of some of the major themes at work in Van Eyck's Double portrait. The reason Vanity holds a mirror in Memling's picture is not only that she is pleased with her own self-image, but also that it is vain to believe beauty will last. In a similar way, Rogier van der Weyden's Braque Triptych in the Louvre is a combination of admonition and epitaph: the skull refers to Adam and original sin as well as the recently deceased Braque himself for whom the work was commissioned, warning the viewer of the inevitability of death. (81)


In signing his name just above the mirror, the artist may have included himself in this symbolic constellation, as part of the play of presence and absence operating throughout the picture. On another level, 'Johannes van Eyck fuit hic' may be an assertion of the veracity of his painted representation, like his inscription on the Leal Souvenir (Loyal Remembrance) portrait in London (Fig. 26), (82) which was itself most likely a posthumous memento or secular epitaph, too. It was signed: 'Transacted on the 10th day of October in the year of our Lord 1432 by Jan van Eyck'--language far more explicit in its association with legal documents. (83) This seems to suggest that the painter was attesting to the truth of the likeness.


A few key aspects of the earlier stages of work revealed by infrared reflectography support the view that Van Eyck was working to accentuate the death of Costanza, suggesting that she either died in the process of the painting's creation or that--determined to create a work of great significance and complexity--he elaborated his message in the course of its execution, necessitating changes to the programme. No underdrawing can be found for the dog or the chandelier. The mirror was larger and originally octagonal. Perhaps the eight-sided mirror was given ten sides so that the hopeful, post-mortem scenes of the Harrowing of Hell and the Resurrection could be added? The smaller mirror also made room for the signature. (84) The high backed chair with its figure of a praying St Margaret was also not underdrawn. (85)

Changes to the underdrawing of Arnolfinis raised right hand (Fig. 27) have been interpreted as being intended to improve the plasticity of the hand, with out affecting the meaning or significance of the oath swearing gestured. (86) To my eye, however, his gesture seems--in the final solution--to be more directed to his wife than to the viewer. As Panofsky says, 'the husband gingerly holds the lady's right hand.' (87) In the underdrawing, Arnolfini had a better grip; now it slips through his fingers, as perhaps did Costanza herself. (88)


The present article is dedicated, most appropriately, to my husband Joseph Koerner, whom I married the day after I finished this article.

I would also like to acknowledge Lorne Campbell whose conscientious and thorough scholarship has always been an inspiration to me, and whose pioneering research in this particular case provided the groundwork for the present study. Whether or not he will agree with my interpretation, it was his admirable treatment of the picture which provided the primary sources for this article.

(1) E. Panofsky, 'Jan Van Eyck's Arnolfini Portrait', Burlington Magazine, vol. LXIV (March 1934), pp. 117-27. Reprinted in M. Levey (ed.), The Burlington Magazine: A Centenary Anthologyy, New Haven and London, 2003, pp. 42-49.

(2) There have certainly been challenges, but they had little impact on the non-specialist readership. See, for example, L. Benjamin, 'Disguised Symbolism Exposed and the History of Early Netherlandish Painting', Studies in Iconography, vol. II, 1976 pp. 11-24; J.B. Bedaux, 'The Reality of Symbols: the Question of Disguised Symbolism in Jan van Eyck's Arnolfini Portrait', Simiolus, vol. XVI, 1986, pp. 5-28; L. Seidel, 'Jan van Eyck's Arnolfini Portrait: Business as Usual?', Critical Inquiry, vol. XVI Autumn 1989, pp. 55-86; M. Carroll, 'In the Name of God and Profit: Jan van Eyck's Arnolfini Portrait', Representations, vol. XLIV, 1993, pp. 96-132, and E. Hall, The Arnolfini Betrothal: Medical Marriage and the Enigma of Van Eyck's Double Portrait, Berkeley, Los Angeles, London, 1994.

(3) 'Een trauwinghe van eenen man ende vrauwe/die van Fides ghetrauwt worden'; quoted in L. Campbell, National Gallery Catalogues: The Fifteenth Century Netherlandish Schools, London, 1998, p. 198.

(4) Ibid., op. cit., p. 198.

(5) Panofsky, op. cit., p. 123.

(6) As pointed out in Bedaux, op. cit., pp. 7-8.

(7) The following inventory, excerpts are all taken from the painstakingly researched catalogue entry in Campbell, op. cit., pp. 174-211.

(8) Ibid., p. 174.

(9) Ibid., p. 174.

(10) Ibid., p. 176.

(11) Campbell, op. cit., Book I, lines 443-44: Promittas facito, quid enim promittere laedit? Pollicitis diues quilibet esse potest

('See that you promise: what harm is there in promises? In promises anyone can be rich') Ibid., p, 176.

(12) Ibid., p. 176. In a recent lecture on titles (Courtauld Institute, 19 May 2003), David Ekserdjian made the pertinent observation that as a rule the less an inventory-taker knew, the more verbose the description tended to be.

(13) A vara measures 84 cm; Campbell, op. cit., p. 176. These are the correct dimensions, so it would seem likely the compiler of the inventory had the picture before him.

(14) Ibid., p. 176.

(15) Ibid, p. 192.

(16) Hall, op. cit., p. 4.

(17) Campbell, op. cit., p. 198, was the first to publish an opinion on the Ovid inscription, and believes it was added later, partly because it is not mentioned in any of the earlier descriptions of the picture, Hall did not know the lines of Ovid, only that the protagonists were said to be 'deceiving each other' He believed that a marbled frame sounded right for Van Eyck, but that the satirical nature of the inscription did not seem appropriate; see respectively Hall, op. cit., p. 137, note 9, and p. 6.

(18) See A. Dulberg, privatportrats: Geschichte und Ikonologie einer Gattung im 15. und 16. Jahrbundert, Berlin, 1990, plates 61, 186, 189.

(19) For a thorough study of renaissance portrait formats and covers, as well as iconography, see ibid.

(20) Ibid., p. 153.

(21) C.S. Lewis, The Allegory of Love: A Study in Medieval Tradition, New York, Oxford, 1958, pp. 7, 8, where, in a passage on the mocking tone of the Ars amatoria, Lewis wrote: 'this is a pretty instance of the vast change which occurred during the middle ages ... But then one would have to acknowledge that it was a very consistent misunderstanding'.

(22) Campbell, op. cit., p. 193.

(23) Ibid., p. 195. The documentary evidence from that date, which is relatively substantial, gives no indication that he had previously been married. He had no children except two illegitimate ones--his money went to Jeanne Cenami and then to the family in Lucca.

(24) For a succinct account of the Trenta family's patronage of Jacopo delta Quercia in Lucca, see J. Turner (ed.), The Dictionary of Art, 34 vols., London, 1996, vol. XVI, pp. 841-42 (entry by E.M. Richter).

(25) Campbell, op. cit., p. 195. The dowry was 800 florins or more, at the discretion of her grandfather, since her father was dead; this was art average amount for someone at her level of society, even on the small side if it was only 800; see C. Klapisch-Zuber, Women, Family, and Ritual in Renaissance Italy, translated by L. Cochraine, Chicago, 1985, pp. 220-21.

(26) 'Only two are living, Paulo and Johi. Costanza and Lionardo died. Paulo is in Avignon. Johanni is in Lucca.' Florence, Archivio di Stato, Mediceo avanti il principato, xx/40; letter dated Lucca 26 February 1432/33. First cited in Campbell, op. cit., p. 194, note 178, p. 208.

(27) Ibid., p. 198.

(28) Hall, op. cit., p. 65.

(29) Klapisch-Zuber, op. cit., p. 183.

(30) Ibid., p. 190.

(31) G. Duby (ed.), A History of Private Life: II Revelations of the Medieval World, Cambridge, MA, and London, 1988, p. 126; see also Hall op. cit., plates 2, 5, 6, 8, 9, for several examples of marriage scenes depicting brides with long hair.

(32) J. Huizinga, The Autumn of the Middle Ages, Chicago, 1996, pp. 101-102.

(33) Campbell, op. cit., p. 191. To cite another example of spatial distortion, in Van Eyck's Virgin in a Church in Berlin, the Virgin stands as high as the clerestory of the nave.

(34) C. Wilson, seminar lecture, Courtauld Institute of Art, Spring 1992, instancing among other works the Washington Annunciation.

(35) Duby, op. cit., p. 194.

(36) Ibid., p. 195.

(37) Ibid., p. 82.

(38) Klapisch-Zuber, op. cit., pp. 220-21, 225, 230, 279.

(39) Huizinga, op. cit., p. 326, and p. 19, for the practice of dressing in colours to show political allegiance.

(40) See also Hall, op. cit., plate 11, for Jean Mielot presenting his translation of the Traite sur l'oraison dominicale to Philip the Good, c. 1457, Brussels, Bibliotheque Royale Albert 1er, MS 9092, fol. 1r.

(41) Huizinga, op. cit., p. 53, 56; see also p. 55, for a discussion of pleurants.

(42) Duby, op. cit., p. 84.

(43) Ibid., p. 143.

(44) Ibid., p. 146

(45) Ibid., p. 257.

(46) Klapisch Zuber, op. cit., pp. 118-19.

(47) Lewis, op. cit., p. 104.

(48) Ibid., p. 20.

(49) Ibid., p. 42.

(50) Huizinsga, op. cit., p. 103.

(51) Ibid., p. 41.

(52) Ibid., p. 104.

(53) Campbell op. cit., p. 178, figs. 4, 5.

(54) Ibid., p. 205, note 23: 'Aloisius-Kolleg, Bad Godesburg: formerly on an oak panel, 88 x 55 cm [almost the same size] ... Attributed by Buchner to the Meister der Aachener Schrankturen, it was discovered beneath a painting of the 1520s representing a Virgin and Child with Heinrich Krain. Krain was a canon of St Gereon at Cologne. The painting of two corpses on the reverse was done at the same time as the double portrait, which is known only from photographs of the back of the paint layer, taken when the Virgin was removed from its support.' See also Buchner, op. cit., pp. 173-75, 219, plate 196; Alfred Stange, Kritisches Verzeichnis der deutsehen Tafelbilder vor Durer, Munich, 1967, vol. I, p. 77, no. 230: 'Hochzeitsbildnis ... Vorderseite: Brautpaar. Ruickseite: Gestalt des Todes.'

(55) Dulberg, op. cit., p. 251, no. 216, plate 127, for yet another example, by a North German Master, c. 1500 (Evangelische-Lutheranische Kirchengemeinde, Laurenberg an der Elbe). There are other double portraits which were designed more as emblems of loss than of marriage: one such is a half length portrait of Ladislaus v (1440-57), King of Hungary and Bohemia, and Magdalena, daughter of Charles VII of France. A later copy, now in Budapest, shows them as bride and groom. Ladislaus died suddenly, on 23 November 1457, just before their wedding.

(56) E. Buchner, Dos Deutsche Bildnis der Spatgotik und der fruhen Durerzeit, Berlin, 1953, fig. 202.

(57) Dulberg, op. cit., pp. 153-63.

(58) K. Cohen, Metamorphosis of a Death Symbol the Transi Tomb in the Late Middle Ages and the Renaissance, Berkeley, 1973, p. 44.

(59) Dulberg, op. cit., p. 154.

(60) For a thorough study of these tombs, sec Cohen, op. cit.

(61) E. Panofsky, Tomb Sculpture: Its Changing Aspects from Ancient Egypt to Bernini, New York, 1992 (lectures given in 1956), p. 55.

(62) Hall, op. cit., p. 46; see also p. 10, fig. 3.

(63) Panofsky, op. cit. in n. 61 above, p. 58, figs. 230, 232.

(64) Hall, op. cit., p. 114; Campbell, op. cit., p. 189: The dog--which resembles a Brussels griffon--was clearly a family pet.

(65) Hall, op. cit., p. 10. For Margaret of Austria's tomb, see Cohen, op. cit., figs. 75-77; Panofsky, op. cit. in n. 61 above, figs. 342-45.

(66) See Cohen, op. cit., p.45. There may also be a link to later double portraits with a dog at the foot of the woman, such as Tobias Stammer's full-length individual panels of Pannerherr Jacob Schwytzer von Zurich and Elizabeth Lochmann of 1564, in the Offentliche Kunstsammlung, Basel (for an illustration, sec M. Bendel, Tobias Stimmer: Leben and Werke, Zurich and Berlin, 1940, pp. 166-67, figs. 2-3.

(67) Campbell, op. cit., p. 190.

(68) Hall. op. cit., p.87.

(69) Campbell, op. cit., p. 187, who adds: 'Her mother Isabel de Sousa had been Lady in Waiting to Isabella of Portugal, Duchess of Burgundy, and Alienor had resided with her mother at the Burgundian court.'

(70) Ibid., p. 187.

(71) Ibid., p. 186.

(72) Ibid., p. 200.

(73) Huizinga, op. cit., p. 58.

(74) Ibid., p. 57.

(75) Campbell, op. cit., p. 187.

(76) Hall, op. cit., p. 117.

(77) A case in point is the representation of St Margaret on the right wing of by Hugo van der Goes's Portinari Altarpiere.

(78) Campbell, op. cit., pp. 186-87.

(79) Ibid., p. 188, 'Often described somewhat inaccurately as a Passion cycle, the ten miniscule medallions embellishing the frame include as well the Harrowing of Hell and the Resurrection, and thus the imagery really epitomizes the central Christian mystery of Christ's triumph over sin and death'. See also Hall, op. cit., p. 122.

(80) For Memling's Triptych of Earthly Vanity and Divine Salvation, (c. 1435), see Dirk de Vos, Hans Memling, exh. cat., Stedelijke Museum, Bruges, 1994, pp.112 15, no. 26.

(81) For Rogier van der Weyden's, Triptych of Jean Braque (c. 1452-53), see Dirk de Vos, Rogier van der Weyden: The Complete Works, Antwerp, 1999, pp. 268-53, no. 19.

(82) Campbell, op. cit., p. 201.

(83) Hall op. cit., p. 2.

(84) Campbell, op. cit., p. 201.

(85) Ibid., pp. 182-84.

(86) Hall, op. cit., p. 127.

(87) Panofsky, op. cit. in n. 1 above, p. 47.

(88) During the 1430s. Giovanni di Nicolao may have fallen on hard times, as documentary evidence concerning his business dealings significantly diminishes in these years. In 1452, he was still living in Bruges, but there is little evidence in the archives of financial or other successes. There is a second portrait by Jan van Eyck, in Berlin, of the man represented in the London Double portrait. Unfortunately, there is as yet no further information on the sitter's identity to be gained from the Berlin picture's provenance or history. See Campbell, op. cit., p. 194.

Margaret L. Koster is a graduate of the Courtauld Institute and of Columbia University, New York. She is the author of articles on Hugo van der Goes and on artistic interchange between Italy and the North. She is currently completing a monograph on the Portinari Altarpiece.
COPYRIGHT 2003 Apollo Magazine Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Koster, Margaret L.
Article Type:Critical Essay
Date:Sep 1, 2003
Previous Article:The Milan International Fine Art & Antiques Show Palazzo per le Belle Arti ed Esposizione Permanente Via Turati 34, Milan, Italy 5th-13th April 2003.
Next Article:A forgotten Assumption of the Virgin: the reredos at St Andrew, Sandford-on-Thames, Oxfordshire.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |