Ghent Altarpiece by Hubert and Jan van Eyck

Ghent Altarpiece

The Ghent Altarpiece was made to change what its audience could see. Shut, its wings present an Annunciation above kneeling patrons and two saints painted as sculpture. Open, they disclose a vast assembly around a bleeding Lamb, beneath enthroned sacred figures, musicians and the first human couple. The familiar photograph of the open work captures only one of these states. Hinges, frames and the movement between them belong to its meaning as much as its astonishing faces, flowers and jewels.

Completed in 1432 for the chapel of Joos Vijd and Elisabeth Borluut in Ghent’s church of Saint John, now Saint Bavo’s Cathedral, the polyptych remains in the cathedral for which it was commissioned. Its frame credits Hubert van Eyck with beginning the work and Jan with completing it, and associates its presentation with 6 May 1432. That testimony provides an essential starting point, although it does not settle the contribution of each painter. Nearly six centuries of worship, travel, separation, copying and conservation have also shaped the object encountered today. This overview follows those connections across the ensemble; the individual panel studies examine its components more closely.^[1]

A painting with two public faces

The cathedral’s guide gives an overall scale of approximately 3.4 metres high by 4.4 metres wide. More useful than imagining one immense flat canvas is recognising a constructed arrangement of oak paintings, some fixed and others hinged. The upper tier contains the enthroned group, angels and Adam and Eve; the lower tier gathers worshippers around the Lamb. Folding the wings replaces this outward spread with a narrower exterior. Traditional accounts emphasise opening on important feasts, when the interior’s strong colours would have produced an exceptional change. The exact timetable should not be assumed to have remained identical throughout the work’s long liturgical life.^[2]

Look first at the relationship between the two horizontal registers. Below, paths and groups of people carry the eye inward from both sides. Above, upright figures and tall compartments establish a much more measured rhythm. The Lamb stands close to the centre of a broad green expanse; the enthroned figure occupies the centre of a dense arrangement of precious surfaces. These are different kinds of pictorial space, held together by repetition, symmetry and the conspicuous vertical line through the middle. The work can be apprehended immediately as an ordered whole, then examined as hundreds of separate incidents of looking, reading, walking and prayer.

Its divisions remain visible even where the landscape appears continuous. A procession passes towards a panel edge and continues beyond it, while the painted horizon encourages the eye to cross the join. The frames interrupt that movement without destroying it. In the upper tier, by contrast, changes of setting become more emphatic: the first couple stand in narrow niches, the musicians cluster around their objects, and the central figures sit before decorated hangings. The painters did not disguise every compartment as a single seamless room. They used the physical divisions to establish relationships between different kinds of presence.

Founders, chapel and continuing worship

The commission belonged to a particular household and a particular place within Ghent. Frederik Buylaert and Erik Verroken’s French study reconstructs the founders’ social position through family history and transactions, challenging a simple picture of wealthy townspeople buying an impressive devotional object. Aristocratic standing, inherited relationships and the preservation of family memory mattered alongside financial resources. Their account connects sales and property arrangements in the 1420s with the couple’s ambitious patronage. Childlessness made the continuation of remembrance especially significant, but it should not be made into a complete psychological explanation of the paintings. The study’s value is to restore the work to a network of land, kinship and social representation: the chapel was an enduring institution as well as a setting for an extraordinary image.^[3]

The Belgian State Archives supplies a particularly tangible connection between those institutions and the founders’ estates. Its account of a 1446 inventory, drawn up after Elisabeth Borluut’s death in 1443, describes property whose income supported an annual memorial observance in the Vijd chapel. Land at Vracenbrouc, associated with earlier reclamation and investment, thus remained linked to prayer after both founders had died. This is an institutional description of an archival document, rather than a claim that the inventory itself explains every painted detail. It nevertheless changes the scale of the commission. The altarpiece’s survival depended on continuing arrangements for income and observance; the founders’ intentions extended beyond the moment when a finished painting could be unveiled and admired.^[4]

A daily Mass foundation followed on 13 May 1435. The conservation history published by KIK-IRPA records provision for two priests and a sexton, together with the practical requirements of light, bread, wine and ornaments. Such details place the retable above an altar that had to be staffed and supplied. They also complicate the familiar statement that the wings opened only on feast days. The same historical study considers possible opening during the daily foundation Mass, while acknowledging gaps in the evidence and later restrictions. There was a history of use, rather than one securely recoverable schedule for every century. The fixed panels and movable shutters served repeated actions whose rhythms cannot be reconstructed from a modern exhibition photograph.^[5]

On the exterior, the founders’ coloured clothing distinguishes them from the simulated stone saints between them. Their hands meet in prayer, and their bodies turn towards the sacred centre. Their portraits preserve individual features without telling us what their private thoughts were. Historian Jan Dumolyn cautions against reading Joos’s wrinkles or the shape of his mouth as evidence of hypocrisy, vanity or an unpleasant temperament. He also explains why devotional purpose and social prestige were compatible in the fifteenth century. That is a useful correction to an either-or account of patronage. A public demonstration of standing could support a sincere religious foundation; an unflattering-looking face does not reveal the sitter’s moral character. Their continuing visibility is itself part of the memorial project’s success.^[6]

Oak, paint and the work of two brothers

Jan’s documented career places this undertaking within a wider professional life. The National Gallery records his activity at the court of John of Bavaria from 1422 and his service to Philip the Good from 1425, including the Portuguese embassy of 1428. He was a court servant, traveller and painter whose work also reached patrons outside the ducal household. He died in 1441. Those dates matter when reconstructing the altarpiece’s development, but a journey cannot automatically explain a particular rock, plant or architectural detail in the painting. The documented career establishes opportunities and obligations; it does not supply a diary of the work’s execution. The distinction also prevents an attractive geographical resemblance from becoming an unsupported record of where either brother painted a passage.^[7]

The supports offer a different kind of chronology. Pascale Fraiture’s dendrochronological study establishes that a plank in Eve’s panel came from the same oak as a plank in the Hermits below. Three of the four measured sequences in Adam and Eve could be dated securely; a fourth remained less certain. The report gives 1412 as an earliest possible date for manufacture of the supports, while stressing that the interval between felling and painting is unknown. Baltic provenance links the wood to a wider trade in painting materials. A shared tree is evidence of material association across the ensemble, not a signature by Hubert or Jan. Equally, a tree-ring date cannot be read as the year when a particular face or landscape was painted.^[8]

The paint was built over a prepared surface rather than applied directly to bare oak. The technical account by Geert Van der Snickt and colleagues describes a chalk-and-glue ground, dark underdrawing and a translucent priming layer, followed by coloured underpainting and the final modelling and detail. This sequence allowed drawing, local colour and small finishing touches to perform different tasks. It helps explain how sharply defined ornaments coexist with soft transitions in skin or distant air. The work’s celebrated oil technique was an achievement of control and adaptation, not evidence that the brothers invented the medium. What survives is the product of successive decisions made above an absorbent preparation, with changes possible between a drawn plan, the first painted forms and the finished surface.^[9]

Research on the central lower panel now gives the division of labour a more substantial material basis. KIK-IRPA distinguishes an initial elaborate composition associated with Hubert from a second phase associated with Jan, who altered the landscape, introduced the fountain and revised other elements. The institute identifies a further phase involving such features as gilded rays and the enlarged dove, while treating the circumstances of that intervention separately. The argument combines paint structure, imaging, comparison and the frame inscription, whose authenticity was confirmed during the recent campaign. It is more precise than assigning every old-looking passage to one brother and every sophisticated one to the other. It also applies to a studied sequence in the lower composition; it is not a completed attribution map for every panel in the polyptych.^[10]

The upper register contains materials beyond ordinary paint: gilding, silver and tin-based relief decoration all require investigation. KIK-IRPA’s current project treats its seven panels as a distinct conservation phase. The components have different histories. The three central upper pictures travelled to Paris; most wings later passed through Berlin, where their painted faces were separated by sawing in 1894. Adam and Eve followed a different route through Brussels. Comparing passages therefore requires consideration of later interventions alongside original authorship.^[11]

The exterior: an announcement within architecture

The closed work rewards prolonged attention before it is treated merely as a prelude to the interior. Gabriel and Mary stand apart across an extended chamber. Between them are openings and domestic fittings; above, prophets and sibyls occupy the curved spaces under the frame. Below, donors flank the two Johns. Lukas–Art in Flanders’s account of the restored exterior emphasises discoveries in the original frames, including painted stone effects over silver leaf. The outside was never simply an absence of colour. Its restrained tonal range, coloured human figures, fictive sculpture and varied framing created a carefully graded display. Recent treatment recovered differences that earlier overpaint had flattened, making the architecture and its simulated substances more articulate again.^[12]

The Annunciation is an encounter conducted across an unusual amount of space. Gabriel’s body and raised gesture face into the room, while Mary turns from her reading at the opposite end. The emptier central compartments become active intervals between them. A towel, a vessel and the view outside give the chamber a particular density without placing another principal actor at its centre. The viewer must carry the greeting across those intervals, following a relationship that is stronger than the interruption of the shutters. The folds of pale clothing, the angles of the architecture and the lettering all contribute to the same movement through a room that is both intimate and conspicuously constructed.

Luke’s account supplies the announcement, the promised birth and Mary’s acceptance. It also gives the future child a place in Davidic kingship, connecting the domestic-looking event to a much larger history. The painted greeting and reply make speech visible, while the orientation of Mary’s words associates her response with the divine presence above. The image does not illustrate every moment of Luke’s chapter, nor should the Old Testament prophet in the upper compartment be confused with Zechariah, John the Baptist’s father, who appears in a separate annunciation earlier in the Gospel. Reading the scriptural passage identifies what is being announced; attending to the painting shows how the artists distribute the event across bodies, objects, light and physical divisions.^[13]

The two central figures below appear carved, yet the shadows around their feet and niches have themselves been painted. They stand beside donors whose flesh and clothing seem to belong to another category of reality. That difference is immediately visible without requiring a decision that one is less spiritually important than the other. The saints’ simulated material gives the donors something like sculptural company within a chapel, while the upper Annunciation opens a more expansive imagined space. In the shut state, the lower and upper zones offer distinct experiences of proximity: figures against niches below, a room stretching beyond the shutters above, and prophetic witnesses looking into the unfolding arrangement from its highest recesses.

Opening into a different kind of space

Opening reverses the balance. Green ground occupies the lower field, and the wide landscape accommodates groups moving from several directions. Strong red, blue and green give the upper figures great visual weight. Gold-coloured ornaments and light-catching surfaces appear against darker passages. The central line is conspicuous, but it does not force every person into an identical pose. Some kneel; some read; others approach on foot or horseback. The broad organisation is stable enough to hold this variety together. Looking across the work is therefore different from following one procession down a street: several routes converge upon a centre that already seems timeless and complete.

The upper and lower halves also ask for different bodily responses from the viewer. The large enthroned figures are encountered frontally, at a scale that makes their jewellery and patterned textiles visible as parts of monumental bodies. The smaller people below invite scanning, comparison and recognition. A small face can be observed as closely as a large one, but it does not command the same immediate distance. Adam’s projecting foot at the upper edge adds another complication, apparently bringing a body across the boundary of its compartment. Throughout the ensemble, scale and framing govern how near a figure seems, even when the eye knows that every figure occupies a painted surface.

The inscriptions extend this structure into language. Weale’s early twentieth-century account transcribes the wisdom language around Mary, the Baptist’s accumulated titles and the fountain’s identification as life-giving water proceeding from the throne of God and the Lamb. These are different verbal functions: praise, identification and a statement linking one pictured object to a larger sacred order. They encourage connections across the compartments rather than merely supplying names beneath portraits. Weale remains a useful historical witness for their transcription and for the ensemble before the 1934 theft, although his older attributions and conjectural identifications require separate assessment. An inscription should be read in its location: above a head, along clothing or around a basin, its position helps determine what the words do.^[14]

The first couple and the cost of violence

Adam and Eve occupy the outer ends of the upper register. Their skin is presented against dark niches, while the surrounding company is heavily clothed. Hands and leaves cover their bodies, and Eve holds a small fruit. Above them, two miniature scenes appear as stone reliefs. The contrast between the broad expanses of flesh and the tiny narratives above is particularly striking. Bodies that seem nearly available to touch stand beneath events compressed into narrow architectural spaces. The painted frame thus encloses both individual presence and a history extending beyond those individuals; the first couple cannot be understood only as demonstrations of anatomical skill detached from the rest of the work.

Genesis 4 supplies the story of their sons’ offerings and the killing of Abel by Cain. The text distinguishes their occupations and describes the acceptance of one offering and rejection of the other, without giving the simple explanation that one brother presented a more expensive gift. Its warning about sin precedes the murder, and the questioning of Cain follows it. In the small painted scenes, offering and violence become two selected moments within that larger narrative. They introduce death and failed brotherhood above the assembly gathered around the Lamb. The comparison is visible within the altarpiece’s arrangement, but the biblical passage should still be allowed its own complexity rather than reduced to a formula about generous and ungenerous worshippers.^[15]

A historical interpretation by L. Aerts makes the connection explicitly typological: Abel, as innocent victim and giver of a lamb, foreshadows Christ. Writing in 1920, Aerts argued that the violent scene did not disrupt the sacred ensemble but reinforced its account of suffering and salvation. His explanation is valuable as a documented stage in the work’s interpretation, especially because he contested readings that treated everything in the upper register as if it occupied one literal heavenly location. It is not a recovered statement of the brothers’ intentions. Nevertheless, it articulates a relationship the painting makes available: the little image of an innocent man’s death above Eve can be considered alongside the central Lamb whose blood is shown below.^[16]

Linda Seidel’s study asks a related question about naturalism itself. Adam’s reddened, weathered hands have often been taken as proof of a closely observed living model. Seidel instead considers what those hands contribute to a representation of labour, bodily knowledge and humanity after Eden. She also draws attention to the conflation of moments: Eve retains the fruit although the couple already cover themselves with leaves. On this account, the figures are deliberately constructed statements rather than neutral records of two models standing in a studio. The argument is an interpretation, not a medical diagnosis of the painted bodies. It helps explain why the altarpiece can combine highly particular appearances with a sequence of sacred events that does not unfold in ordinary narrative time.^[17]

A Lamb at the centre of worship

Revelation 5 describes a Lamb standing despite having been slain, receiving the scroll and the worship of a universal assembly. It is an image of victory inseparable from sacrifice. The Ghent painting transforms that vision: its Lamb stands on an altar, and a stream of blood enters a chalice. It does not reproduce the scriptural creature’s seven horns and seven eyes as a literal inventory. The contrast demonstrates how closely related images can be without being interchangeable. Revelation provides the triumphant sacrificial Lamb and the expanding company of worshippers; the painting gives that account a particular bodily appearance and places it within a setting whose altar and vessel invite further liturgical associations.^[18]

The Gospel of John provides another strand. John the Baptist identifies Jesus through the Lamb of God designation and testifies to the Spirit’s descent in the form of a dove. That scriptural relationship helps explain why the Baptist has a substantial place in an ensemble centred upon the Lamb. The two Johns on the exterior should nevertheless remain distinct persons and pictorial roles: the Baptist is the prophetic witness associated with Christ’s coming, while the Evangelist is the saint traditionally connected with the Gospel and the apocalyptic vision. Reading their attributes and positions together makes the outside more than a line of unrelated devotional figures. It prepares witnesses and language that become central when the shutters disclose the inner assembly.^[19]

Susan Jones’s Metropolitan Museum account relates the ensemble to the liturgy of All Saints and to the Eucharist celebrated before it. This connects the diversity of holy figures with the actual bread, wine, vessels and actions of worship in the chapel. The painted altar did not operate in isolation from the real one beneath the retable. The relationship helps explain the chalice receiving blood and the fountain’s life-giving imagery, while leaving room for more than one textual source. Jones also stresses that no single explanatory programme has resolved every element. The work’s coherence does not depend on finding one book whose sentences can be assigned mechanically to all its figures; liturgical practice and several traditions of sacred writing overlap within the image.^[20]

At the centre, the Lamb remains very small beside the scale of the whole assembly. Its isolation on the altar makes that size effective. The surrounding angels form a close ring; more distant groups occupy the meadow; the viewer looks past the fountain towards them. The blood is represented as a narrow, definite connection between body and vessel, rather than a stain spread through the scene. White fleece, pale cloth, a red altar covering and reflective metal establish sharp local contrasts inside the larger green field. The painters concentrated attention through position and difference, allowing a relatively small figure to organise an extraordinarily extensive gathering.

The enthroned figure and the limits of a label

The identity of the large central figure above has prompted enduring disagreement. Maximiliaan Martens sets out two important relationships: read horizontally beside Mary and the Baptist, the figure recalls the Deesis, with Christ enthroned between intercessors; read vertically above the dove and Lamb, it can participate in a Trinitarian arrangement with God the Father at the top. Other scholarship considers a deliberately inclusive or ambiguous representation of divinity. These are interpretations of combined visual and textual evidence, not alternative names selected at random. The attributes support several connected theological claims. Calling the figure simply Father or simply Christ without qualification would conceal the question that makes its position within the whole work so important.^[21]

Visually, the central figure differs from its neighbours through both scale and display. Its red garment spreads across the body and gathers into a mass of folds below. The headdress, sceptre, jewellery and patterned hanging produce a dense concentration of ornament. Mary bends towards her book; the Baptist turns and gestures. The contrast is between a frontal, commanding centre and figures whose inclinations direct attention into a relationship with it. Yet the rich clothing also joins them. Pearls, embroidered borders and shimmering backgrounds recur across the three compartments, providing a material continuity that is easier to observe than to convert into a single unambiguous theological title.

Watch: Jan van Eyck in context

Choose a Smarthistory close reading of the Arnolfini Portrait or the closed and opened Ghent Altarpiece.

Jan van Eyck, The Ghent Altarpiece (2 of 2)

Smarthistory

Follow the heavenly figures and the gathering around the Lamb when the wings open.

Watch on YouTube

▶03 · InteriorThe Ghent Altarpiece, openSmarthistory▶01 · PortraitThe Arnolfini PortraitSmarthistory▶02 · ExteriorThe Ghent Altarpiece, closedSmarthistory

The cloths behind Mary and the Baptist have generated another interpretation. Hélène Verougstraete and Wim Verbaal propose that damaged lettering in the applied decoration refers to the deceased Hubert, whom they read as Lubrect, and that it supports a memorial understanding of the upper group. Their reconstruction combines repeated motifs, technical images and difficult signs. It should be distinguished from the securely accepted frame testimony naming the brothers and from the current conservation project’s continuing investigation of upper-register authorship. A proposed reading of damaged ornament cannot silently replace the established chronology of Hubert’s death in 1426. Its significance lies in the question it opens: whether remembrance of a painter could have been incorporated into sacred decoration as well as recorded on the frame.^[22]

Books, prophecy and Marian presence

Mary appears twice in radically different circumstances. Outside, she responds within an architectural interior; inside, she is a richly dressed reader beside the enthroned figure. Her book remains an active part of both appearances. In one state she turns from reading towards an announcement; in the other her downward attention creates a quiet counterweight to the frontal centre. The difference can be seen in the placement of hands, the inclination of the head and the weight of the clothing. Opening the shutters changes not only the number of figures visible but the kind of presence granted to a person already encountered on their exterior.

Grantley McDonald proposes Bridget of Sweden’s Revelations as a further source for the altarpiece’s imagery and inscriptions. His study follows connections between words, figures and the sacramental action at the real altar, rather than treating each inscription as an isolated quotation. Such an approach is useful because a sacred name on one object can echo a name elsewhere, directing the eye across the composition. McDonald’s argument concerns a contribution to an intricate programme, not an exclusive key that displaces biblical and liturgical sources. The survival and circulation of Bridget’s writings in the Low Countries make the proposal historically meaningful, while the lack of a surviving commission document prescribing the complete imagery still matters when judging how securely a specific connection can be demonstrated.^[23]

Angela Maria La Delfa develops the Marian aspect of that comparison. She relates the Virgin’s appearance and attributes to Bridget’s descriptions and considers how Marian privileges and devotion entered the visual language of the early fifteenth century. Her account places the altarpiece within a receptive environment for female mystical writing in Flanders and Brabant. The argument extends beyond resemblance in an individual feature: hairstyle, inclined pose, clear complexion and the Virgin’s exalted status are considered together. These relationships remain scholarly interpretations of images and texts. They do not justify projecting later definitions of every Marian dogma unchanged into 1432. They do, however, enlarge the devotional context in which the two representations of Mary could have been understood.^[24]

Hearing the image

The angel panels provide a particularly vivid distinction between seeing evidence of music and recovering an actual performance. In her study of the 1435 foundation, Barbara Haggh connects the priests to the church’s continuing body of singers and to the timetable of worship. She finds an institutional place for musical skill, including the later service of a priest known as a tenor singer. Crucially, she also states that there is no evidence that polyphonic singing was prescribed or included in the daily Mass before the Lamb. Painted open mouths and instruments cannot fill that documentary gap. The foundation made sound part of the chapel’s life, but the specific music of a fifteenth-century service cannot be reconstructed merely by looking at the angels.^[25]

Laura Slater’s Oxford study examines the pictorial relationships more closely. The singers gather at a lectern decorated with Saint Michael defeating evil; the instrumentalists occupy a related setting with patterned tiles beneath them. Their frame texts concern praise and thanksgiving, including the call to praise with strings and organ. Such details bring sound into the ensemble’s larger account of heavenly worship. The figures’ concentration and differing mouths make their activity unusually persuasive, but the image does not become a complete recoverable score. The important connection is between the represented act of praise, the visual objects through which it is organised, and the possibility of worshippers hearing their own prayers or music in the chapel below.^[26]

Observe how the two groups divide attention. In one panel, several faces turn towards a shared open book; in the other, hands work across instruments whose forms occupy a substantial part of the available space. Clothing overlaps, and bodies stand close enough for small differences of gesture to become noticeable. The repeated concentration creates a visual equivalent of coordinated action without making every participant identical. At this scale, an eye directed downwards or a finger resting near an edge can change the rhythm of the group. Those distinctions help the upper register remain animated even where the figures neither travel through landscape nor take part in a conventional unfolding story.

Sacred lettering and a composite natural world

The lettering requires the same care as the music. Jacques Paviot’s investigation, with Dina Goren’s contribution, distinguishes Greek or Hebrew forms from Latin expressions written in unusual characters. On the central robe, royal and divine titles can be recognised beneath inventive letter substitutions; elsewhere sacred names and the formula AGLA belong to a Christian history of devotional and protective usage. Their presence is not straightforward proof that the painter read Greek and Hebrew fluently. Nor is every difficult mark meaningless ornament. A painted word can combine semantic content, learned appearance and decorative effect. Close comparison of the actual forms is more reliable than assuming that any unfamiliar alphabet supplies a hidden biographical message or a simple cipher awaiting solution.^[27]

The landscape likewise combines observation with construction. The City of Ghent’s botanical account notes plants belonging to different seasons and climates within the same setting. Spring flowers do not oblige the viewer to date every other plant to a single spring morning. Mediterranean forms can stand beside vegetation familiar farther north. The abundance creates a place of continuing fertility rather than a survey of one surviving meadow. This distinction leaves the painters’ observational precision intact: a carefully described leaf or flower can be recognisable even when its neighbours would not grow or bloom beside it. It also cautions against converting the whole landscape into an exact map of a journey, with every exotic species treated as a dated entry in Jan’s travels.^[28]

Eve’s fruit demonstrates the limits of botanical certainty. It is often called an etrog or citron, but Diego Rivera and colleagues, studying citrus history and identification, argue that its size and form are more consistent with a limetta. Their comparison addresses a recognisable painted shape; it does not supply a specimen that can be tested. The alternatives should therefore remain visible rather than concealed beneath the familiar word apple. For the overall programme, the important point is that a particular-looking fruit participates in the account of the first couple. Its exact identity can sharpen interpretation, but should not carry an elaborate argument as though a scientific consensus had settled every feature of the object in Eve’s hand.^[29]

Within the lower landscape, small plants interrupt the green surface with individual shapes. Some form upright stems; others spread close to the ground. Leaves catch light differently from cloth, stone or flesh, and the small contrasts make the meadow appear inhabited before the eye begins to distinguish the human groups. Farther away, the visual scale contracts into trees, buildings and a pale horizon. The same surface thus offers near inspection and distant extension. Moving between those scales is one of the pleasures of the painting: the viewer can examine a flower as a local event and then return to the broad movement of the gathering around the altar.

Early admiration and the evidence of copies

Lucas de Heere’s ode, published in 1565, records an encounter with a work already famous. He praises the extraordinary flesh of Adam, the apparent distinctions between the angels’ voices, the varied faces and the enduring colours. He also identifies supposed portraits of the brothers among the riders. The poem is evidence of what a sixteenth-century admirer believed and valued, rather than proof of those portrait identifications. Its emphasis is revealing: bodily presence, music, variation and colour mattered alongside the sacred subject. De Heere asks the reader to look closely at an achievement whose separate effects seem to exceed the normal capacity of painting. His praise shows that the altarpiece’s minute particularity was recognised long before modern magnification made it convenient to examine online.^[30]

A different kind of testimony survives in Michiel Coxcie’s copy, commissioned by Philip II and made in 1557–59. M Leuven’s account of its 2013 reunion describes Coxcie working for an extended period within a wooden enclosure at the Vijd chapel. The project reproduced the ensemble on an ambitious scale, but also adapted it to a later artistic language and included new identifications and portraits. Its subsequent dispersal among collections in Brussels, Berlin and Munich gives the copy its own complicated history. It is indispensable as a witness to the original before later losses, yet cannot be treated as a colour photograph made in the sixteenth century. Comparison must ask what Coxcie saw, what he changed and what later happened to his own paintings.^[31]

The Royal Museums of Fine Arts of Belgium’s record for Coxcie’s Judges panel makes that distinction particularly concrete. Its reverse depicts John the Evangelist, and the catalogue explicitly identifies the four Evangelists as exceptions to the copying of the original shutters. A later viewer who assumes every copied exterior figure duplicates the 1432 programme would therefore make a basic error. The double-sided wooden painting also preserves the practical logic of a wing, with two states carried by one support. As a museum object it has dimensions, acquisition and deposit histories of its own. A useful comparison concerns both imagery and construction: a copy can transmit the organisation of the original while altering what appears within that organisation.^[32]

Dispersal changed the work itself

The City of Ghent’s historical account describes the altarpiece being hidden during the iconoclasm of 1566 and subsequently kept at the town hall before its return to the church in 1584. Protection could require dismantling and removal from the very setting that gave the painting its function. A later cathedral fire in 1822 prompted another emergency evacuation, during which the central lower panel split. These episodes are more than dramatic interruptions in an otherwise continuous life. Each movement required people to handle, pack or separate vulnerable supports. A painting conceived as a joined and hinged ensemble could survive only by temporarily ceasing to operate as one. Survival therefore left physical consequences as well as stories of rescue.^[33]

The nineteenth-century separation was particularly consequential. The central paintings returned from Paris after the Napoleonic period, while most wings were sold in 1816 and reached the Prussian royal collection in 1821. In Berlin their fronts and backs could eventually be displayed simultaneously because the supports had been sawn apart. Adam and Eve were excluded from that sale and later acquired by the Belgian state. The present arrangement brings together parts that once followed different national and institutional routes. It also conceals, in a frontal photograph, the fact that some shutters no longer possess their original physical relationship between two painted sides. Reassembly restores a recognisable composition without undoing the cutting of its wood.^[11]

The nude figures have a separate history of reception. Martens treats the story that Joseph II was offended by them in 1781 as a rumour, rather than a securely documented motive that should be repeated as fact. The later clothed substitutes by Victor Lagye are tangible evidence of changing expectations. Their animal-skin coverings altered how the first couple could be presented while the original figures remained outside the assembled display. Such replacements remind us that an audience’s acceptance of nudity is historically variable. The same carefully described bodies could be admired as feats of painting, withheld as unsuitable or replaced by versions adjusted to a different idea of religious decorum. Those responses belong to the work’s history without determining what its first viewers thought.^[34]

War, restitution and the contested national treasure

The Second World War carried the altarpiece far beyond Ghent again. Greg Bradsher’s account for the US National Archives traces its evacuation to Pau in 1940 under the care of Jacques Jaujard, its seizure and movement through Neuschwanstein in 1942, and its later storage at Altaussee. Following recovery, the work travelled through the Munich collecting point and was flown to Brussels on 21 August 1945; the Belgian handover followed on 22 August. These dates should be distinguished from subsequent ceremonial presentation. The account draws on military and restitution records, making clear that return involved packing, transport, responsibility and negotiation. The paintings did not simply reappear in the cathedral as soon as the mine was reached by Allied forces.^[35]

Jenny Graham’s study examines how that return intensified the altarpiece’s position in Belgian cultural and political life. It also confirms that four central panels had been taken to Paris in 1794, a history broader than the three upper panels discussed in the current upper-register project. Graham follows the controversy surrounding the 1950–51 conservation in Brussels: public anxiety about renewed removal, the authority of scientific treatment and fears concerning ownership and return. International participants included specialists from the Louvre, Rijksmuseum and National Gallery alongside Belgian and American colleagues. Conservation was consequently also a negotiation over who could decide the future of a church-owned work regarded as a national treasure. Her analysis shows how a material treatment became inseparable from postwar anxieties about identity, institutional authority and the continued place of the altarpiece in Ghent.^[36]

The missing Judges and the surviving replacement

The theft of 10–11 April 1934 removed the Just Judges and the exterior John the Baptist. The Baptist was recovered that May, but the Judges remain missing. KIK-IRPA’s French dossier distinguishes the evidence concerning Arsène Goedertier from a conclusive account of exactly who carried out the crime. It also gives a more complicated chronology for Jef Van der Veken’s replacement than the single date often attached to it. He began painting in 1939, resumed in 1945 and made further adjustments around the 1950–51 treatment. The visible replacement belongs to that later history. It restores the position of the riders in the ensemble, but it does not make the original painting materially present or remove the need to distinguish old photographs, Coxcie’s witness and the modern copy.^[37]

Van der Veken’s methods also entered a debate about how early Netherlandish painters worked. KIK-IRPA’s Dutch account describes his egg-based reconstruction and the subsequent confirmation, through the 1951 examination, of an oil-based binder in the original. Close visual imitation had not by itself recovered the brothers’ technique. This is an especially useful warning in an account of such convincing surfaces: matching an appearance and matching its manufacture are different achievements. The replacement can preserve a composition, help the ensemble function visually and remain a distinct twentieth-century object. Its own materials require care, and their ageing need not follow the behaviour of the fifteenth-century paintings around it. The technical difference is part of the evidence, not a reason to deny the replacement’s historical interest.^[38]

The RKD’s account of the Van der Veken archive identifies drawings, paintings, documentation and working equipment as resources for investigating that practice. The archive matters because the same skilled individual worked across copying, restoration and the production of objects that entered the market with misleading identities. These activities should not be collapsed into a single category. The acknowledged Ghent replacement has a different status from an object sold as an undisclosed old master. Records of working methods can clarify how a convincing surface was produced and help distinguish the interventions of a restorer from older paint. They also allow the replacement to be studied through evidence of its own manufacture, rather than treated only as a regrettable blank filled by an anonymous modern hand.^[39]

Looking beneath the admired surface

The present research campaign began with questions of physical survival as well as appearance. Getty’s project history records condition assessment and emergency stabilisation in 2010, followed by support for training and technical documentation. Anne van Grevenstein at the University of Amsterdam and Ron Spronk, associated with Queen’s University and Radboud University, were among the project leaders. The public Closer to Van Eyck platform launched in 2012. This international structure brought structural care, conservation experience and imaging into one sustained investigation. Its importance was not simply the production of larger photographs. It created records through which decisions could be compared and assessed, while helping train people responsible for the future care of paintings on wood.^[40]

Getty’s account of the expanded platform explains the importance of comparing images made before, during and after treatment. Visible-light photographs can be studied beside infrared and X-ray material, with several views available together. Its report that roughly seventy per cent of the closed exterior’s original paint had been obscured by overpaint has a specific scope: it is not a percentage for the entire altarpiece. Such distinctions matter when a dramatic discovery enters general circulation. The platform’s sequence also prevents a cleaning-stage image from being mistaken for a finished result. A dark patch, pale gap or unevenly treated garment has to be understood in relation to when the photograph was taken and which layers had then been removed.^[41]

Marie Postec and her colleagues describe how the recognition of extensive overpaint emerged gradually during treatment. Old varnishes had masked both condition and intervention, while the prior fame of the work made the possibility of widespread repainting difficult to accept. Certain reddish fills transmitted X-rays rather than standing out as obvious dense losses; familiar imaging alone had therefore not revealed the whole problem. Close observation, cross-sections and newer chemical methods had to be interpreted together. The account is especially valuable because it gives the trained conservator’s eye a central role without treating expertise as infallible. The investigation advanced by testing surprising observations, revisiting assumptions and comparing forms of evidence whose limitations were different.^[42]

The University of Antwerp’s Phase I account explains how macro X-ray fluorescence scanning helped map the exterior. Measurements across more than eight square metres were collected over sixty working days, with the surface divided into thirty-seven scanning areas. Elemental distribution maps revealed forms and losses hidden by later paint, while microscopic examination of cross-sections helped locate those signals within the layer structure. A map of lead, mercury or iron is therefore evidence about materials, not an ordinary photograph of a complete hidden painting. Its interpretation depends on stratigraphy and other observations. This combination supplied a stronger basis for assessing the surviving original than a visual impression that the old surface simply looked dark or disappointing.^[43]

Why removing paint required a separate decision

Martens’s discussion of conservation ethics makes the conditions explicit. A suspected later layer must first be distinguished from a change made by the original artist. Its date, purpose and relationship to damage require investigation; the possibility of safely revealing what lies beneath also matters. The desired result cannot be a literal return to 1432, since ageing, cracks and other irreversible changes remain. In the exterior’s case, intermediate varnish and the condition of the older paint helped make removal possible. Scientific imaging strengthened the assessment, but did not eliminate historical and aesthetic judgment. The question was not whether all old additions should automatically disappear. It was whether a particular intervention could recover significant original work without imposing an unacceptable risk upon it.^[44]

The Lamb’s changed face is a clear example of this distinction in practice. KIK-IRPA explains that recent treatment exposed the smaller, forward-looking face beneath later overpaint. The earlier partial cleaning of the 1950s had left the much-discussed appearance of additional ears. Investigation also identified a smaller earlier version of the body, but the enlarged contour was retained because the evidence did not securely establish whether that early alteration belonged to the original painters or another intervention. Face and body therefore did not receive an identical treatment. The decision followed differences in evidence. It would be misleading to present every change as newly painted by restorers, or to assume that detecting an older form necessarily required uncovering it.^[45]

Details from Ghent Altarpiece

1

Compare the Lamb’s face during conservation

Published 2018 treatment comparison: overpainted face at left, exposed earlier face at right. KIK-IRPA, 19 June 2018. Saint Bavo’s Cathedral © Art in Flanders vzw, photo KIK-IRPA. Complete published comparison, unchanged.

The difference remains important when looking at the restored scene. The Lamb’s direct gaze now contributes strongly to the encounter with the viewer, but its significance comes from the exposed painting and its context, not from a modern desire to make an animal more expressive. The surrounding cloth, altar, angels and chalice still organise the same concentration of attention. A comparison photograph can demonstrate the alteration of visible features; the conservation record explains why those features could be revealed. Both are needed. Without the record, an unfamiliar face is easily mistaken for a new invention; without looking at the surface, the technical account loses the visual event that made the discovery consequential.

Digital research and its boundaries

Some supports introduce an additional difficulty: an X-ray records both painted sides together. Zahra Sabetsarvestani and colleagues’ 2019 study uses a neural-network method to separate information in examples from Adam and Eve and the exterior scenes on their reverse. The aim is to make a superimposed radiograph more legible for investigation. It is not to generate a replacement artwork or discover a signature that was never recorded. The resulting separated images are computational estimates that need comparison with the measured data and the known paintings. Their usefulness lies in helping specialists distinguish overlapping structures and ask better questions about condition and technique, while keeping the difference between a physical measurement and a processed interpretation visible.^[46]

Digital crack removal raises a related problem at the scale of the Virgin Annunciate’s book. Aleksandra Pižurica and colleagues explain that cracks can resemble parts of letters, while paint lifting and abrasion affect the surrounding image. An automated repair may improve apparent legibility yet accidentally remove a real stroke of writing or extend a false one. Their research tests methods intended to preserve the continuity of meaningful structures. The lesson for interpretation is practical: a cleaner-looking virtual page is not automatically more truthful than the damaged photograph. A proposed reading must be checked against the unaltered evidence. Digital tools can support the study of inscriptions without converting an inferred completion into surviving fifteenth-century paint.^[47]

Precious surfaces are physical materials

The applied brocades of the upper register make the relationship between illusion and substance unusually complex. Sophie Kirkpatrick and colleagues’ June 2026 preprint studies tin relief decoration through complementary imaging and material analysis. Much of the original tin has altered into corrosion products, including romarchite and cassiterite, with consequences for cracking and separation of the decorative structure. Proposed additional mechanisms, including possible organic tin compounds and interactions involving gilding, remain subjects for further investigation. The paper should be read as a preprint, with those distinctions intact. Research involving French synchrotron facilities is part of this wider effort to understand how the decoration was made and why it has changed. Conserving an imitation textile may require treating an actual layered metallic relief.^[48]

Conservator Kathleen Froyen’s account of three treatment details brings the consequences back to visible form. She describes irreversible alteration in ultramarine and the role of careful retouching in recovering the legibility of a blue robe’s volume. She also explains the difficulty of fragmented tin reliefs and the investigation of floors whose effects depended on transparent colour over metal leaf. These are different problems; removing discoloured varnish cannot solve them all. Her account distinguishes what treatment can reveal from what it can only help the viewer read more coherently. The upper register’s brilliance was produced by several physical systems, and their survival requires decisions adapted to each material rather than one uniform cleaning procedure.^[49]

A dated KIK-IRPA update from 10 December 2024 provides a useful snapshot. At that stage, approximately half the overpaint on the three central cloaks had been removed, while work on Adam and Eve found their original painting comparatively well preserved beneath varnish and retouching. Research on the angels and other upper components continued alongside it. That statement should not be extended to mean that half the entire altarpiece had been treated, or presented as a description of the project’s state in October 2026. Progress reports are evidence at particular moments. Their value increases when the photograph, date, area and stage of treatment remain attached to the result being described.^[50]

The current timetable and the cathedral setting

As checked on 29 October 2026, KIK-IRPA’s revised funding announcement distinguishes two forthcoming milestones: treatment is scheduled to finish in December 2026, while the fully restored ensemble is expected to return to public display in spring 2027. Both remain planned dates. The announcement links additional time and funding to the extent and complexity of overpaint and to the research required for its removal. It should not be rewritten as a claim that the project has already finished. The distinction matters for interpreting contemporary photographs as well as planning a visit: a published view of the complete open ensemble may predate the latest treatment even when it accompanies an up-to-date article about the research.^[51]

The display itself is another historical intervention. Architect and art historian Maaike Blancke explains the choice of the Sacrament chapel and the need for a protective case, controlled conditions and improved access around the work. The new arrangement keeps the altarpiece in the cathedral while addressing requirements the smaller original chapel could not accommodate. She also describes lighting used to reduce the apparent colour mismatch between the Van der Veken replacement and the newly cleaned adjacent paintings. That adjustment is part of presentation, not proof that the copy has become identical to the originals. Temperature, light, security and visitor access now mediate the encounter just as architecture, opening practices and the performance of worship did in earlier centuries.^[52]

The most rewarding way to study the ensemble is consequently to move between kinds of evidence. The open view reveals a network of scale, colour and converging movement; the closed view establishes different relationships between room, niche, prayer and announcement. Surviving documents explain what supported worship, while the wood and paint preserve traces that documents cannot supply. Copies and old photographs retain information about appearances that have changed or disappeared. None substitutes completely for the others. Together they allow the altarpiece to be understood as a functioning sacred object, an achievement of painting and a material survivor whose history remains visible in the very surfaces through which its extraordinary world is encountered.

Related reading

Gabriel: the announcing angel on the Ghent Altarpiece

Adam on the Ghent Altarpiece

Eve on the Ghent Altarpiece

References

1. Saint Bavo’s Cathedral, The Ghent Altarpiece, installation and frame attribution. View Source
2. Saint Bavo’s Cathedral, panel guide, closed exterior, section I and general introduction. View Source
3. Frederik Buylaert and Erik Verroken, Un retable noble : le contexte socio-historique de l’Agneau mystique, in Danny Praet and Maximiliaan P. J. Martens, editors, L’Agneau mystique – Van Eyck. Art, Histoire, Science et Religion, Paris, 2019, pp. 50–115, especially pp. 52–58 and associated notes; author-posted text. View Source
4. Archives de l’État en Belgique, Faites la connaissance de Joos Vijd, le maître d’ouvrage du célèbre Agneau mystique, 3 June 2020. View Source
5. Bart Fransen and Cyriel Stroo, editors, The Ghent Altarpiece: Research and Conservation of the Exterior, KIK-IRPA, 2020, pp. 12–14, foundation and historical opening practices. View Source
6. Jan Dumolyn, The pimples of Joos Vijd, City of Ghent, 28 November 2025. View Source
7. National Gallery, London, “Jan van Eyck”: documented career, origins and range of patrons. View Source
8. Pascale Fraiture, dendrochronological analysis of Adam and Eve (2013), conclusion, p. 25. View Source
9. Geert Van der Snickt and colleagues, Dual mode standoff imaging spectroscopy documents the painting process of the Lamb of God, Science Advances 6, eabb3379, 2020, pp. 1–2, original painting structure. View Source
10. KIK-IRPA, “The mystery of Hubert van Eyck and the Ghent Altarpiece solved”: lower-register phases and division of work. View Source
11. KIK-IRPA, “Restoration and research of the upper register”: materials, differing panel histories, research and current schedule. View Source
12. Lukas–Art in Flanders, The real Van Eyck revealed, restored exterior and original frame effects. View Source
13. USCCB, Luke 1:5–38, the two announcements by Gabriel. View Source
14. W. H. James Weale, The Van Eycks and Their Art, 1912, pp. 40–54, Ghent Altarpiece descriptions and inscriptions. View Source
15. United States Conference of Catholic Bishops, Genesis 4:1–10 and explanatory notes on 4:1–16 and 4:4. View Source
16. L. Aerts, Het lam Gods der gebroeders Van Eyck. De hoofdfiguur, 1920, pp. 1076–1077 and associated note on Maastricht. View Source
17. Linda Seidel, “Adam and Eve: Shameless First Couple of the Ghent Altarpiece,” Different Visions, 2008: hands, labour, artistic making and inherited gestures. View Source
18. USCCB, Revelation 5:1–14 and notes. View Source
19. USCCB, John 1:19–36 and notes. View Source
20. Susan Jones, “The Ghent Altarpiece,” Metropolitan Museum of Art, 2002: arrangement, chapel light, liturgy and programme; attribution discussion supplemented here by later KIK-IRPA research. View Source
21. Maximiliaan Martens, “What is depicted in the Ghent Altarpiece?”, Visit Gent, 15 August 2025: horizontal and vertical readings and divine-mystery proposal. View Source
22. Hélène Verougstraete and Wim Verbaal, The Ghent Altarpiece and the Inscription on the Cloths of Honour behind Mary and John the Baptist, 2022: a proposal concerning the neighbouring backgrounds. View Source
23. Grantley McDonald, A Further Source for the Ghent Altarpiece? The Revelations of Bridget of Sweden, Oud Holland 128, 2015, pp. 1–16, especially opening discussion and pp. 2–3. View Source
24. Angela Maria La Delfa, The Virgin in the Ghent Altarpiece and the Revelations of Saint Bridget of Sweden, De Medio Aevo 14, 2020, pp. 83–98, introduction and conclusion. View Source
25. Barbara Haggh, “The Mystic Lamb and the Golden Fleece,” Revue belge de Musicologie, 61, 2007, pp. 5–59, especially 8–9, foundation and musical evidence. View Source
26. Laura Slater, “Singing to Heaven in the Ghent Altarpiece,” MALMECC, University of Oxford, 2019, lectern, frame inscription and heavenly worship. View Source
27. Jacques Paviot, with Dina Goren, “Les inscriptions grecques et hébraïques de Jan van Eyck,” 2006, especially pages 64 and 68–69: pearl lettering, royal titles and Hebrew-derived divine names. View Source
28. Visit Gent, “A heavenly garden,” botanical identifications and combined seasons. View Source
29. Diego Rivera and colleagues, “Analysis of ‘Marrakesh limetta’,” Scientia Horticulturae 293, 2022, 110688, especially section 3.5.1: botanical comparison and the sour-limetta proposal. View Source
30. Lucas de Heere, Den hof en boomgaerd der poësien, 1565, ode XI, pages 29–31, in W. Waterschoot's 1969 edition: a sixteenth-century poetic response to Adam. View Source
31. M Leuven, French account of the reunited Coxcie copy, 1 October 2013: commission, adaptation and the copied Deity's later collection history. View Source
32. Royal Museums of Fine Arts of Belgium, Michiel Coxcie, Judges with Saint John the Evangelist on the reverse, 1557–1559, inventory 6696. View Source
33. City of Ghent, The adventures of the Ghent Altarpiece. View Source
34. Maximiliaan Martens, “The naked truth about Adam and Eve,” Visit Gent, 24 October 2024: reception, clothed substitutes and photographs after varnish removal. View Source
35. Greg Bradsher, The Monuments Men in August 1945: The Belgian Treasures, US National Archives, The Text Message. View Source
36. Jenny Graham, The Ghent altarpiece after World War II: Restitution, restoration, and redemption, International Journal of Cultural Property 28 (2021), pp. 343–367; published online 25 February 2022. View Source
37. KIK-IRPA, Rendez-vous avec les Juges intègres, 4 June 2010, pp. 3, 7–9, reconstruction chronology and urgent conservation. View Source
38. KIK-IRPA, De zaak Van der Veken, 2005, pp. 5–6, egg-based reconstruction and reassessment of early Netherlandish technique. View Source
39. RKD, “Echt nep of net echt? Vervalsingen in de schilderkunst,” 2024, archive contents and research into working practice. View Source
40. Getty Foundation, Closer to Van Eyck, project overview. View Source
41. Alexandria Sivak, The Ghent Altarpiece in 100 Billion Pixels, Getty, 7 November 2017. View Source
42. Marie Postec and colleagues, Découvertes spectaculaires sur l’Agneau mystique des frères Van Eyck: le rôle essentiel du restaurateur, 2016, pp. 153–171, especially pp. 155–157 and 161–163. View Source
43. University of Antwerp, Phase I: The Ghent Altarpiece: finding Van Eyck, MA-XRF and cross-section research. View Source
44. Maximiliaan P. J. Martens, “Afblijven of Afnemen?”—Leave it or take it away: ethical considerations on the removal of overpaintings. The case of the Ghent Altarpiece, 2015, author version, pp. 1–7. View Source
45. KIK-IRPA, “Eyckian Lamb of God reveals secrets,” 2020, earlier body and retained contour. View Source
46. Zahra Sabetsarvestani and colleagues, “Artificial intelligence for art investigation: Meeting the challenge of separating X-ray images of the Ghent Altarpiece,” Science Advances, 2019, DOI 10.1126/sciadv.aaw7416: method, results and limitations. View Source
47. Aleksandra Pižurica and co-authors, Digital Image Processing of the Ghent Altarpiece: Supporting the painting’s study and conservation treatment, IEEE Signal Processing Magazine 32, no. 4, 2015, pp. 112–122, section IV. View Source
48. Sophie Kirkpatrick and colleagues, Multimodal Characterization of the Tin Relief Decorations on the Van Eyck’s Ghent Altarpiece and Their Degradation Process, June 2026 preprint, DOI 10.21203/rs.3.rs-9941906/v1, author-posted full text. View Source
49. Kathleen Froyen, “The conservation in three details,” 19 September 2025, ultramarine and Mary's robe. View Source
50. KIK-IRPA, upper-register conservation announcement, 10 December 2024: original paint, varnish removal and retouching. View Source
51. KIK-IRPA, “Additional funding for the third phase of the conservation of the Ghent Altarpiece”: revised treatment and return timetable. View Source
52. Maaike Blancke, “A new exhibition space for the Ghent Altarpiece,” colour matching and display lighting. View Source