by Tom Gurney
Tom Gurney BSc (Hons) is an art history expert with over 20 years experience
Published on June 19, 2020 / Updated on October 14, 2026
Email: tomgurney1@gmail.com / Phone: +44 7429 011000
A familiar face with an uncertain name
A woman looks out from a deep red setting. Her dark blue-green dress and pale undergarment make a broad shape below the softly modelled face. A string of pearls draws the eye towards her neck, while her hands hold a folded fan. Fur rests across one shoulder. The portrait presents luxurious textures without turning the sitter into a flat display of possessions: her face remains calm, direct and unusually close to us.
The Gemäldegalerie in Berlin[1] calls this painting Portrait of a Woman, inventory 762A. It credits Peter Paul Rubens, describes the support as oak panel, and gives the present painted image as 97.7 × 71.8 centimetres. The collection says the woman's features resemble Isabella Brant, Rubens's first wife. That careful wording is central to this page. The inherited route names Isabella, but the museum's current title does not establish that she is certainly the sitter.
The Berlin record[1] also reveals that the portrait has more than one making phase. A smaller first version belongs to about 1622–26. The panel was enlarged at the sides and bottom no earlier than about 1637. Its structured date of “around 1637” reflects the later state of the object; it cannot mean Isabella sat for Rubens then, since she died in 1626. This physical history is as important to an honest caption as the face itself.
At a Glance
- ArtistPeter Paul Rubens [1]
- Current titlePortrait of a Woman [1]
- SitterFeatures resembling Isabella Brant; identification uncertain [1]
- First versionAbout 1622–26 [1]
- EnlargementNo earlier than about 1637 [1]
- MediumOil on oak panel [1]
- Present image97.7 × 71.8 cm [1]
- CollectionGemäldegalerie, Berlin, 762A [1]
Contents
- What can we actually see?
- Why does Berlin use a neutral title?
- How did the panel change over time?
- Which Isabella portraits offer useful comparisons?
- What do date and ownership tell us?
What can we actually see?
The sitter wears a white underdress beneath the richly coloured velvet. Darker passages absorb parts of the shoulders and bodice, while lighter fabric at the sleeves catches the light in crumpled folds. Pearls and a pendant form small bright accents against the chest. The fan brings her hands into the lower centre of the composition. Those details are visible in the reproduction and make the portrait legible as a carefully arranged half-length likeness.
Berlin's catalogue identifies the velvet as blue-green, the ground as a strong wine red, and the fur as crossing the sitter's left shoulder. It contrasts the loose handling of the background with the smooth, careful work in the face. That difference rewards close looking. The red need not be read as a literal room with documented furnishings: here it functions as a field of colour against which the pale face and dark costume stand out.
The face is painted with a quieter rhythm than the clothes. A little shadow shapes the cheeks and the area around the eyes; the mouth is small and reserved. The head is held upright, but the torso and hands introduce gentler shifts of direction. The portrait gives the sitter an alert presence without a dramatic action. It is tempting to convert that expression into a story about her mood. A painted likeness, however, cannot tell us what she was thinking during a sitting or whether its expression was changed in a later phase.
The jewelry and fur are also easy to overinterpret. They signal a costly presentation, but the Berlin entry does not identify a particular ceremony, commission or emblem for them. The fan is an object to be described before it is assigned symbolic meaning. The safest reading starts with the painter's contrasts: light against dark, smooth face against brisker ground, a poised upper body against the movements of the hands and cloth.
Why does Berlin use a neutral title?
The museum says the woman has the features of Isabella Brant; it does not title the object Portrait of Isabella Brant. These are different levels of identification. A resemblance can invite comparison with securely named portraits, but it is not the same as a contemporary inscription, commission record or unbroken sitter tradition. The website's old page name remains a useful route to the object; the displayed title should follow the holding collection's more cautious formulation.
Isabella was a real and frequently represented member of Rubens's family. The Rubenshuis biography[3] records their marriage in 1609 and her death on 20 June 1626. Berlin adds that she was the daughter of Antwerp's city secretary Jan Brant and that her sister married Rubens's elder brother Philip. Those facts explain why Isabella is a plausible candidate for a portrait made in Rubens's household. They do not by themselves prove that every woman with similar features in his work is Isabella.
Comparison can sharpen the question. The British Museum[2] identifies a drawing of Isabella Brant, around 1620–21 in its structured date field, showing her head and shoulders with a smile. Its catalogue discusses parallels with the Berlin panel, while acknowledging that the Berlin sitter has not always been recognised as Isabella. The drawing supplies a documented likeness for comparison; its medium, pose and degree of finish differ from the oil painting. A reader should not collapse the two into a single artwork or call the Berlin painting a direct copy of the London sheet.
The resemblance question remains meaningful even if the sitter cannot be named securely. It shows how portrait identifications are tested against other images and how collections express uncertainty. A title that simply declares “Isabella Brant” would conceal the Berlin museum's reservation. A title that simply ignores the possible connection would conceal why this particular portrait has long drawn interest. Both the portrait and the question deserve to be visible.
How did the panel change over time?
The smaller first version is dated about 1622–26, within Isabella's lifetime. Berlin says the image was enlarged along the sides and lower edge no earlier than about 1637. That means the size we see today is not necessarily the size at which the first likeness was conceived. Added space can change how a figure sits within a picture, how much of the costume is shown, and how close the sitter seems to the edges. The museum does not spell out every brushstroke belonging to each phase in its public summary, so the page does not assign individual folds or accessories to a specific campaign.
An altered panel is still the historical artwork, not a damaged puzzle to be mentally trimmed back to an imagined original. The later enlargement is part of what the Berlin collection holds and measures. At the same time, a caption that gives only “1637” risks implying a sitting in that year. The most informative short description gives both the earlier portrait phase and the later expansion, with their approximate dates and different evidential limits.
The change raises questions the public record does not settle. We are not told in that short description who requested the enlargement, what use prompted it, or whether the sitter's identity was certain to the person who changed the panel. A posthumous enlargement could be compatible with an Isabella identification, but it is not proof of one. Nor does the later work make the face a wholly 1630s invention. Berlin explicitly places a smaller first version in the 1620s.
Physical sequence matters because a painting is more than the single image transmitted in a web photograph. The support is oak, the current dimensions belong to the enlarged state, and the museum's dating includes interpretation of that state. Thinking in phases lets the viewer ask a better question than “What year was this painted?”: when was the likeness first made, and when did the object acquire its present shape?
The small reproduction on this page cannot be used to map the joins of the oak support or to decide precisely where one campaign of paint ends. It shows the current composition as a whole. For the dimensions, material and sequence of enlargement, the Berlin collection record is the controlling evidence. This difference between what a photograph reveals and what an object catalogue documents is practical for readers: the fan and fur can be described from the image, but the phases of the panel need the museum's research. The image is an entry point to the painting, not a substitute for its physical examination.
Which Isabella portraits offer useful comparisons?
The British Museum drawing[2], numbered 1893,0731.21, is in black and red chalk with brown wash and white heightening. It records Isabella with swept-back hair and drop earrings. The museum says no painting directly corresponds to it, although there are parallels with paintings in Cleveland and Florence and with a Van Dyck portrait in Washington. That distinction prevents a neat but unsupported chain from “drawing” to “Berlin painting.” The London sheet is a useful visual comparator, not a documented preliminary design for this panel.
The Cleveland Museum of Art[4] holds another Rubens Portrait of Isabella Brant, about 1620–25, oil on wood, accession 1947.207. Its entry describes an intimate expression and notes that Rubens first painted Isabella's face and added the hands later. That is an alteration within a separate portrait. It cannot be transferred to the Berlin work as evidence of exactly how its own enlargement proceeded. The comparison shows that portraits of Isabella could have distinct formats and working histories.
An earlier, more public image is Rubens and Isabella Brant in the Honeysuckle Bower[5] in Munich, dated about 1609–10. The Alte Pinakothek says it was probably made for their marriage and describes its larger representative format. The couple sit together under honeysuckle, rather than the woman appearing alone against a red ground. That image can establish Isabella's place in Rubens's portrait practice, but its early date and double-portrait design make it a different kind of work.
The National Gallery of Art in Washington[6] holds Anthony van Dyck's Isabella Brant, 1621, accession 1937.1.47. The Gallery says Van Dyck probably drew on a likeness Rubens had made and records that its painting was formerly attributed to Rubens before the Van Dyck attribution became settled. The Washington picture is useful precisely because maker and sitter must be kept separate. It is not Rubens's Berlin panel, and a visual similarity does not override the different catalogued authorships.
What do date and ownership tell us?
Berlin's public entry places the first version before or around the end of Isabella's life and the expansion more than a decade later. The Rubenshuis account[3] identifies the cause of her death as plague in June 1626. That chronology rules out a living sitting for Isabella in 1637. It does not tell us whether the woman in the first version was Isabella, nor does it say why Rubens or someone working with the panel enlarged it. Those are separate historical questions.
The Gemäldegalerie records a much later collection event: purchase in 1903 from the collection of Empress Victoria Friedrich at Schloss Friedrichshof. That is a documented route into the Berlin museum, not a provenance back to Rubens's own studio. A gap between the seventeenth-century making and the twentieth-century acquisition should not be filled with invented owners. The current museum number 762A is the most useful anchor for readers checking this particular panel against lookalikes.
The painting is therefore legible through three kinds of evidence: its visible composition, the museum's account of its physical phases, and comparison with independently catalogued portraits of Isabella. Each supports a different claim. The visible face invites an identification, the chronology limits what a later date can mean, and the comparative works keep the proposal open without turning it into certainty. That approach gives the sitter room to remain a person in a particular portrait rather than a name attached to every similar image.
Rubens artworks directory
Frequently Asked Questions
No secure identification is asserted in the current Berlin Gemäldegalerie record. It calls the painting Portrait of a Woman and says the sitter's features resemble Isabella Brant's.
Berlin dates a smaller first version to about 1622–26 and says it was enlarged at the sides and bottom no earlier than about 1637. The two dates describe different phases.
No. Isabella died in June 1626. The sitter identification and the later expansion must be considered separately.
No. The Berlin work is an oil painting on oak panel, inventory 762A. The British Museum holds a separate chalk-and-wash drawing of Isabella, 1893,0731.21.
The Gemäldegalerie gives the present painted image as 97.7 × 71.8 centimetres. That is the enlarged state, not a measurement of the smaller first version.
Use the Staatliche Museen zu Berlin object page[1] for inventory 762A and the collection's current title and description.
Discussion
Look closely at the contrast between the face and wine-red ground.
Share a close observation or a reliable source about Berlin 762A.
Reader Insights
How should the museum's cautious sitter identification shape its title?
What changes when an initially smaller portrait is enlarged?
Which Isabella portraits make useful comparisons, and why?
Join the Conversation
Add your reading below, distinguishing visible details from object history.
References
- Staatliche Museen zu Berlin, Gemäldegalerie, Bildnis einer Frau , 762A. View source
- British Museum, Rubens, Portrait of Isabella Brant , 1893,0731.21. View source
- Rubenshuis, biography of Peter Paul Rubens. View source
- Cleveland Museum of Art, Rubens, Portrait of Isabella Brant , 1947.207. View source
- Alte Pinakothek, Rubens and Isabella Brant in the Honeysuckle Bower, 334. View source
- National Gallery of Art, Van Dyck, Isabella Brant , 1937.1.47. View source
