The Signing of Paintings by Rembrandt and His Contemporaries

Relatively little attention has been paid to the intriguing practice of signing paintings in the seventeenth-century Dutch Republic. Whereas in the past a painting was accepted as a work by a certain artist because it bore a signature, today this no longer seems so certain for works from this period. Contemporary sources that could provide some clarity in this matter are scarce. Moreover, differences are evident among seventeenth-century painters not only in the frequency and context of signing works—whether their own or those created with or by their pupils—but also in the application of various forms of signatures. Of the twenty paintings in The Leiden Collection made by Rembrandt van Rijn (1606–69) or produced in his workshop, thirteen bear a Rembrandt signature, either with his monogram or his first name in cursive script.¹ The signatures span nearly his entire career, beginning with Unconscious Patient (Allegory of Smell), which was part of a series of the Five Senses originating around 1624–25, and extending to Portrait of a Seated Woman with Her Hands Clasped, dated 1660. The proportion of signed paintings in the Leiden Collection roughly corresponds to the percentage of signed works in Rembrandt’s entire painted oeuvre: of the 349 paintings Ernst van de Wetering recorded in 2017 as works by Rembrandt, whether or not produced in collaboration with assistants, 257 bear a signature.²

In the seventeenth century, signing practices among artists varied. For instance, the high percentage of signed works by Rembrandt highlights a significant difference between his oeuvre and that of Flemish artist Peter Paul Rubens (1577–1640) in Antwerp. Rubens did so very rarely: we know of only five signed paintings in his extensive oeuvre.³ This distinction reflects the divergent practices regarding signing works of art in the Northern and the Southern Netherlands more broadly.⁴ However, dissimilarities in these practices can be observed within the Northern Netherlands as well. They concern not only the frequency of signing but also the way in which it was done. The following essay first considers the name by which a person was known in the seventeenth century. It then discusses Rembrandt’s various signatures on his paintings, comparing them to those of other artists, particularly the signatures of his teachers and other painters from his immediate circle. Finally, it examines the practice of signing, drawing attention to the scarce seventeenth-century documents in which signatures are mentioned and considering the question of when Rembrandt signed his work.

“Rembrantio nomen est”— “Rembrandt is the Name”

A person’s identity is determined by various factors, for example their place of origin or profession. A name also contributes greatly to an individual’s identity. Until the beginning of the seventeenth century, many people in the Dutch Republic used their baptismal name in combination with a patronymic, formed from the father’s forename followed by “son” or “daughter.” For men, once they became employed, it was customary to add their professional designation. In the 1600s, the use of surnames steadily became more popular, as individuals adopted the name of their place or region of origin or that of their occupation. Some surnames, particularly those related to place of origin, were added by later commentators to help identify particular individuals. For example, in 1604, Karel van Mander (1548–1606) used the name Lucas van Leyden (1494–1533) in his Schilder-boeck, while it can be deduced from documents that the artist’s contemporaries knew him as Lucas Hu(y)gensz.⁵ Rembrandt had already been provided for at birth, for his father was called Van R(h)ijn, and—while this was not an obvious practice at the time—the artist also went by this surname. A receipt he drew up in 1630 for the apprenticeship fees he received from Isaac de Jouderville (ca. 1612–45/48) begins thus: “ick, Rembrant Harmensz. van Rijn” (I, Rembrandt Harmensz. van Rijn).⁶

For a long time, however, people in the Dutch Republic considered a person’s given name as their actual name. For example, guild members were often addressed as “master” followed by their first name—even if they had a surname. In Leiden, Coenraet Adriaensz van Schilperoort (1577–1636), a contemporary of Rembrandt, was known as “Mr. Coenraet.” Landscapes by him are credited as such in Leiden estate inventories.⁷ Similarly, in 1645, a wine merchant in Leiden gave his father-in-law in Amsterdam his household effects, including a portrait of his wife “gedaen bij mr. Ysack” (“done by Master Isaac”).⁸ The portraitist must have been the Isaac de Jouderville mentioned above.

In the case of artists with a very common forename, using it would not have been practical. But a name like Coenraet was not easily confused with another artist, nor, for that matter, was Isaac. The same certainly applied to Rembrandt, a name rarely found outside of Holland and Zeeland in the seventeenth century and, moreover, one that occurred very rarely even in these coastal regions.⁹ In sources from the early part of his career, Rembrandt’s work is already referred to by his first name. In 1628, Joan Huydecoper (1599–1661) recorded the purchase of a small painting by “Rembrant” in his cash book, after initially writing the name incorrectly as “Warmbrant.”¹⁰ Although this may have simply been an error, one wonders if Huydecoper, from Amsterdam, was not yet familiar with the name of Rembrandt of Leiden. As we shall see below, if this work was signed, it must have borne Rembrandt’s monogram and not his first or last name. The fact that he was already known by his first name before signing his work with it is also apparent from the autobiography that the Hague resident Constantijn Huygens (1596–1687) wrote in Latin between 1629 and 1631. He mentions Rembrandt only by his first name (“Rembrantio nomen est”) in the same breath as Jan Lievens (1607–1674), whose patronymic Huygens added to the artist’s common first name.¹¹

Rembrandt’s Signature

When Rembrandt signed documents, he did so with his first and last name, while in official deeds he also added his patronymic (“Rembran[d]t Harmensz van Rijn”). As noted above, Rembrandt’s receipts for the guardians of Isaac de Jouderville, drawn up between the beginning of May 1630 and mid-November 1631, were signed in the latter manner. In these documents, he spelled his forename “Rembrant,” ending in -nt. By contrast, he initially signed works of art with his initials only. In his early years in Leiden, he used several types of monograms. His monogram on his earliest signed painting, dated 1625, is a capital R. He subsequently added the letter H, sometimes with a lower-case or capital f (for fecit, Latin for “made”) and usually followed by a date. Rembrandt kept this monogram until 1627. In 1628, the year in which he first provided his monogrammed and dated prints, he included an L (Leydensis, “from Leiden”)¹² after his two initials. Rembrandt only used the L for a short while, abandoning it in the course of 1632, one year after he began working for Hendrick Uylenburgh (ca. 1584/9–1661) in Amsterdam. His relocation to that city was probably gradual, for it seems that he continued to work in his Leiden studio through the first half of 1632.

During Rembrandt’s transition from Leiden to Amsterdam, he began to paint portraits, a genre not known from his Leiden years. He signed his first two likenesses “RHL 1631,” a monogram he continued to use in early 1632, as, for instance, in the Portrait of Marten Looten (fig 1). This sitter holds a note on which can be read “xj. Januyary” and, farther down, the monogram and date “RHL 1632.” During this year, the artist began to add “van Rijn” to his monogram, as seen in The Leiden Collection’s Young Girl in a Gold-Trimmed Cloak (fig 2). In 1632 and 1633, Rembrandt began signing his paintings with his first name only, written in cursive script, initially ending with -nt and then with -ndt. Such signatures can be seen in three paintings from 1633 in The Leiden Collection: Portrait of a Young Woman (fig 3), signed with only -nt; and Portrait of a Man in a Red Coat (fig 4) and Bust of a Bearded Old Man (fig 5), both of which are signed with -ndt. Rembrandt continued signing paintings with only his first name until his death in 1669.¹³ While his monograms include his initials in various characters, his first-name signatures generally exhibit the same, somewhat idiosyncratic cursive script.¹⁴

Both monograms and names written out in full can be found in signed works of art by Rembrandt’s teachers, as well as by his Leiden companion Jan Lievens. Only a few paintings by his first teacher, Jacob Isaacsz van Swanenburgh (1571–1638), are known, three of which are signed—one with the monogram “IVS” and two with his full name. Based on this modest number, little can be said about his signatures. This is not the case, however, with Rembrandt’s second teacher, Pieter Lastman (1583–1633), who frequently signed his paintings with his monogram, “PL,” or his signature, either “Pietro Lastman” or “PLastman,” as found in David Gives Uriah a Letter for Joab in The Leiden Collection (fig 6). Dated works by Lastman indicate that he alternated between monograming and signing. In his younger years, Jan Lievens, who also studied under Lastman, signed his paintings considerably less often than Rembrandt did. While Rembrandt signed about 70 percent of his paintings in Leiden, Lievens signed less than 30 percent.¹⁵ The majority of Lievens’s signatures consist of the letter L, as in The Leiden Collection’s Bookkeeper at His Desk (fig 7), or the monogram “IL” seen in Card Players and in Self-Portrait (fig 8), also in The Leiden Collection. Occasionally, he signed his paintings “J. Lievens,” or the Latinized version of it, “J. Livius.” While all four artists thus inscribed their works with monograms as well as signatures, Lastman and Lievens used both forms simultaneously, while Rembrandt shifted over time from monogram to signature.

As noted, the percentage of Rembrandt’s signed paintings is high: approximately three quarters of his currently known oeuvre bears his name. This percentage would undoubtedly have been higher originally, given that signatures may have disappeared when canvases and panels were cut down at a later time. In addition, pictures that are either part of a series or considered pendants are generally not all signed individually. Not infrequently, only one work of a series or in a pair of pendants is signed, whereby the series as a whole can be considered as signed. This is the case with Rembrandt’s series of the Five Senses.¹⁶ Of the four paintings from this series now known, only Unconscious Patient (Allegory of Smell) is signed: the monogram “RHF” appears on the drawing on the back wall, a tronie of an old man (fig 9). Of the six surviving paintings of the Passion Series that Rembrandt painted between 1632 and 1646 for the stadholder’s court in The Hague, only The Ascension of Christ (1636), The Resurrection (1639), and The Adoration of the Shepherds (1646) are signed.¹⁷ Similarly, of the three busts painted on copper plates covered in gold leaf—Old Woman Praying (ca. 1629–30),¹⁸ Laughing Man (ca. 1629–30),¹⁹ and Self-Portrait (1630)²⁰—only the latter work bears a signature. The identical format and corresponding, highly unusual support of these three paintings indicate that they were conceived as a series.²¹

In 1632, Rembrandt began signing his work with his first name in cursive script. Famous Italian Renaissance painters such as Leonardo, Raphael, and Titian, who were known by their first names, are often cited as sources of inspiration for Rembrandt’s decision to sign in this manner. However, the practice was not unique to these Italian painters. As already indicated, in the Dutch Republic, the given name had long been considered to be an individual’s actual name; nevertheless, signing a painting with a given name was very unusual in Rembrandt’s time. One other Dutch painter with an uncommon first name who also did so is Hercules Segers (1589/90–1633/40). Of his five known signed paintings, one is signed “herkeles segers” and three “hercules segers,” but one bears only the artist’s first name, “Hercules.”²²

The Seventeenth-Century Practice of Signing

The question of why artists signed their work in the sixteenth and seventeenth centuries is not easy to answer.²³ Painters had to comply with the rules of their local guilds; however, guild regulations do not contain any prescriptions about signing paintings. For example, the ordinance of the Delft Guild of Saint Luke of 1611, with addenda recorded in later years, contains no provisions whatsoever regarding the signing of works of art. Similarly, the 1590 charter of the Haarlem artists’ guild does not mention signing, nor does the extensive, unofficial draft of 1631, which was not formally sanctioned by the magistrate of this city. The same situation is found in Amsterdam, where the guild charter of 1579 includes nothing about signing, nor do the additions and amendments to it in later years.

The fifteenth article of the Hague painters’ confraternity, issued by the city on 21 October 1656, states that a disciple or pupil who has progressed to the point of signing his work was obliged to pay a contribution to the confraternity. This meant that he was then considered a master. The fact that an artist signed his name was apparently taken for granted, which is not to say that everyone actually did so. This regulation offers no insight into the choice of individual artists to sign or not to sign, and there are virtually no other documents dealing with this subject. An article in an addendum made to the Utrecht ordinance of the Guild of Saint Luke issued on 15 March 1651 (first issued in 1644) also mentions something about signing without giving insight into the choice of when or how to sign by individual masters, although this passage has often been misinterpreted.²⁴

Cultural and cognitive changes in the early modern period, in which the notion of individuality gradually came to the fore, may have encouraged artists to distinguish themselves emphatically from their colleagues by means of their signature. Sales may also have played a significant role in the matter of signing. The landscape painter Jan van Goyen (1596–1656) frequently signed not only his paintings but also many of his drawings with a clearly legible name. However, he generally did not mark the sketches he kept for his own use.

Rudi Ekkart has argued that there were local differences as to whether portraits produced in the Dutch Republic were signed or not. The late sixteenth-century Delft painters Jacob Willemsz Delff (ca. 1545/50–1601) and Herman van der Mast (ca. 1550–1610), for example, usually signed their portraits, as did their younger fellow townsman Michiel van Mierevelt (1567–1641) from 1607 onward. In contrast, likenesses made in Amsterdam before 1620 were seldom signed.²⁵ In subsequent years, Ekkart suggests, a noted increase in the number of signatures applied to portraits made in that city was related to artists’ growing sense of self-awareness.²⁶

The decision to sign a work need not have been solely the reserve of the maker, however; it might also reflect the will of the patron. Especially in the case of portraits, which were usually commissioned, the influence of the sitter or commissioner was crucial. It applied to the determination of the format, the pose of the sitter, and how the portrait was painted, as well as the application of an inscription, such as the subject’s age or the painting’s date, and the placement of a signature. For example, when they engaged Michiel van Mierevelt to paint four portraits of members of the House of Orange, the burgomasters of Delft specified that the artist should sign the portraits. Van Mierevelt accepted the commission and promised “syn eygen naem te sullen teykenen” (to sign [the pieces] with his own name).²⁷ The patrons’ request undoubtedly had to do with the destination of the likenesses: they were to hang in the formal burgomasters’ chamber in the town hall, where distinguished guests could view the portraits produced—and signed—by the famous Delft artist.

In addition to patrons or buyers of paintings, art dealers also may have influenced the practice of signing works of art. From the beginning of the seventeenth century onward, more and more paintings were sold by dealers rather than by their makers. The professional art dealer was a new phenomenon in the early years of the Dutch Republic; previously, painters, merchants, and agents had been engaged in the art trade. In 1621, the schilderijverkopers (painting sellers) in Amsterdam joined the Guild of Saint Luke. Without this membership and the citizenship that was a prerequisite for it, they could not trade in paintings. In this city of professional art dealers, early documented examples of this occupational designation include Guilliam Bouwens (ca. 1561–after 1627), who is called “schilderijcooper” (painting buyer) in 1625;²⁸ and Willem Sybrantsz. van der Bent (d. 1652), referred to as “schilderijcooper” in 1642 and, after his death in 1656, as “coopman van schilderijen” (painting merchant).²⁹ Hendrick Uylenburgh, who played an important role in Rembrandt’s career, first worked in Amsterdam as “coopman” (merchant; 1627–29) before being listed as “cunsthandelaer” (art dealer; 1631, 1634), “schilderhandelaer” (painting dealer; 1636), and “schildervercoper” (painting seller; 1637).³⁰ Bouwens came from Antwerp, where the professional art trade began significantly earlier than in Amsterdam. Painters, though, continued to trade in art as well, their winckels (shops) designated both workshop and salesroom.

It was in the dealers’ interest that the paintings they offered bore a signature. No matter how good, an anonymous work of art generally fetched less than one by a known painter. Interesting in this respect are the documents relating to two flower pieces shown to the board members of the Antwerp guild for appraisal in 1684, both signed by Nicolaes van Verendael (1640–91). The guild’s verdict was that the works were originals and could be sold as such. Four months later, Van Verendael himself appeared in the guild hall. He had, as he testified, painted over the pieces in question but did not consider them to be his own work; rather, he signed them because he had had a hand in them.³¹ The artist made a statement about this situation before a notary, and his desposition provides additional information. It reveals that a frame maker involved in the art trade had obtained the two flower pieces at an early stage and showed them to Van Verendael. The frame maker asked him to paint over them for a considerable fee conditioned on Van Verendael “sijnen naeme op deselve twee stucxkens soude moeten stellen” (signing these two pieces). Initially, the painter refused to do this. Only when the frame maker promised he would not trade the pictures in Antwerp did Van Verendael comply. A related statement shows that the flower pieces were by the artist’s former pupil.³²

This anecdote raises the question of what value a contemporary attributed to the signature on a painting in the Dutch Republic. Did it give him a guarantee of autography? It does seem that owners of paintings placed their trust in signatures. Hence, it was with good reason that the art dealer and frame maker urged Van Verendael to sign his work. This may have also been the case with Meleager and Atalanta by Abraham Janssen (1576–1632), now in Le Havre,³³ on which the pristine monogram “AB. I.” and, beneath it, the year “1625” came to light during restoration. In 1627, Hendrick Uylenburgh owned a painting by Janssen of this subject. Since we know that this dealer had previously purchased art in Antwerp and that Janssen, who worked there, rarely if ever signed his paintings, it is not implausible that this work was signed by its maker at the buyer’s behest.³⁴ The signature seems to have had a different meaning for makers than for buyers. When artists added a signature to paintings by pupils or advanced assistants that they had painted over or retouched, this did not automatically mean that they considered them to be autograph works; rather, the signature authenticated such paintings as studio pieces. Similarly, if a painter collaborated with another artist, this was not necessarily reflected in the signature.

An interesting source providing insight into this practice is the notebook of Adriaen van der Werff (1659–1722). This painter from Rotterdam kept meticulous records of the paintings he produced between 1716 and 1722. He noted the subject of each painting, the name of the person for whom he made the work, and how much he asked for it. He determined the asking price on the basis of the number of dashes he recorded, whereby a dash represented a day’s work. He kept a record not only of the time he had spent on a painting, but also of that of his brother Pieter van der Werff (1665–1722), who contributed to almost all the paintings recorded.³⁵ Several of the works mentioned in the notebook can be identified with extant paintings. It is clear that Pieter’s share in the production was not expressed in the signature. Even in the exceptional case that Pieter spent considerably more time on a painting than his brother, it only bore Adriaen’s name. For example, the “Shepherd with the two dancing women” recorded in the notebook, on which Adriaen worked for nine weeks and Pieter for no less than thirteen, is signed only “Chevr vr/werff. fec. 1718.”³⁶

The Practice of Signing in Rembrandt’s Workshop

Whereas Pieter van der Werff was an independent artist hired by his brother Adriaen, Rembrandt’s situation was different: he was the head of a workshop in Amsterdam with pupils and assistants. It has often been noted that Rembrandt predominantly attracted advanced pupils, often called “disciples,” who had already completed an apprenticeship elsewhere and who were employed by him in his production. According to the German painter and biographer Joachim von Sandrart (1606–88), who spent several years in Amsterdam from 1637 onward, the sale of paintings and prints by Rembrandt’s pupils contributed significantly to his income.³⁷

Data on the sale of Rembrandt’s works is extremely scarce, but a note by Rembrandt himself provides some insight. He recorded the sale of certain studio works on the back of a drawing, which can be dated around 1636. Rembrandt documented the names “Fardynandus” and “Leendert,” as well as the subjects of three of the paintings that were sold, namely a “Flora,” a “Vaandeldrager” (standard-bearer), and an “Abraham.” Rembrandt also noted several prices, including five guilders for which he had sold “Leenderts floorae.”³⁸ This and other recorded prices do not indicate drawn copies, as is sometimes assumed, but rather paintings: a drawing would not have sold for more than one guilder in those days. Because paintings by Rembrandt of the above subjects from 1634, 1635, and 1636 are known, the pieces mentioned in the note are most likely painted copies or variants made by pupils after examples by the master. Whether these works were signed is not known.

The fact that Rembrandt included on the back of this drawing the names of his pupils who made these paintings could indicate that he sold them not under his own name but as works made by advanced pupils under his supervision. “Fardynandus” refers to Ferdinand Bol (1616–80) and “Leendert” to Leendert van Beijeren (1619–48), both of whom are documented as Rembrandt’s disciples shortly after the mid-1630s. The 1638 inventory of Leendert’s father’s estate mentions paintings of “Flora” and  “Abrahams Offerhande” (Abrahams’s Sacrifice)—the latter described as “groot” (large)—and three more pieces “naer Rembrandt gecopieert” (copied after Rembrandt). The relatively low price paid for such studio work must mean the buyers knew that these pieces—signed or not—were absolutely not by the master himself.

Whether the paintings by Rembrandt’s pupils that he sold, as his note on the back of the drawing indicates, were signed and, if so, how cannot be determined with certainty. A Rembrandt signature can be found on assorted painted copies and variants based on  prototypes by the artist and, for various reasons, considered to be apprentice works made in his workshop.³⁹ This is the case with the Portrait of Rembrandt in Oriental Dress in The Leiden Collection attributed to Isaac de Jouderville (fig 10). However, the signature “Rembrandt ft. 1641” on that work was added later, probably after 1818. An auction catalogue of that year, where the painting was recorded for the first time with a detailed description, does not mention a signature.⁴⁰ This painting was copied in Rembrandt’s workshop after his Self-Portrait in Oriental Dress with a Spanish Water Dog, which is signed “Rembrant. f 1631.”⁴¹ The copy was made before the painter introduced the dog in the prototype, between 1631 and 1633, and therefore not in 1641, as the signature on the copy would have us believe.

If a signature forms an inextricable part of the painted image, then the act of signing is directly connected to the genesis of the painting. This is evident when the signature is placed in the wet paint, as in the undated Bust of a Young Bearded Man from ca. 1656–58 in The Leiden Collection (fig 11). Yet even if it can be established that a signature is part of the genesis of a painting, this does not clarify whether Rembrandt signed solely his own work or also pictures wholly or partly made by his workshop assistants. Looking for an answer to this question, it is interesting to consider the exploratory research of the handwriting experts of the Dutch Forensic Laboratory.⁴² They subjected the signatures on 88 paintings from the period 1632–42, 15 of which are signed “RHL van Rijn” and 73 “Rembrant” or “Rembrandt,” to a comparative handwriting analysis. The first group was too small to draw any conclusions. Of the second group, they considered the signature to be autograph in only 23 cases. As the examination was based on photographs, they took no account of the actual size of the signature or its material condition. The conclusions of the forensic research were compared to the Rembrandt Research Project’s (RRP) views on the attributions of the paintings in question. It emerged that three paintings with a signature recognized as autograph by the handwriting experts were, according to the RRP, not by Rembrandt, but probably made in his workshop—which goes to support the idea that Rembrandt did sign pictures executed by or with help from his assistants.⁴³

In their publication, the forensic handwriting analysts point out that art historians with little expert knowledge of handwriting should be cautious in assessing signatures on paintings. Nevertheless, for example, we deem the signature on one painting attributed to Rembrandt’s workshop (possiby Ferdinand Bol), the Man in Oriental Costume in The Leiden Collection (fig 12), a copy after a prototype by Rembrandt in Chatsworth House (ca. 1639), as not autograph.⁴⁴ The very regularly formed and spaced letters deviate too much from the handwriting that is qualified as autograph by the handwriting experts. With the information available to us, it is not possible to determine by whom and when the painting was signed.

The examination of the signatures by the forensic handwriting experts proved innovative, but it left many questions unresolved. They chose to examine the signatures according to their own methodology, analyzing the signatures as if they were written with pencil or pen, rather than painted, and without any knowledge of the paintings that could influence their results. And while it would be interesting to continue this kind of investigation, some conclusions must be drawn from the exploratory research about the way in which a signature on a painting should be examined. First, when assessing signatures, differences between those painted with a brush and those written with a pen or pencil should be taken into account. Moreover, the handwriting experts’ assessment was based on photographs of the signatures and disregarded all kinds of relevant information, such as the size of the signature, its place in the painting, the manner of application—for example in the wet paint or only after the work was completed—and the condition of the painting, often worn or worked up. A forensic handwriting investigation should be integrated with art historical research and combined with the expertise of other specialists, including painting conservators.

Only then would it be useful to reexamine the signature on Rembrandt’s The Anatomy Lesson of Dr Nicolaes Tulp (1632),⁴⁵ for example, which the handwriting experts put forward as an example of an inauthentic signature because the letters deviate from those of signatures considered trustworthy. The signature on this painting, however, is integrated into the scene: it is placed on a roll of paper hanging on the back wall, which may help to explain the different writing. Something similar happens to the signature on other Rembrandt paintings such as Susanna and the Elders (fig 13) in Berlin, in which the signature is placed at an angle so that it follows the perspective of the painting. This phenomenon also occurs in early paintings by Rembrandt with monograms integrated in the scene, as in the case of the Unconscious Patient (Allegory of Smell) (fig 9) and the Portrait of Marten Looten (fig 1).

At What Point Did an Artist Sign?

A fascinating aspect of Rembrandt’s art is his creative process, especially the changes he made in the course of executing his works. The complex genesis of Susanna and the Elders (fig 13), for example, can be clearly discerned, informed by drawings, written sources, and technical research. This documentation also provides us with an exceptionally good sense of the time span of the painting’s entire creation process. Rembrandt began this ambitious work in the 1630s and changed it drastically over the years. It must have been completed in some form in 1642, when his wife, Saskia, died, because it is mentioned in a document relating to her estate. But Rembrandt then continued to work on the painting until as late as 1647, when he sold it to Adriaen Banck for 500 guilders, as indicated in a statement by the latter. The year 1647 is also inscribed with the painting’s signature, which may therefore indicate both the date of its completion and the time it was sold and left the workshop.⁴⁶

In the case of the Portrait of Petronella Buys in The Leiden Collection (fig 14), too, it is highly likely that the date 1635 accompanying the signature refers to the moment when it left Rembrandt’s workshop. An inscription on the back of the painting, applied later in the seventeenth century, reads: “Jonckvr. Petronella Buijs: sijne Huijsvr. Naer dato getrout aen de Hr. Borgerm. Cardon” (Jonkvrouw Petronella Buijs, . his [Philips Lucasz] wife/ after this married to Burgomaster Cardon). The term Jonkvrouw refers to her unmarried status. She must, therefore, have sat for Rembrandt before 27 August 1634, the day she wed Philips Lucasz in the Nieuwe Kerk in Amsterdam. The pendant depicting her husband would have also been created in the same period.⁴⁷ The year would have been applied when both portraits were completed or delivered, about the exact date of which there is no certainty. Although the couple left for the East Indies on 2 May 1635, the portraits were not intended for the sitters themselves, but rather Petronella’s sister and brother-in-law, in whose possession they are mentioned in 1653 and 1655.⁴⁸

The year inscribed with the signature usually marks the end of the creation process, but this does not always seem to be the case. Rembrandt signed his Self-Portrait in Oriental Dress with a Spanish Water Dog “Rembrant.f . . . 1631” upon completion after having added the dog. However, the spelling “Rembrant” does not appear on other works from the year 1631. This spelling of the name does occur a little later, namely in 1632 and 1633. Due to the Self-Portrait signature’s close resemblance to that on a painting signed and dated 1633, the Rembrandt Research Project concluded that Rembrandt did not sign the Self-Portrait in 1631; rather, they propose he added the signature in 1633 after the addition of the dog.⁴⁹ Rembrandt would thus have antedated the Self-Portrait to 1631, in all probability the year it was finished in the form copied in Rembrandt’s workshop as the Portrait of Rembrandt in Oriental Dress (fig 10) in The Leiden Collection.⁵⁰

As Rembrandt’s Self-Portrait attests, the presence of the artist’s signature, whether or not accompanied by a date, is just one thread in a complex web of clues about the circumstances of a seventeenth-century painting’s production. The twenty paintings in The Leiden Collection that can be attributed to Rembrandt and his workshop not only reflect the high percentage of signed works in Rembrandt’s oeuvre but also afford great insight into the various signature forms he used throughout his career. More broadly, the signatures of Rembrandt and his contemporaries reflect a wide range of customs and forms of signing works of art. The study of signatures can provide insight not only into the practices of individual artists and their workshop production but also into the requirements of their clients and art dealers.