In this powerful and enigmatic painting, a man holding an ornamented Ottoman sword sits in the foreground of a dimly lit interior.1 Portrayed half-length and in three-quarter view, he engages us with steady blue eyes and a muted expression. His restrained smile and focused gaze reflect little about his state of mind, yet somehow render him familiar and approachable. He wears old-fashioned clothing consisting of a purple jerkin over a white blouse with ruffled cuffs gathered tightly at the wrists. Beneath a plush purple beret, the man’s long brown curls extend onto a black cape that envelops the side of the wooden chair. A gold chain and medallion hang prominently across his chest, with areas of impasto — as well as those on the silver and gold ornament on the sword — depicting the reflected light that descends from the painting’s upper-left corner.
Man with a Sword, which is signed and dated 1644, has a complex and fascinating history, much of which has only come to light in recent years. After having been out of the public eye for nearly two decades, the painting reappeared on the art market in 2013.2 Prior to its sale, it underwent a series of technical examinations in order to address questions regarding its attribution and dating.3 Although nineteenth- and early twentieth-century scholars viewed the painting as an exceptional work by Rembrandt, by the second half of the twentieth century most scholars no longer accepted its attribution.4 Horst Gerson was the first to raise doubt, remarking in 1968 that “to judge from the photograph, the attribution to Rembrandt is not very convincing.”5 In 1970, Bob Haak and Pieter van Thiel, then members of the Rembrandt Research Project, considered the work to be an eighteenth-century pastiche.6 Subsequent proposals in the 1980s and 1990s that the painting should be attributed to Govaert Flinck and Ferdinand Bol never gained acceptance.7
Technical research undertaken in 2012/13 demonstrated a number of reasons why the attribution of this work had proved to be such a conundrum for Rembrandt scholars. The discovery of a quartz ground, which is consistent with and specific to Rembrandt and his workshop’s practices starting with The Night Watch (ca. 1639–42), indicates that the painting must have originated in Rembrandt’s workshop.8 The signature and date, moreover, were also found to be integral to the underlying paint layers.9 X-radiography, infrared reflectography (IRR) (fig 1), and macro XRF scanning, however, revealed a number of compositional changes—particularly in the costume and position of the hands—that raise fascinating questions about the painting’s conception and execution.
X-radiography and infrared images reveal that the man originally wore contemporary dress consisting of a white shirt with a high, horizontal neckline and collar, and unadorned cuffs at the wrists.10 The beret was added in a second stage; a large plume, which was partly articulated with a series of rapid scratches into the wet paint, was also added. A medallion, which probably held the feather in place, was positioned near the brim of the beret (fig 2). Both the feather and the medallion were thereafter overpainted.11 The sitter’s hands were also substantially reworked. While the sequence of brushwork on the sitter’s left hand is not discernable in the X-radiograph, the changes to the right hand are evident (fig 3). It appears originally to have been placed higher and further to the left, with the thumb raised, to accommodate the sitter’s grip on either another object or a part of his dress.12 In the background, a red curtain that hung in the upper right corner was also overpainted. Strikingly, X-radiographs indicate that during the second campaign, only minor adjustments were made to the sitter’s face. It became somewhat rounder, the right nostril and nose were enlarged, and the bottom lip was slightly extended. The man’s bangs replaced a center-parted hairline he had in the initial stage of the painting (fig 4).
These technical examinations have raised the fundamental question of whether Rembrandt originally conceived the painting as a formal portrait before it was changed into a generic depiction of man with a sword. Ernst van de Wetering, who attributed the original concept for the painting and the underlying portrait to Rembrandt, reached this conclusion, but also proposed that a member of the artist’s workshop overpainted and modified significant portions of the composition.13 He suggested that Rembrandt may have left the so-called original portrait unfinished, or that the patron rejected it, a circumstance known to have occurred with other paintings by Rembrandt.14
Differences in handling between the modeling of the face and the figure’s costume and sword reinforce Van de Wetering’s assessment of this work. He concluded that the head and hair (“largely unaltered”), the shaded parts of the left hand, parts of the black cape (“raw, bravura brushwork”), and the signature and date were by Rembrandt, while a member of the workshop was responsible for the second stage of the painting’s execution.15 Indeed, the smoothly blended and well-modeled brushwork on the face, with areas of warm, pink flesh tones, is entirely characteristic of Rembrandt, while the modeling of the purple jerkin, collar, and chain is rather superficial in character. Finally, impastos on the medallion and on the ornamental features of the sword do not convincingly convey, as Rembrandt would have done, the three-dimensionality of these pictorial elements.
Questions of workshop participation related to Man with a Sword are analogous to those related to a few other Rembrandt paintings from this period. For example, Man with a Hawk (fig 5), signed and dated 1643, and its pendant, Woman with a Fan (fig 6), were recently reattributed to Rembrandt and workshop. Much as with Man with a Sword, changes made to Man with a Hawk during a second campaign were concentrated in the man’s costume and the position of his arm.16 As with the Leiden Collection painting, the figure’s face, executed with smooth, thin brushwork and “fluent transitions and a slight sfumato” in a manner consistent with Rembrandt’s style, was left largely intact and not overpainted.17 The execution of Woman with a Fan reflects some of these same stylistic characteristics. As Van de Wetering has suggested, the woman’s face and necklace “could have been painted by Rembrandt,” whereas her costume was likely executed by a member of the workshop. The artist responsible for the woman’s costume may have been the same one responsible for the revisions in Man with a Hawk.18
Could the artist who was involved in the production of these pendants also be the same artist responsible for the alterations in Man with a Sword? The similar patterns of revisions in Man with a Hawk and Man with a Sword are striking.19 However, an important difference in the genesis of these works merits mention: neither of the portrait pendants was transformed from a formal portrait featuring contemporary dress into a tronie-like image with fanciful costume, as was done with Man with a Sword. Rather, the Man with a Hawk and the Woman with a Fan were intended from the beginning as portraits historiés (sitters portrayed in the guise of historicizing costume).
The imaginative alteration of Man with a Sword’s attire links this work to the pictorial tradition of historicizing tronies that Rembrandt executed over the course of his career. Combinations of a beret or bonnet, jerkin, cape, and gold chain, which are derived from sixteenth-century fashions, already appeared in Rembrandt’s work in the 1620s.20 He used these motifs again in his self-portraits in the early 1640s, such as the Self-Portrait at the National Gallery in London from 1640 (fig 7), and the Self-Portrait at Windsor Castle from 1642 (fig 8). Although Man with a Sword does not portray Rembrandt himself—despite the assertions of many nineteenth- and early twentieth-century scholars—its strong similarities with the exotic character of his tronies provide compelling visual evidence for his close involvement in stages of its production.21
Understanding the period of Rembrandt’s career in which Man with a Sword was executed presents numerous challenges, largely because we know relatively little about how his workshop functioned during the mid-1640s. After he completed The Night Watch in 1642, Rembrandt’s artistic production slowed; he received fewer commissions for portraits, and almost turned away from self-portraiture until the 1650s.22 Did the workshop at this moment function similarly to the model of the late 1630s, when it has been understood that “Rembrandt worked together with pupils and assistants on the same compositions” and developed “an integrated approach to workshop production”?23 Were collaborations of this nature, which may have taken various forms, or the overpainting of Rembrandt’s works by his pupils commonplace?24
Technical examinations and careful stylistic assessments of Man with a Sword indicate that this painting’s distinctive character was the result of a complex creative process involving Rembrandt and his workshop. While Rembrandt initiated, signed, and dated the work, it appears that Rembrandt engaged a student or assistant to amend his initial concept.25 The hand of the other artist involved in that process has not been identified, but the painting’s quality and visual power indicate that Rembrandt must have guided its transformation while it was in his studio.26 Although the reasons for this reworking are unknown, Man with a Sword provides insights into the character of Rembrandt’s workshop in the mid-1640s. Its transformation and reappearance after years of obscurity also offer a fascinating demonstration of how insights into a work of art change over time, depending on the availability of the work itself, as well as on the evolution of research techniques.