The Card Players

Two men face one another across a red table. The one on the left holds a fan of pale cards and a pipe; his opponent studies a smaller hand. A dark bottle stands between them. The picture looks familiar as The Card Players , but it is only one of five oil paintings Cézanne made of this subject. The lead image shows the Musée d’Orsay’s two-player canvas, RF 1969 (FWN 684). The Courtauld Gallery in London owns another pair; a much larger pair is in a private collection in Qatar. The Metropolitan Museum of Art owns a scene with three seated players and a standing watcher. At the Barnes Foundation, a child joins that four-person company. The shared title conceals paintings of strikingly different scale, arrangement and mood. ^{[1] [2] [3] [4] [5]}

The series was made in the 1890s, but its internal order is still debated. The range of five catalogued canvases gives us no single scene to reconstruct, no documented winner and no secure transcript of a game. What we can follow is Cézanne’s sustained exploration of people brought together by a little table: how a body occupies a room, how one person’s attention meets another’s, how small cards can command a large painting. Individual studies of smokers and players complicate the story further. Some may precede a group composition; at least one related portrait has been argued to follow the Met canvas. The five paintings are therefore more revealing as variations tested against one another than as steps on an assumed path to a final image. ^{[6] [7]}

The Five Canvases at Different Scales

The two crowded rooms

The Met painting measures 65.4 × 81.9 centimetres. Three men sit around a shallow white table: the central player wears a large dark rounded hat, while a fourth man, also hatted and smoking a pipe, stands behind the man at left. A rack of four pipes hangs on the pale blue wall, and a gold curtain presses down the right edge. The player on the right seems almost too large for the room. His blue back, sleeve and dark hat occupy a formidable area, while the central player’s pale clothing opens a narrow interval between the darker figures. On the table, the cards are surprisingly small. The scene gives far more space to the men’s bent bodies and to the wall behind them than to the game’s visible equipment. ^{[2] [8]}

At the Barnes, the composition expands to 135.3 × 181.9 centimetres, and the difference reaches beyond physical enlargement. A child appears between the bareheaded central player and the man in blue. The standing smoker has a more isolated position at the back. High on the wall are a little shelf with a vessel, a framed dark rectangle and a pipe rack; a broad ochre curtain fills the upper right. The white table remains the apparent anchor, but its playing surface is darker and more patterned than in the Met version. It carries a pipe as well as cards. The child’s face makes the game a scene watched from within the adult circle, while the standing smoker watches from farther away. These altered relationships are why the Barnes canvas deserves a separate close reading , rather than a caption describing it as the Met painting made larger. ^{[3] [9]}

The Barnes Foundation observes that the size approaches a format conventionally associated with history or mythological painting. Its account helps us grasp the challenge of making a game among workers fill such a space. The figures are imposing enough to meet us at near human scale, yet none performs an eloquent narrative gesture. Looking between the Met and Barnes works, one notices instead how a slight shift of posture or the insertion of one child changes the social geometry. Both rooms are full of people, though the participants remain absorbed in small rectangles of paper. ^[9]

Three pairs

The remaining canvases each show two men facing one another, but they vary greatly in size. FWN 685, now in a private collection in Qatar, measures 97 × 130 centimetres; the Courtauld’s FWN 686 measures 60 × 73, and the Orsay’s FWN 684 only 47 × 56.5. The current catalogue dates them respectively 1892–96, 1891–96 and 1893–96. Those overlapping ranges do not tell us in which order Cézanne worked on them. ^{[1] [4] [5]}

In the Courtauld canvas, the left man sits upright in a tall tan hat, a pipe extending from his mouth. His cards flare outward like a small fan. The right man, in a paler coat and lower grey hat, bends toward his hand. They face one another without exchanging a glance. Their coats leave a dark channel for a bottle, while the red-brown table pushes towards us. The panelled wall remains shadowy; the table’s hanging cloth and angular legs occupy much of the foreground. The Courtauld dates this canvas 1892–96 and records Samuel Courtauld’s 1929 purchase and 1932 gift. ^{[1] [10]}

The Orsay painting shows a tighter encounter. Its bottle rises between the men like a narrow dark screen. The smoker at left holds a few mostly pale cards just beside it, while the right player’s grey cards are drawn close to his hands. Their coats almost fill the side edges, and the red tabletop is crowded by forearms. There is little room for the eye to rest between the heads, bottle and cards. T. J. Clark reads this compression as distinctive, particularly beneath the table where the men’s knees claim the remaining space. The Orsay dates the painting 1893–96; its commentary’s older-versus-younger symbolism is an interpretation, rather than a documentary identification of the sitters. ^{[4] [11] [12]}

The Qatar picture makes a paired game expansive. A broad orange-red cloth spreads from the men’s hands to the near edge of the much larger canvas. The left player’s high dark hat and blue-grey coat face a man in a pale coat whose shoulder fills the right side. Their heads stay apart above a slender bottle; the left player holds a fan with red and blue marks, while his opponent’s cards show mostly grey backs. The dark wall is divided into horizontal and vertical bands behind them. Compared with Orsay’s dense knot of hands and bottle, the Qatar table opens a long painted interval between the players and the viewer. It is a large two-person image, so fewer figures did not necessarily mean a smaller artistic task. ^[5]

Why the Order Is Still an Argument

It is tempting to arrange the canvases as a smooth movement from five people to four to two. Such a sequence would turn the series into a lesson in simplification. The present evidence allows a more complicated account. The Met and Barnes have overlapping 1890–92 catalogue ranges; the paired paintings have wider and overlapping ranges too. A catalogue date estimates a likely period for an individual object, not every painting session or a precise place in a five-work chain. Earlier scholars have argued for differing orders, and Clark explicitly presents his own as hypothetical. He favours the Met before the Barnes, then the three pairs, but his preference is part of a close reading rather than an established calendar. ^{[2] [3] [11]}

The Met’s 2022 conservation account reports a different sort of evidence. Research carried out for the 2010–11 exhibition used X-radiography, infrared reflectography and close surface examination. It found extensive revision in the Met canvas and helped overturn an older assumption that the smallest paintings necessarily came last. The account discusses the small paintings in relation to the multi-figure canvases; it does not deliver a complete, settled order for all five. Its findings matter because they give a physical counterweight to a story based mainly on how a composition looks. A compact painting need not be a final reduction, and a large one need not be a first attempt. ^[6]

The Met picture itself preserves visible and imaged changes. Conservator Charlotte Hale describes adjustments to figures, table and chairs. The museum’s essay points out the unresolved right player’s hand, which the mass of his blue sleeve almost conceals. These particular findings belong to the Met’s accession 61.101.1. They help us see a picture apparently calm at first glance as a record of prolonged decisions. They cannot automatically describe the material history of the Barnes or Courtauld canvases. In the series as a whole, the lesson is methodological as well as visual: arrangement and finish must be judged in relation to each work’s actual surface and examination. ^{[6] [8]}

Satish Padiyar likewise questions a one-way movement toward fewer people. He considers how the paintings build a world among figures, especially in the Barnes composition, and how changes of scale disturb an easy evolutionary narrative. His account relies in part on technical work that has circulated among scholars; the shorter public Met conservation essay gives the directly available findings here. Reading the pictured series without a settled end point lets the large gathering and the quiet pair remain equally deliberate solutions to different spatial problems. ^{[13] [6]}

Studies, Workers and the Problem of Names

The paintings grow out of sustained work with individual figures. Drawings, watercolours and paintings of men holding cards or smoking pipes do more than furnish a list of preparatory fragments. A person portrayed alone may later appear inside a group, but the direction is not always certain. The Nelson-Atkins Museum’s Man with a Pipe presents the standing smoker as the subject of an independent canvas. Brigid Boyle discusses Nancy Ireson’s proposal that it may postdate the Met composition, based on differences in the sitter’s turn and the angle of his pipe. If so, the isolated portrait would continue an experiment already under way in the group scene. The possibility unsettles the convenient formula that every single-figure study came first. ^[7]

The Nelson-Atkins technical study is valuable precisely because it is specific. Diana Jaskierny reports evidence consistent with possible dilute blue drawing on that museum’s Man with a Pipe , even though carbon-sensitive infrared imaging did not reveal a conventional dark underdrawing. She describes rapid passages of wet paint and later adjustments around parts of that portrait. Those observations concern Nelson-Atkins 2015.13.6. They show the care needed when speaking about how Cézanne worked, while leaving the paint structure of each Card Players canvas to its own examination. ^[14]

A letter dated 13 February 1891 gives a glimpse of modelling at the Jas de Bouffan, the Cézanne family estate near Aix. Paul Alexis told Émile Zola that Cézanne was painting there with a worker posing. Alexis was relaying a report and refers to one unnamed worker. The letter is important evidence for the setting of figure work around this time. It does not identify the worker, date a completed canvas or demonstrate that the five men and child in the Barnes painting sat together around a table. The passage is known here through Dombrowski’s citation and translation; Alexis’s original letter edition has not been independently checked. ^[15]

Later testimony offers possible names but asks for equal precision. Michel Fraisset’s research separates Paulin Paulet , a gardener associated with the Jas, from a model called père Alexandre . Earlier accounts muddled them under the name “Alexandre Paulin.” Fraisset draws on recollections from Cézanne’s son, an interview with Paulet’s daughter Léontine and family photographs preserved much later. He connects the standing pipe smoker in the Met and Barnes scenes with père Alexandre and Paulet with figures in the paired compositions. Orsay’s own account identifies a pipe smoker as père Alexandre; current catalogue entries do not make all these identifications definitive. The name correction is useful, while the distance between a sitting in the 1890s and later recollections keeps individual assignments qualified. ^{[16] [12]}

Fraisset also discusses testimony identifying the child in the Barnes painting as Léontine Paulet. That association is attached to the five-person painting alone. In the Met canvas no child appears, and in the three paired pictures the two players occupy the whole human scene. The child’s uncertain name does not diminish her visible role: she gives the Barnes gathering another kind of attention, closer to the cards than the standing smoker and distinct from the men who hold them. ^{[16] [3]}

Cards, Tables and Painted Space

The cards are among the smallest things in the paintings, yet they generate almost every large bodily arrangement. Each bent neck and lowered gaze leads towards a hand. At the Met and Barnes, the three seated players form a curved enclosure around the table. In the paired compositions, the opposing profiles and elbows make a more direct bridge over it. Some cards show pale backs or coloured faces, while others nearly merge with fingers or paint on the table. The game remains recognisable even as its paper pieces become difficult to separate from the material of the picture. ^[15]

André Dombrowski makes this double character of the cards central to his reading. A playing card is an image that someone handles as a little object: it can be turned, hidden, fanned out, placed face down. Cézanne’s paintings, he argues, exploit that traffic between visible design and physical thing. The reading is especially suggestive at the Courtauld, where the left player’s fan of white cards flashes against a dark sleeve, and at the crowded tables where small rectangles approach the colour and angle of surrounding strokes. Dombrowski also compares transformation decks and contemporary ideas about play; he does not claim evidence that Cézanne owned such a deck or knew those theories. The painted hands and cards themselves supply the clearest point of entry. ^[15]

The table works differently at each scale. In the Met and Barnes it has a white front, a drawer and legs that make a stable, nearly frontal frame below a busy top. This furniture seems to hold the group in place while the heads and arms tilt towards the game. In the Courtauld picture, red cloth and reddish wood push forward, the bottle marks a middle line and the broad underside of the table fills the lower centre. In the Orsay version, the tighter framing makes the space under the table part of the compression. Clark’s account of the changing rooms asks us to notice what is enclosed and what is left open, particularly the watchers in the two larger scenes and the pair who meet only through the game. His preference for one version over another is an argument about experience, rather than a ranking the paintings demand of every viewer. ^[11]

Margaret Iversen offers another way to watch the surfaces. She reads Cézanne’s small, separate marks as allowing variation and chance within an overall series of choices. Her final pages attend particularly to the Courtauld pair: contrasting hats, bearing and proximity introduce a trace of humour alongside their absorption. The reading keeps us from treating stillness as proof that every edge was predetermined. On the Courtauld canvas, the warm tabletop is built from changing reds and ochres; the paler right coat has blue and green touches; the black wall behind the bottle refuses to be a smooth void. An immobile game scene can remain visibly active as painting. ^[17]

Padiyar’s analysis of the Barnes table gives yet another emphasis. He sees it as a shared object that gathers people into a world they inhabit together, while each retains a separate bodily position and attention. This helps articulate why the Barnes canvas can feel sociable without showing conversation or exchanged glances. His later philosophical framework is an interpretive one; the visible anchors remain the hands, cards, pipes, people and table. Charlotte de Mille’s comparison with improvisation in music offers a different metaphor for variations that retain a common theme. It is a way of discussing looking and pictorial ambiguity, not evidence that a particular musical composition supplied Cézanne with a method. ^{[13] [18]}

Cézanne himself once reached for the playing card as a comparison in a very different setting. Writing to Camille Pissarro from L’Estaque on 2 July 1876, he described the red roofs against blue sea in landscape motifs for Victor Chocquet as “comme une carte à jouer” — “like a playing card” (our translation). The remark predates this series by many years and refers to a landscape arrangement. It does show that the flatness and vivid division of a card could serve him as a visual analogy. It cannot be made into an advance explanation of these players or a statement of their intended meaning. ^[19]

What We Know of the Game — and What We See

None of the five paintings supplies enough reliable detail to identify the particular game or its rules. The gestures show holding, studying and playing cards; they do not reveal stakes, score or outcome. Dombrowski notes the temptation to deduce a game from visible hands but leaves it unresolved. That uncertainty need not make the scenes abstract. It sharpens attention to the social facts the pictures do show: the players share a table, possess separate hands and direct their eyes downwards. A pipe in a mouth or on a tabletop is a small sign of a pause, but it does not tell us whether someone is relaxed, anxious or waiting for a turn. ^[15]

The solemn bearing of the farmworker models has invited large claims about the dignity of labour. The Barnes Foundation’s point about grand scale supports that discussion; the paintings also retain awkwardness, silence and uneven proximity. The multi-figure works do not arrange their men as a harmonious public assembly. An observer stands apart. The child in one canvas watches among adults. In the paired scenes, the players occupy opposite ends of a common surface. The seriousness comes from the arrangement we can see, rather than access to private thoughts. ^{[9] [11]}

The question of setting is similar. A wall, pipe rack, curtain, shelf and framed object give the Met and Barnes paintings a room; dark panelling gives the Courtauld pair a backdrop. The records do not secure a particular café. Treating these places as built pictorial interiors lets their differences matter. The Met’s pale wall opens behind the central group, though the right edge is hemmed in by curtain and coat. The Barnes wall rises high enough to place small objects above the people, extending the event into a wider social room. The Courtauld reduces the surrounding world to dark planes so that the players and red table dominate. ^{[8] [9] [10]}

How the Paintings Reached Their Viewers

The five works did not become a unified public series at one moment. Dealer, collector and exhibition histories took different paths. The Met canvas passed through Ambroise Vollard and other Paris dealers, then to Stephen C. Clark in New York; it entered the museum through his bequest in 1960. The Barnes painting also has a substantial dealer and collector history, ending with Albert C. Barnes; one proposed 1923 exhibition identification in its catalogue remains tentative. The Courtauld records a line through European dealers and collectors before Samuel Courtauld’s 1929 purchase and 1932 gift. Such paths matter because later viewers often knew one canvas through reproduction before they could compare several at the same scale or in the same room. ^{[2] [3] [10]}

The Qatar canvas supplies an earlier public point of comparison. The catalogue records it with Ambroise Vollard and Bernheim-Jeune, then the Pellerin family, and lists it in the 1907 Salon d’Automne retrospective after Cézanne’s death. It was catalogued for the 2010–11 Card Players exhibition but was not shown there. This distinction matters when imagining what a visitor could actually compare: the scholarly catalogue could discuss all five while the room itself contained fewer originals. Its route from Paris dealers through Pellerin and later ownership also explains why one of the most physically expansive paired compositions is less familiar through direct museum viewing today. ^[5]

The 2010–11 Courtauld and Met exhibition made comparison a central experience. It brought paintings and related studies into conversation and prompted the technical examination discussed above. Modernist Games , the 2013 volume edited by Padiyar, arose from that context and brought six markedly different readings of the series into print. The arguments in it range from Clark’s rooms and Dombrowski’s handled images to Padiyar’s shared social world and Iversen’s chance in making. Their differences are useful: the paintings resist a single account that would explain every version in the same language. ^{[20] [6]}

Later reception also changed what viewers expected these works to do. Gavin Parkinson examines how Surrealist writers approached Cézanne, including André Breton’s response to The Card Players in the 1930s. Parkinson argues that Breton’s association with a two-player scene belongs to the history of Surrealist reading, while the exact canvas behind it is not securely identified. A later viewer can find strangeness in the tense pairing without that response being evidence for Cézanne’s intended story. Parkinson also shows that Surrealist Cézanne was a contested figure, not the object of one uniform admiration. The series had acquired new meanings long after the models posed. ^[21]