One can often tell a cinephile by the rituals they establish. For my part, I begin every summer by revisiting Jacques Tati’s Monsieur Hulot's Holiday (1953), the feature debut of his most beloved character. I can no longer remember what drew me to this habit outside of a strong association of the season with the smooth jazz theme to the film “Quel temps fait-il à Paris?”, written by Alain Romans. Revisiting the film last summer, I decided for the first time to put on the 1953 version of the movie instead of the 1978 version I usually watch, which is labeled “definitive” by Les Films de Mon Oncle, the foundation responsible for the restoration and rerelease of Tati’s films. Outside of one addition to this later cut, I was unaware of the differences between them, and couldn’t find much information about the original release. Almost immediately, I was shocked to see things previously so familiar to me suddenly appear in a different light. The film seemed even fuller than it had previously, and in some ways it was: the original release ran 98 minutes, while the later version was trimmed down to 86. More than this, the tone had shifted in significant ways, the social satire sharper and a bit edgier, enough to alter what I thought I understood about the film.
If anything unifies the variety of critics who come together to adore Tati, it’s the belief that his films reward repeated viewings. The evolution of my own relationship to Monsieur Hulot's Holiday mirrors the response from many others: I first found it very likable, but only after a few further revisits did I grow to love it. The film is clearly experimental, but because its experiments are delivered with modesty, it is hard at first to gain a full view of what they achieve. One of the reasons for this, I suspect, owes to the ways in which Tati strives to point out how much we take for granted in narrative film. Among other things, he does this through his renunciation of the close-up. The description of the film by Tati critic Michel Chion as a child, that it was a “film without faces,” is certainly a resonant one.1 Strictly speaking, there are no close-ups in the film and only a handful of medium shots, most important among them the arrival of Hulot at the check-in counter of the hotel where he introduces himself to the audience (and is even courteous enough to spell it out for us, “H-U-L-O-T”).
This hesitancy to eliminate our distance from the characters encourages us to take in more of the scene in any given frame, though we may not realize it at first. Among the things that appear more clearly through repeated viewings is how well every character on holiday with Hulot is defined. We’re used to Hollywood extras, ambient and basically anonymous, whose purpose onscreen is closer to furniture than to character. It takes a few viewings to realize how different Monsieur Hulot's Holiday is in this respect: nearly every character, circumscribed by some key trait, possesses an interlocking adventure comparable to that of the title character and equally deserves our attention. It’s easy to miss this at a glance. After all, it’s Hulot’s holiday, so we’re told, and it seems only appropriate to assume he holds a certain centrality to the film. Perhaps he is more responsible for the action here than in PlayTime (1967), where the character is so diffuse that he seems to disappear for long stretches of time. Yet it becomes clear that the unusual ensemble approach to plotting that is perfected in PlayTime finds its origin in Monsieur Hulot's Holiday. Tati’s comedy is often compared to the great slapstick comedians of the silent era—Linder, Chaplin, Keaton—and, at first, the comparison holds true. His early role in René Clément’s Soigne ton gauche (1936) is a solid example of one of the most reliable narrative archetypes of the genre, the amateur boxer. But the comparison holds no meaning by the time he arrives at Hulot; there is no equivalent humility in any of the other great works of slapstick, no comparable attempts to erase oneself from the center. Hulot is for Tati neither what the Tramp is for Chaplin (an index of the social problems of his time, who personifies its suffering) nor what Antoine Doinel is for Truffaut (a diary of the author’s fantasies, if not his realities). He’s more elusive. The difference is not of quality, but of genus.
Tati released three versions of Monsieur Hulot's Holiday: the first in 1953 (withdrawn in 1959 by Tati), the second in 1962 (taken out of circulation and held in escrow by the French courts following Tati’s bankruptcy in 1967), and the final version in 1978. Considering the first and the last versions together, I began to wonder how we might understand the description of the latter as “the one the director intended.” Ordinarily, when we use the term “director’s cut,” we use it to set one release against a version compromised by some form of studio interference. Thus, we might be inclined to speak of a “director’s cut” of Orson Welles’s Mr. Arkadin (1955) or Erich von Stroheim’s Greed (1924), two great examples of artists at odds with an unartistic industry.2 But this hardly describes the situation Tati found himself in while making Monsieur Hulot's Holiday. On the contrary, seldom do we have examples of filmmakers with more resources and greater creative autonomy than Tati after the unexpected runaway success of his previous feature, Jour de fête (1949), which according to an estimate by his producing partner yielded some 22 million francs for the director.3 Budgeted at 105 million francs, five times the amount of his debut, Monsieur Hulot's Holiday certainly does not appear to have suffered from financial want, and as it was produced by Tati’s own company, he wasn't artistically constrained like Welles or Stroheim. Moreover, whereas we ordinarily expect a director’s cut to run longer than the commercial release, the final version of Holiday is over ten minutes shorter.
Not unlike the painter in Balzac’s The Unknown Masterpiece, Tati was constantly tweaking and revising his films. Of his six features, Monsieur Hulot’s Holiday saw the greatest number of changes, and Tati himself regarded this final version of Holiday as the definitive one before his death in 1982. Several aspects of this later release are an undeniable improvement, above all the new arrangement of the main theme with expanded instrumentation (the original contained a simple piano score) and the faster narrative tempo. Yet in other cases, the value of the changes is debatable. To choose the best-known example: when Tati finally received legal permission to re-release Holiday in 1977, he returned to Saint-Marc-sur-mer to shoot new footage expanding the gag near the middle of the film in which Hulot’s kayak splits and folds in half on top of him. In homage to the enormous success of Jaws (1975), Hulot now rips four triangular, tooth-like sheets from the ship’s hull as he struggles to escape the boat, leaving the wreck looking (vaguely) like a shark. A new crowd was brought to the beach, who are filmed shrieking and running for the hotel.
To me, this gag has always stood out as awkward. Perhaps it’s simply that it’s the most unbelievable moment of the entire film, running contrary to Chion’s astute observation that Tati’s gags post-Jour are obliged to obey the laws of comic plausibility, whereas this one seems too much a leap of faith.4 The discovery that this scene did not belong to the original release came as no surprise, but it opens up a more interesting question: what do these different versions reveal about Tati’s intentions for the film at different times? How did this work change over time? The more times I watch the original release, the more I come to prefer several aspects of it. Yet I should quickly add that my aim is not to suggest that the earlier release is necessarily superior. I want to make a case for the unique brilliance of the '53 release, not at the expense of the later version but simply on its own terms. They should be taken complementary, as two distinct drafts of a great work that offer us different views of its artist and his aims.
Of the three versions of Monsieur Hulot’s Holiday, only the first and the third are available on home video in the United States, thanks to the complete set of the director’s works released by the Criterion Collection in 2014. In addition, we also have access to the original script for Monsieur Hulot's Holiday via the Complete Jacques Tati set published by TASCHEN Books in 2019, allowing us to compare the original conception for the film with its first and subsequent releases. The absence of the second release makes it difficult to track when all of the changes were implemented by Tati, but we’re aided by the novelization of the film, written by a young Jean-Claude Carrière—one of his first jobs as a writer—and illustrated by Pierre Étaix, released in 1962. Two key differences between the novel and '53 release are worth emphasizing. The first, maybe the most surprising, is the perspective of the novel, which is written from the point of view of the elderly gentleman, played by René Lacourt, who strolls along the beach behind his wife throughout much of the film. It’s unclear whether the responsibility for this decision to promote this character—known in the script simply as “the Strolling Man”—to the center of the novel belongs to Carrière or Tati, but this knowledge does bring to our attention how frequently he appears in the film as a spectator watching the other guests. More than any other film of his, Monsieur Hulot's Holiday is a film of glances, and so much of its brilliance owes to how intricate this exchange of glances grows over the course of the film. The second change is to the ending, which is more in line with the conclusion of the '78 release, from which it’s safe to deduce this change was introduced in the '62 release.
Comparing the 1953 and 1978 versions, the most typical change is the most banal: trimming bits here and there from gags to tighten up the pace of the action. Two shots are cut from Hulot’s first entrance at the Hotel de la Plage, where big gusts of wind blow through the lobby and startle the other guests. Very often, the '78 release cuts the beginning or end of a gag that has overstayed its welcome, such as Hulot’s fearful encounter with a bucking horse near the end of the film. Equally common are the rearrangement of inserts and point-of-view shots.5 Consider, for example, four shots from Martine’s arrival at the beach. When Martine first arrives at the beach in the '53 release, she looks behind her and spots a small family of four sporting big sombreros (described as “Mexican style hats, all worn in different ways” in the original script). Here, the second shot in the sequence is from her point of view, sandwiched between the first and third shots of her looking out toward the beach.
In the '78 release, the order of the shots changes: this time, the fourth shot is from Martine’s point of view instead of the second, and what she sees on the beach—viewed over her left shoulder instead of her right, as in the original release—is another young woman running toward the waves.
A similar change occurs after she crosses the street. In both versions, Martine looks behind her to see two young men gawking awkwardly at her, who we later see attempt to win her over with cigarettes and jazz. In the '53 release, this shot is followed by an establishing shot from the opposite side of the beach.
While in the later one, it cuts to a shot introducing the strolling couple instead, who view this interaction between Martine and the boys from a bit further up the street.
Such changes are common but do little to alter our understanding of the film. Some appear motivated purely by rhythm: the strolling man’s stride lines up perfectly with the speed of theme music, for instance. (The earlier introduction of the strolling couple, who don’t appear until 14 minutes into the original release, may also offer a clue to their unconscious importance to our comprehension of the film’s narrative in later versions). Other times, these changes improve the logic of the scenes. The cross-cutting during the picnic near the end of the film, switching between the hotel guests on the grass and Hulot working hopelessly on his car several miles away, is greatly reduced to allow the scenes to play out more clearly in complete segments. In other cases still, some minor plot points are trimmed down in the final release. Hulot loses a tennis match early in the '53 release, making his crushing victory later in the film more amusing.
Does it matter much to see Hulot fumble at tennis before growing into a star player? Hardly, though for other characters the elimination of narrative beats and details does transform their function. In the '53 release, several of the characters are simply designed more fully, with a bit more personality and certainly with more to say. The character to suffer the most from these changes is Martine’s aunt, whose speaking role is almost entirely cut out of the final version. In the original release, she frequently appears to comment on the action of the film—a foil to Martine’s practically silent performance, an aspect of both releases—but in the final cut her part is considerably reduced, including her role in the original finale, which saw her and Martine in a train car discussing photos they took of the guests during the vacation.6
A similar fate befalls another notable character, the communist-existentialist Pierre Aubert (listed in the script as “The Intellectual”), whose monologues decrying capitalism are reduced in the final release. In one instance, as the character rides alongside Martine to the picnic, Tati snips off the beginning of his monologue referring to Lettrism, a popular avant-garde literary and film movement contemporary to the original film.
The removal of this reference is one of many examples that point to the most regrettable difference, the overall elimination of references to '50s France. Most glaring are the major changes to the audio in two scenes, which are possible thanks to Tati’s refusal to record synced sound. His ability to play with the construction of the film as a whole is primarily thanks to this decision, which left him with an even greater creative liberty to assemble and transform the meaning of the footage in inventive ways by adding or altering sounds in the edit to change or enhance the meaning of the scenes. Consider again Hulot’s introduction at the Hotel. In the '78 release, the patrons listen to an orchestra performance over the radio, interrupted by their bumbling new guest. Yet in the '53 version, this scene is set to a news broadcast listing political items relevant to the '50s audience, above all references to the first Indochina war (the French precursor to the American war in Vietnam).
Another reference to Indochina occurs much later in the film, when Martine checks in on her uncle listening to the radio at night.
Again, in the '78 version this is replaced with music.
Most likely, these references were removed because they were no longer timely for the new audiences seeing the film for the first time 25 years later. Yet this brings us to a familiar philosophical problem: whether a work is better served left as it was in its time, or “updated” for contemporary audiences. For numerous reasons, Tati chose the latter, for the same reasons he created English-language versions of his films rather than release them with subtitles: as his biographer David Bellos notes, Tati wanted his films to appear as though they belonged to the same time and place as the audience every time they were viewed. Yet this decision blunts the satire of the original film. True enough, the kinds of characters he depicts—overconfident communist intellectuals, emotionally distant businessmen, buffoonish tourists—still exist, but the '53 version gains from the historical specificity of its milieu that later iterations neither develop nor recover. The sharpness of its satire and moral meaning is enhanced the more specific it is. It has a real bite to it that is mostly removed by the final release. There’s also a meanness to the original version mixed in with its delights. Without these references to contemporary wars, it is harder to notice, for example, the way that the sputtering of Tati’s car and even more the sounds of the fireworks in the film’s finale resemble mortar fire, an observation that led André Bazin to write, in his celebrated essay on the film, “sound is what gives Monsieur Hulot’s world its depth, its moral contours.”7
Do these changes to the sound, large as they may be, dampen the meaning of Holiday as a whole? To my mind, the answer is no, nor does it lessen its effect on the audience. In every version, including the literary adaptation, Hulot's first excursion remains an feat for the immense scope of its vision, for Tati’s skill as both a performer and an observer of human follies, and for the rich detail with which he fills this small French beach. The gains to be had from the release of these different versions belong, like the three drafts of Marx’s Capital or the many versions of Joyce’s Ulysses, mostly to the eyes of the beholder. For some, maybe even for most, one is plenty.
I find these changes fascinating to consider, less out of a pursuit of the “most authentic” version than for how these drafts open up a work I already love even wider than before, leaving room for discovery while deepening my original appreciation. We should be grateful for the rare opportunity to see so many drafts of the film, and admire how modular the work can be. No single version contains everything Tati has to say, and it’s clear in contrasting the two primary versions that even what he had to say is not the same between them. In view of this, one version cannot be preferred over the other without qualification. Tati’s experience is among the rarest in film history, granted an unusual amount of autonomy during production and an equally unusual amount of creative freedom in the edit. For that reason alone, the chance to see this liberty in action is something to be treasured.
1. Chion, Michel, The Films of Jacques Tati, translated by Antonio D’Alfonso (Toronto: Guernica, 2003), 18. Chion’s comment pairs well with a strikingly similar one made by Roger Ebert in his review of the film: “gradually, we see faces.” Ebert, The Great Movies (New York: Broadway Books, 2002), 300.
2. Though whether we should is an altogether different question. Jonathan Rosenbaum, whose analysis of the various versions of Touch of Evil and especially Mr. Arkadin have clearly influenced this piece, wisely cautions us from an all-too-frivolous, if not simply Quixotic, use of the term. See in particular his essays “Death by a Thousand Directors Cuts” and “Potential Perils of the Directors Cut,” both available at jonathanrosenbaum.net.
3. I owe this and many other historical details in this piece to the biography by David Bellos, Jacques Tati (London: The Harvell Press, 1999), 350.
4. Chion, 23.
5. The recycling of footage owes in part to the fact that the unedited material from the shooting of Monsieur Hulot's Holiday was mistakenly destroyed in the mid-fifties, limiting Tati to work with shots from the original release alone.
6. This sequence appears differently in the original script, which ends in the train car with Martine rather than returning to the beach to see M. Hulot drive away from the Hotel de la Plage, as he does in the final shot of both the 1953 and 1978 versions. See Monsieur Hulot’s Holiday in The Complete Jacques Tati, Volume 2, edited by Alison Castle (TASCHEN, 2019), 79.
7. Bazin, André, “No Script for Monsieur Hulot,” in The André Bazin Reader, translated by Timothy Barnard (Montreal: caboose books, 2023), 346.