The River (Jean Renoir, 1951)

The near-hallucinatory vividness of Technicolor was seemingly made for Jean Renoir, the son of the Impressionist painter Pierre-Auguste Renoir, but this film was his first use of the process. It's the more remarkable because he was working with his nephew (and Pierre-Auguste's grandson), Claude Renoir, the cinematographer, and neither director nor photographer was particularly experienced in shooting landscape, especially in the country of India, which is the real star of the film. The River has a distinctly Western attitude toward the country, viewing it through the eyes of its British residents. It's based on the experiences of Rumer Godden, the English writer who spent her childhood in India. Her screenplay, co-written with Jean Renoir, is about the tensions between cultures, using the Ganges, the titular river, as a symbol of both the eternal and the mutable. Ravishingly beautiful as the film is, it suffers from some major weaknesses in casting. Its central character, the teenager Harriet, is played by Patricia Walters, a nonprofessional who made no subsequent films and never quite seems at ease before the camera. As Capt. John, the American recovering from the loss of a leg during the war, Thomas E. Breen doesn't have the kind of charisma that would seem to have Harriet, her older friend Valerie (Adrienne Corri), and her Eurasian neighbor (Radha Burnier) falling over themselves to attract his attention. (Breen, incidentally, was both a real amputee from a war wound and the son of the enforcer of the Production Code, Joseph I. Breen.) But for those willing to overlook its flaws, which also include a lack of narrative urgency, The River rewards sympathetic attention and, as a film by a Frenchman about the English in India, stands as a landmark in postwar international filmmaking.

La Pointe Courte (Agnès Varda, 1955)

Philippe Noiret in La Pointe Courte

Lui: Philippe Noiret
Elle: Silvia Monfort

Director: Agnès Varda
Screenplay: Agnès Varda
Cinematography: Louis Soulanes, Paul Soulignac, Louis Stein
Film editing: Alain Resnais
Music: Pierre Barbaud

"They talk too much to be happy," says one of the villagers about the Parisian couple (Philippe Noiret and Silvia Monfort) who are spending time in a small fishing community on the Mediterranean. He has been in the town, where he was born, for several weeks, and she arrives to tell him that she wants a separation after four years of marriage. So they wander around, endlessly analyzing their relationship, as life goes on in the village: The townspeople argue with the authorities about where they can fish and about the bacterial count that has been detected in the water; a small child dies; a young man courts a girl over the objections of her father; a festival that involves jousting from boats takes place, and so on. The sophisticated self-analysis of the couple soon begins to look petty against the backdrop of the village's real problems. At one point in their dialogue, a scene is stolen from them by one of the village's many cats, whose playing around on a woodpile behind the couple becomes far more interesting than what they have to say. Aside from Noiret and Monfort, the actors are all actual residents of the village. Varda began as a still photographer, and her sense of composition is notable throughout this film, her first. She had spent time in Sète, the town where La Pointe Courte was filmed, as a teenager, and was photographing it for a friend when the idea of the movie came to her. It's often cited as one of the first films of the French New Wave, of which she became a prominent member, though seven years passed before she made her second and better-known film, Cléo From 5 to 7 (1962). Alain Resnais, who became one of the leading New Wave directors, edited La Pointe Courte, and the striking score is by Pierre Barbaud.

Night and Fog (Alain Resnais, 1955)

Cynics used to say that the surest way to win an Oscar for best documentary was to make a film about the Holocaust. But when Alain Resnais's Night and Fog was released, it not only received no Oscar nominations, but it was confronted by protests. The German government wanted it to be withdrawn from exhibition at the Cannes Film Festival, and the French censors objected to a scene in which a French police officer was shown guarding one of the deportation centers run by the Vichy government during the war. The French censors also objected to a sequence showing bodies being bulldozed into a mass grave. But it's a testimony to the power of Resnais's editing and the narrative written by Jean Cayrol, a survivor of the Mauthausen-Gusen camp, and spoken by Michel Bouquet, that although such images have grown distressingly familiar over the past 60 years, they still have their power to shock the conscience. It sounds tediously moralizing to reiterate, but every time a politician today tries to dehumanize whatever group is currently out of favor, these images should come to mind.

Hiroshima Mon Amour (Alain Resnais, 1959)

I remember the long dorm-room discussion after my friends and I saw this film for the first time, so I was surprised on returning to it after so many years how conventional the elements we talked about now seem. The extended use of documentary clips at the beginning of the film, with the voiceover by the lovers (Emmanuelle Riva and Eiji Okada) arguing about whether she had really seen anything in Hiroshima, seemed to us a bafflingly random way to start a movie. The jump cuts into and out of flashbacks confused us. What, we argued, did it signify that their entwined bodies, seemingly covered with ashes, then began to glow? (Today, I'm afraid some sarcastic voice will pipe up to say, "They've been glitter-bombed.") Why does she refer to her Japanese lover as "you" when she's actually talking about the German she loved during the war? Is the movie really about sleeping with the enemy? Doesn't it trivialize the horror of Hiroshima to bring it down to the level of the background for a love affair? Today, we'd regard those questions as naive, and I'm certain we wouldn't be confused by the film's structure, which is a way of saying that Resnais and his screenwriter, Marguerite Duras, really did succeed in revolutionizing movies. But if no one is startled by jump cuts or unconventional narrative devices today, there remains a raw immediacy about the film that no subsequent imitators have ever quite succeeded in equaling. Much credit also has to go to the score by Georges Delerue and Giovanni Fusco, and to the editing by Michio Takahashi and Sacha Vierny.

Far From the Madding Crowd (Thomas Vinterberg, 2015)

It's not really easy for me to comment on one version of the Thomas Hardy novel after just commenting on another without resorting to comparisons, so I won't even try to avoid them. One unavoidably obvious difference between the 2015 film and the 1967 John Schlesinger version is visual. In Schlesinger's film, despite the fine cinematography of Nicolas Roeg, the interiors seem impossibly overlighted for a period that resorted to candles and oil lamps for illumination. The change in film technology now makes it possible for us to see the way people once lived -- in a realm of darkness and shadows. (We can almost precisely date when this change in cinematography took place: in 1975, when director Stanley Kubrick and cinematographer John Alcott worked with lenses specially designed for NASA to create accurately lighted interiors for Barry Lyndon. Since then, the digital revolution has only added to the arsenal of lighting effects available to filmmakers.) So cinematographer Charlotte Bruus Christensen's adds an element of texture and mystery to Thomas Vinterberg's version that was technologically unavailable to Roeg, and not only in interiors but also in night scenes, such as the first encounter of Bathsheba (Carey Mulligan) and Sgt. Troy (Tom Sturridge), when he gets his spur caught in the hem of her dress. The scene is meant to take place by the light of the lamp she is carrying, which Christensen accomplishes more successfully than Roeg was able to. But the story's the thing, and David Nicholls's screenplay is much tighter than Frederic Raphael's 1967 version -- it's also about an hour shorter. Nicholls makes most of his cuts toward the end of the film, omitting for example the episode in which Troy becomes a circus performer, one of the more entertaining sections of the Schlesinger-Raphael version. I think Nicholls's screenplay sets up the early part of the stories of Bathsheba and Gabriel Oak (Matthias Schoenaerts) much better, though he has to resort to a brief voiceover by Mulligan at the beginning to make things clear. His account of the affair of Troy and the ill-fated Fanny Robbin (Juno Temple) is less dramatically detailed than Raphael's, but in neither film is their relationship dealt with clearly enough to make us understand Troy's character. On the whole, I think I prefer the new version, which is less star-driven than Schlesinger's, but I can't really say whether I would feel that way if I had seen the new one first.

Far From the Madding Crowd (John Schlesinger, 1967)

People always complain about the way movies change the stories of their favorite novels, but screenwriter Frederic Raphael's adaptation of Thomas Hardy's novel shows why such changes are necessary. Raphael remains faithful to the plot, with the result that characters become far more enigmatic than Hardy intended them to be. We need more of the backstories of Bathsheba Everdeen (Julie Christie), Gabriel Oak (Alan Bates), William Boldwood (Peter Finch), and Frank Troy (Terence Stamp) than the highly capable actors who play them can give us, even in a movie that runs for three hours -- including an overture, an intermission, and an "entr'acte." These trimmings are signs that the producers wanted a prestige blockbuster like Doctor Zhivago (David Lean, 1965), which had also starred Christie. But Hardy's works, with their characters dogged by fate and chance, don't much lend themselves to epic treatment. John Schlesinger, a director very much at home in the cynical milieus of London in Darling (1965) and Sunday Bloody Sunday (1971) and New York in Midnight Cowboy (1969), doesn't show much feeling for Hardy's rural, isolated Wessex, where the weight of tradition and the indifference of nature play substantial roles. What atmosphere the film has comes from cinematographer Nicolas Roeg's images of the Dorset and Wiltshire countryside and from Richard Rodney Bennett's score, which received the film's only Oscar nomination.

The Postman Always Rings Twice (Tay Garnett, 1946)

It's one of the most memorable entrances in movies. Actually, her lipstick enters first, rolling across the floor toward him. She is Cora Smith and he is Frank Chambers, the man her husband has just hired to work in their roadside café/filling station. But more important, she is Lana Turner, one of the last of the products of the resources of the studio star factories: lighting, hair, makeup, wardrobe, and especially public relations. And he is John Garfield, one of the first of a new generation of Hollywood leading men, trained on the stage, and with an urban ethnicity about him: His vaguely presidential nom de théâtre thinly disguises his birth name, Jacob Julius Garfinkle. The pairing shouldn't work: She's a goddess, not an actress, whom the publicists had turned into "the Sweater Girl" while claiming that she had been discovered at a drugstore soda fountain. He was the child of Ukrainian-born Jews and grew up on the Lower East Side, trained as a boxer and studied acting with various disciples of Stanislavsky. But the chemistry is there from the moment Frank picks up Cora's lipstick and the camera surveys her from toe to head: white shoes, tan legs, white shorts, tan midriff, white halter top, blond hair, white turban. She reaches out her hand for the lipstick, but he doesn't move, so she comes over and gets it. It's one of the many power plays that will take place between them. The rest is one of the great film noirs, from a studio that didn't usually make them, MGM. In fact, the studio head, Louis B. Mayer, hated it, which is always a good recommendation: He hated Sunset Blvd. (Billy Wilder, 1950), too. (Mayer's tastes ran to Jeanette MacDonald-Nelson Eddy operettas and the Andy Hardy series.) It's the only really memorable movie directed by Tay Garnett, so I suspect a lot of credit goes to the screenwriters, Niven Busch and Harry Ruskin, and to their source, James M. Cain's overheated novel. Cain also wrote the novels that were the basis of two other famous noirs: Double Indemnity (Billy Wilder, 1944) and Mildred Pierce (Michael Curtiz, 1945), so the screenwriters and the director had some powerful examples to follow.

Things to Come (William Cameron Menzies, 1936)

Works of fiction that pretend to depict things as they will be in a specific place and year tend to look a little foolish when that year actually comes. The years 1984 and 2001 didn't turn out to be precisely as George Orwell and Arthur C. Clarke envisioned them. But neither Orwell nor Clarke expected them to: Both were extrapolating from what they saw about the times in which they were writing. Orwell was viewing with alarm the struggle for power in 1949, and Clarke was elaborating on thoughts he had about the relationship of man, technology, and nature -- for good or ill -- in a series of stories beginning with "The Sentinel" in 1948. It's significant that both of these writers were working from a post-World War II point of view. But Things to Come starts from a very different place: England just before the second World War. H.G. Wells's 1933 The Shape of Things to Come was a meditation on a utopia founded on science, replacing religions, and a world government, replacing nationalism. The adaptation of these ideas in Wells's screenplay involves a world on the brink of war at Christmas, 1936 -- less than three years before the world actually went to war. Wells didn't have to wait long to see the ideas in the film superseded by reality. In the film the conflict lasts 40 years, and is devastating to the old order of things. There arises a kind of technotopia, which then has to battle with (and triumph over) reactionary, anti-science forces. We no longer have the kind of faith in technology to solve all problems that Wells possessed -- in fact, if the atomic outcome of World War II is any indicator, technology presents as many problems as it solves for humankind. Things to Come is muddled but fascinating: It raises the right questions while providing unsatisfactory answers. The best things in the film are the ones closest to home. For example, Ralph Richardson's performance as the dictator known as "The Boss" -- a slangy translation of Il Duce. Richardson was one of the three greatest English actors of the mid-century, but unlike Laurence Olivier and John Gielgud he rarely got a chance to show his stuff in movies. Here, his eccentric manner is the key to the role, and he plays it to the hilt. Unfortunately, Menzies, a gifted designer, wasn't much of a director, and he surrounds Richardson with inferior performers: Margaretta Scott, who had a long career once she grew accustomed to film acting, here recites her lines as if reading them for the first time and assumes poses copied from silent film vamps. For contemporary viewers, the most interesting things about the film are the set designs by Vincent Korda and the fantasias about what people will be wearing in 2036 -- which in Wells's scheme of things is the year of the first voyage around the moon. The costumes are credited to John Armstrong, René Hubert, and the Marchioness of Queensberry. (Her given name was Cathleen Mann; a portrait painter and costumer, she was married to the 11th Marquess of Queensberry from 1926 to 1946.)

H.G. Wells, Pearl Argyle, and Raymond Massey on the set of Things to Come

Days of Wine and Roses (Blake Edwards, 1962)

This melodrama about alcoholic codependency threatens to fall into didacticism, becoming a latter-day temperance lecture, but is rescued by the fine performances of Jack Lemmon and Lee Remick as Joe and Kirsten Clay. He's a ladder-climbing public relations man and she's the secretary to one of his clients; they fall in love, get married, have a child, and turn into self-destructive lushes. Eventually, with the help of Alcoholics Anonymous, and after a couple of harrowing relapses, he climbs out of it, but she refuses to admit that she has a problem that can't be solved with "will power." The film is unexpectedly bleak for one made with a solid Hollywood budget and two big stars -- both of whom received Oscar nominations -- directed by a man more famous for the Pink Panther movies and for his marriage to (and films with) Julie Andrews than for a serious problem drama. Fortunately, the film has a point to make: that alcoholism is a disease that manifests itself differently in each person who suffers from it. Joe, being a sociable type whose job has always involved drinking with clients, is the kind of person who benefits from the sense of community that AA provides. Kirsten, on the other hand, is a loner: an only child with a doting father (Charles Bickford), who when we first see her doesn't drink at all and is given to taking long walks alone on San Francisco's Fisherman's Wharf. It's Joe who introduces her to alcohol, which softens the rough edges of life -- without it, she says, everything looks "dirty." She feels comfortable denying her problem, even when it affects her marriage and her child so severely: At one point, she sets fire to their apartment in an alcoholic haze. They love each other, but she's unable to express her love for Joe unless he drinks with her. The screenplay by JP Miller is a reworking of his TV drama that appeared on Playhouse 90 in 1958, starring Cliff Robertson and Piper Laurie. There is a bit too much Hollywood gloss on the film, including an Oscar-winning title song by Henry Mancini and Johnny Mercer, but the thoughtful core of the narrative manages to surface because everyone resisted the tendency to paste an easy resolution of the Clays' problems on the end of the film.

Fanny and Alexander (Ingmar Bergman, 1982)

Artists' reputations often take a severe hit as time passes: No one thinks Walter Scott was as great a novelist or poet as his contemporaries did; today, he's read only by specialists, and then often grudgingly. So it's not surprising to find people who think that the directors who revolutionized filmmaking in the 1950s and '60s, like Ingmar Bergman, Federico Fellini, and Akira Kurosawa, are mannered and overrated. Some of Bergman's early films, I think, are: Smiles of a Summer Night (1955) is better in its Stephen Sondheim musicalized version; Wild Strawberries (1957) is a meditation on aging by someone who hasn't aged; The Seventh Seal (1957) and The Virgin Spring (1960) strive for a mythic quality that the material and the medium won't bear. But in his later career, after he had ceased to be the darling of the "art houses," he made some intensely personal films that have the warmth and humanity that he could only feint at in the early ones. Critics tend to prefer his films about women, Persona (1966) and Cries & Whispers (1972), but I find him most genuine in his exploration of childhood, especially his elegant reworking of Mozart's The Magic Flute (1975), which sees the opera through childlike eyes, and Fanny and Alexander, which works with a kind of double-vision: We know what's going on in the various sexual combinations and permutations of the Ekdahl family and their lovers, but we also have the point of view of Fanny (Pernilla Allwin) and especially Alexander (Bertil Guve) to elevate them from the merely physical and sometimes sordid into the realm of mystery. If seen from the point of view of the children, this is a kind of ghost story. Alexander will carry into adulthood the experience of seeing two ghosts: one benign, his real father (Allan Edwall), and one malign, his stepfather (Jan Malmsjö). It is also a fable about two kinds of imagination: the artistic and the religious. And we know which side Bergman comes down upon rather heavily.