Dog Day Afternoon
(Warner Bros.)
At least once a week there's something in the newspapers that is believable only because it actually happened. And in the current strained relation between fact and fiction in our culture, fictionmakers of prose and film sometimes let life do the inventing for them, then use fictional techniques of conviction on the nearly incredible material. The producer Martin Bregman sponsored Serpico (TNR, Jan. 19, 1974) which was made in that impulse; out of the same impulse he now presents the much-superior Dog Day Afternoon.
Late on a steaming August day in 1972 two robbers held up a Brooklyn bank. (The picture guarantees the facts in a prior note; I remember the story but am relying on this guaranty of accuracy.) Before they could get away they were spotted, and police surrounded the bank. The robbers had eight hostages, so the police, joined by the FBI, began negotiating. In the course of the parleys it came out that the leader of the pair, a young man with a wife and children, was in this robbery because he wanted money to pay for the transsexual operation of another young man, a homosexual who was his second "wife." The FBI arranged for a getaway plane in return for the hostages' safety. At the airport the FBI killed one robber and captured the other, the leader. The hostages were unhurt. The leader is now in prison. The homosexual "wife" somehow financed the operation and is now a woman living in New York.
Three others who worked on Serpico are also involved here: the director Sidney Lumet, the editor Dede Allen, and the star Al Pacino. Their picture, though it runs a bit long, is generally gripping, sometimes fine, more often funny. And eventually unsatisfying.
The screenplay is by Frank Pierson, who wrote Cat Ballou and Cool Hand Luke and worked with Lumet on The Anderson Tapes. I assume that much of the dialogue was loosened up by the actors in performance in moderate improvisation, but Pierson at least provided some good scenarios. The script's structural defect comes from the fact that it was strapped to the story as it happened.
The photography by Victor J. Kemper, who made The Candidate, supplies just sufficient color (an important point), and Dede Allen's editing is one of the sureties of excellence in a wobbly world. Note, for instance, in the street scenes outside the bank, how deftly she works between long high shots and the close shots, how easily we are given the proceedings both as a news event in the world and as personal drama.
But the film's quality—and it has a quality, sad, desperate, ludicrous, entangled—comes from the collaboration of Lumet and his actors. First, right down the line, he has cast every part perfectly. Then, with his principals, he has worked with just enough control and enough relaxation to create tight naturalistic surfaces that evoke ambiguous inward states—a dialectic that parallels Allen's editing.
My admiration for Lumet and Pacino has had discernible limits in the past, but here, happily, I can change my tune. Lumet, at 50 much less of a show-off than Lumet at 35 or 40 (The Fugitive Kind, The Group), directs with simplified apparatus and deepened perceptions. Pacino is not called on here to radiate quiet power, at which he failed in both Godfather films. Here he is fortune's fool—a pawn who is being played at the same time that he thinks he's masterminding the game. This summer I went to a racetrack for the first time in 10 years, and outside the gate there was a tout selling a tip sheet for a dollar—eight sure winners. His elbows were sticking out of his ragged jacket, but he had eight sure winners. Pacino, about to lick his condition in the world and his personal/sexual troubles by being smarter than the world, has his figurative elbows sticking out. The pseudo-superiority, this pathos of self-deceived bravado, Pacino handles very well.
The film begins like a comedy of imbecilic errors. There are three bandits at first, but one quits just because he has "bad vibes." The second is John Cazale—Pacino's executed brother in Godfather II—here tacitly effective as an ex-convict who, in his quiet way, would rather die than go back to prison. He and Pacino bungle a good deal, comically, as they line up the tellers and rifle the bank, even though Pacino has worked in a bank and has a lot of inside info. (Eight sure winners.) Luck and faulty forethought are against them. A bank shipment has left only $1100 in the vaults. When Pacino cleverly burns the record book that lists the bill denominations, the smoke goes out an air vent and alerts a neighbor who calls the police.
Negotiations between Pacino and a detective, Charles Durning, immediately move on to the Treaty of Locarno level that these talks always have. Big crowds quickly gather behind street barricades. The talks soon become street theater, and Pacino soon senses that he has, perversely, become the hero of the drama. (Nicely appreciated by Lumet.) When his mother is summoned for a tearful, profane street scene with him, she turns out to be, with cunning aptness, Judith Malina of the Living Theater, noted for its street activities.
Pacino's wife has also been sent for. We get a glimpse of police calling on a fat emotional woman in a squalid apartment. But when a patrol car approaches, announced as bringing the wife, out steps an unshaven young man in a hospital bathrobe—a homosexual who has been in Bellevue with drug troubles. He is taken to the barber shop across from the bank that is serving as police command post. Then follows the high point of the film—a scene that is surely going to be a locus classicus of naturalistic acting for a long time to come—a telephone conversation between Pacino in the bank and, just across the street, his "wife," played nervously, delicately, free of cliche, by Chris Sarandon.
Lumet and Allen handle it superbly. It begins in the barber shop with Sarandon seated, almost incidentally, at the side of the frame, and the camera casually saunters up to him as the coils of his life have done. Gradually he is centered, as the shots of Pacino at the other end are cut in; and (under Lumet's good guidance) they play a moving scene of affection, resentment, subtle horror and farewell, Pacino's lonely brown eyes speaking almost more eloquently than his voice, Sarandon fighting for breath and life against the passionate suffocation of this madman across the street. They say goodbye, knowing they will never meet again.
The film's biggest act of daring is not in its use of homosexuality, which is no longer so great a novelty, but in the fact that it is proletarian homosexuality and that it is left quite unexplained. Pacino's blubbery wife, his strident mother, his stolid accomplice, all accept the fact that he has a male "wife." Theater, film and TV have dealt with homosexuality generally as the province of the cultivated and "artistic." But this is hardhat country. Hubert Selby explored prole homosexuality in Last Exit to Brooklyn. Has anyone else? The matter-of-course acceptance of Pacino's relationship here is not a try for sensationalism nor a manifest of latter-day broadmindedness: it's an acknowledgment of conditions that are taken for granted by that sector of society. No apologies and no psycho-sexual explanations, just acceptance in this supposedly puritanical class, that's the picture's quiet bombshell—assuming that the script is sticking to the facts.
Yet it's that very adherence to fact that finally makes the picture unsatisfying. Good art often mixes modes but always to some point selected by the artist. Here the final effect is only of a crazy series of events: the clownishly bungled robbery, the comedy of the hostages mixed with their real danger (Cazale is quite willing to kill them), the floppiness of the fat wife, the naked telephone dialogue, the snappy G-man ending—tonality after tonality tumbling by. But life, as has been well established, is no artist; and when Gazale's body is wheeled past Pacino at the end and he looks at it numbly, what are we supposed to feel? Pathos? Society's triumph? Something else? I don't know. The facts of the dog day don't clarify Pacino's character—as hero or social victim or moral anarchist or psychic deviant or anything else—and the script is bound to the facts. No matter how good a film is segmentally, if we don't know how we're supposed to feel at the end, it can hardly be called satisfying.
On the other and substantial hand, most of those segments are very good. If the whole is less than the sum of the parts, if there really is no sum of the parts, those parts those parts are extraordinarily well made.
This article originally ran in the October 4, 1975 issue of the magazine.
Stanley Kauffmann is the film critic for The New Republic.