Akira Kurosawa’s film, Rashomon, is famous as an unsolvable puzzle: multiple clashing viewpoints, with no truth to be found. If we view the film through the eyes of the sociology of fighting, however, one of the four witness accounts is true to life. The others are largely false.
The film tells the story of a murder and rape set in Japan during a lawless period of the 12th century. There are four witnesses.
The bandit says he was aroused by the sight of a beautiful woman on horseback being led through the woods by a samurai. The bandit offers to show the samurai where a cache of weapons is hidden in the forest; when they arrive there, the bandit seizes the samurai from behind and ties him up, then lures the woman to the forest glade and rapes her. Afterwards, she tells the bandit that she can’t live with the shame of being seen by two men, and that one of them must die. The bandit unties the samurai and gives him his sword back. They fight heroically in classic samurai style. The bandit brags about it after he is captured: no one ever clashed swords with me for twenty strokes; we fought like tigers until I killed him with the twenty-third stroke. But the woman had run away.
The woman says that after the rape, she rushed to her husband and cut him free with her dagger. But in his eyes she saw only a cold look of loathing. The bandit was gone. She tells her husband that she can no longer live with him, and asks him to kill her. When he refuses, she loses consciousness, then awakens to find she has stabbed him. She tries to kill herself but she hadn’t the strength. So she ran away.
The dead man’s story is told by a spirit-medium, to testify before the police investigator. After the rape, the samurai remained tied up, listening to his wife’s conversation. The bandit tells her that her virtue is stained, so that her husband won’t take her back; why not marry him instead? She suddenly cries out, Kill him! I can’t marry you as long as he lives. The bandit angrily knocks her down, and asks the samurai what he should do with her, kill her or let her live? While the samurai struggles to answer, the woman escapes into the forest, and the bandit cuts his bonds and disappears. After a long silence, the samurai hears someone crying: it is himself. He finds his wife’s dagger and stabs himself.
These three witnesses comprise the story as written in 1922 by Ryunosuke Akutagawa, which Kurosawa used as the basis for his 1950 film. But Kurosawa now adds a fourth witness. In the original, there is a woodcutter who finds the body. In the film, Kurosawa has the woodcutter tell his story, not to the police inspector, but to a small group of listeners at Rashomon gate, where they are waiting out a rainstorm. The woodcutter, hiding among the trees, saw the rape and its aftermath. The two men fight, but not at all in the heroic samurai style. Both are tense and fearful; they hang back, make sudden charges and retreat again. They swing wildly and can’t keep their feet, falling into the bushes in their uncontrolled rushes. Finally the samurai drops his sword and gets tangled in branches-- an easy target, finally, for the bandit to stab him through the heart.
Real fights in sociological observation
The fourth version is true to life. Fighting in films and in literature has almost always been depicted inaccurately, portraying fighters as more heroic than they really are. In the last 20 years, as real fights have been captured on video and CCTV, sociologists find a very different pattern. Fighters are tense and mostly incompetent. They swing wildly, shoot inaccurately and hit the wrong targets. Because of the tension, angry disputes often end in standoffs before they get going; most fights abort without a clear winner. This is the pattern in fist-fights as well as gun violence. One-on-one confrontations are the hardest to carry off; most such fights abort. The exception is where there is an audience who cheers on the fighters, making it more like a boxing match or a duel, where the social pressure of the group keeps them fighting. Fights between evenly matched antagonists have the highest tension; violence is successful mainly when it consists in the strong attacking the weak, catching them off guard in an ambush, or by a group ganging up on a single individual-- the most common pattern of violence seen in riots.
In Rashomon, the fighters are evenly matched-- the bandit versus the samurai. At the very beginning, the bandit attacks the samurai from behind and ties him up; attacking from behind is a favorite tactic of robbers, giving a psychological advantage, avoiding the tension that results when the contenders stare in each other’s face. Later-- in the woodcutter’s account of the sword fight-- they are evenly matched, and hesitant to fight. The woman goads them into fighting, screaming that neither is acting like a real man; but once the fighting begins she is terrified, can scarcely bear to watch the fight, and runs away. This fight fits the pattern of the most difficult kind of confrontation-- one-on-one, without a supporting audience.
And here the fighters are very incompetent. Their sword-swings and lunges are clumsy; they screw up their courage, then run away; they have trouble staying on their feet. This clumsiness is common in cell-phone videos of fighters, whose wild swings often throw themselves off balance; in street confrontations with guns (as among rival gangs), there is a lot of wild firing that misses its target. The threat of a violent confrontation generates a surge of adrenaline, the flight-or-fight hormone, tensing up the body to go either way. The mythology of fighting pretends that the adrenaline surge (called “heart” or “courage”) always presses forward in a determined attack; in reality, most fighters either retreat or at best keep themselves on the spot by an effort at self-control, their body pulling two ways at once. This is the mechanism that produces heavy breathing and trembling limbs, with the result that fighters often can’t control their fists or their weapons.
Confrontational tension in sword-fighting
This is what we see in the woodcutter’s account of the fight: they are tense, breathing heavily, gasping for breath, wearing themselves down as the fight proceeds. The bandit only wins because his opponent loses his sword and becomes tangled in the bushes. Suddenly it turns into an unfair fight, the strong attacking the weak, and this is when the bandit gets enough control over his bodily tension to kill him. After the fight, he is so debilitated that he can barely walk away.
Almost all our evidence of realistic violence in video and first-hand observation comes from modern times, in fights with fists or guns. Does it also fit sword fights? Certainly there are a lot of Hollywood movies and TV series of medieval heroes and sword-and-sorcery dramas, showing sword-fighters in the mythical heroic mode: never afraid, always attacking and counter-attacking; far from being clumsy, they make acrobatic moves, especially when the hero has to swirl around fighting his way through a crowd of opponents on all sides. This is also the style of kung-fu films and Chinese flying-dragon films, where the acrobatics are enhanced by computer-generated images. And Japanese samurai films-- including those made by Kurosawa in his long career after he made Rashomon-- also show Zen-inspired warriors, flashing their lightning sword-thrusts and making the graceful moves of a martial arts school routine. All this means is that sword-fighting films, both the Western and the East Asian versions, are designed to be an entertaining spectacle. It’s all done in the studio, and the film editing. Movie sword-fights are no more accurate than movie fist fights or gun fights.
Here are two pieces of direct evidence. A samurai in 1864 just before the Meiji revolution in Japan describes a night-time encounter on the streets of Edo (Tokyo):
“The time had already turned an hour past midnight-- a cold and clear winter night with the moon shining brightly overhead. Its silent, white beams made me feel unusually chilly for no good reason. I walked along the broad, vacant street-- no one in sight, absolutely still. Yet I remembered that strolling ruffians had been appearing every night, cutting down unfortunate victims at dark corners.
“I saw a man coming toward me. He looked gigantic in the moonlight, though now I would not swear to his stature at all. On came the giant.
“ ‘I cannot run back,’ I thought, ‘for the rascal would only take advantage of my weakness and chase me more surely. I had better go ahead. And if I go ahead, I must pretend not to be afraid. I must even threaten him.’
“I moved diagonally to the middle of the street from the left side where I had been walking. Then the other fellow moved out too. This gave me a shock, but now there was no retreating an inch. If he were to draw, I must draw too. As I had practiced the art of iai, I knew how to handle my sword.
“ ‘How shall I kill him? Well, I shall give a thrust from below.’
“I was perfectly determined that I was going to fight and felt ready if he showed the slightest challenge. He drew nearer...
“Now there seemed no alternative. If the stranger were to show any offense, I must kill him. At that time there was no such thing as police or criminal court. If I were to kill an unknown man, I would simply run home, and that would be the end of it. We were about to meet.
“Every step brought us nearer, and finally we were at a striking distance. He did not draw. Of course I did not draw either. And we passed each other. With this as a cue, I ran. I don’t remember how fast I ran. After going a little distance, I turned to look back as I flew. The other man was running, too, in his direction. I drew a breath of relief and saw the funny side of the whole incident.
“Neither had the least idea of killing the other, but had put up a show of boldness in fear of the other. And both ran at the same moment... He must have been frightened; I certainly was.” [The Autobiography of Fukuzawa Yukichi, 236-37]
During the Tokugawa period of the 1600s, when the civil wars had ended, the Shogun required all the great lords and their samurai to spend every other year in Edo. The samurai spent much of their time in sword-fighting academies, where the graceful movements of stylized exercises and mock duels were practiced. It kept the samurai ethos alive, but in fact it was almost all for show. There was very little real fighting, while carrying swords and displaying the elegant etiquette of the sword schools became a key part of the samurai code. Japanese sword-fighting turned into the equivalent of boxing gyms, which in England and the United States became popular for men at just the time when modern law-and-order was eliminating duels and most real fighting.
The bandit’s idealized sword-fight
The bandit’s account portrays the usual mythology of fighting. It is all very honorable; the bandit cuts the samurai free so that they can have a fair fight, just man-on-man, to decide who will get the woman. Both look like they were trained in a sword-fighting school, making all the proper feints and maneuvers. The bandit is bragging, showing off afterwards to the police and declaring that he expects to die sooner or later and he repents of nothing. His claim that they crossed swords 23 times is implausible-- not that it couldn’t have happened, but a real fighter in the midst of adrenaline rush would find most of it a blur. (This is typical of cops describing their experiences in a gun-fight, where time is distorted and they often are unaware of how many shots they fired.) In the bandit’s version of the Rashomon fight, both men perform their moves like a ballet or a sword-school exercise. This is the opposite of the woodcutter’s version, which shows the fighters sweating profusely. Their breathing is so heavy that it fills the sound track.
The other two versions
We can rule out the bandit’s version, and accept the truth of the woodcutter’s version, as far as the sword-fight goes. What about the other two stories-- could they be true? There is no fighting in either of these. In the woman’s version, she cuts her husband loose, but he refuses to fight for her. In the dead man’s version, he also refuses to fight, refuses to let the bandit kill his wife (even though in his view she has betrayed him), and kills himself. Both accounts are self-serving. The woman says she killed her husband while she was blacked out. This is plausible; losing conscious awareness can happen during extreme states of adrenaline rush. After she runs away, she tries to kill herself by drowning, but fails. This too is plausible, since the majority of suicide attempts fail, and women’s attempts fail more frequently than men’s.
If we leave aside the sword-fights, comparing the four stories one after another turns up something unexpected: the woman is the central character driving the overall plot. And she becomes increasingly dominant from one version to the next.
In the first version, she resists the rapist ineffectively: she lunges with her dagger, he dodges, he overpowers her. Then she turns on sexually: the camera shows her hands clutching the bandit’s back as the rape proceeds. (This is the bandit’s story, and it sounds like rape mythology, that a woman enjoys it.) Afterwards, she demands that the two men fight over her, and they comply.
In the second version, she takes the initiative to free her husband; but when he refuses to fight for a dishonored woman, she goes into a fury, demanding that he kill her to relieve her shame. When he refuses that too, she flips out of ordinary consciousness, and stabs him.
In the third version, she switches tactics strategically (as her husband sees it). After the rape, she implores the bandit to take her with him; he agrees. Then she implores him to kill her husband; this makes the bandit angry. He offers to let the husband decide whether he should kill her or not. Her husband is now reduced to complete passivity, and the woman successfully escapes the bandit.
From scene to scene, she becomes more central; in the fourth version, she dominates most of the action. After the rape, the bandit is won over by her, and begs her to marry him, even promising to give up crime for her. She’s not letting anyone tell her what to do; she breaks away with her dagger and frees her husband. But her husband takes the same line as the second and third versions, refusing to fight for a worthless woman: “You’ve been with two men. Why don’t you kill yourself?” In version two, she asked him to kill her, but now she switches tactics: she rushes to the bandit, calling on him to wait, crying she is only a helpless woman. “Stop crying. It’s not going to work any more,” her husband says. This gets the bandit to take her side: “Stop bullying her.” She laughs angrily, “If you’re my husband, why don’t you kill this man?” Annoyed at his cowardice, she turns to the bandit: “I was sick of this tiresome daily farce. I thought you could save me. But now I see you’re as petty as my husband.” She laughs hysterically at both men, and they hesitatingly begin to fight.
As an actor would say, she takes over the scene. A micro-sociologist would say she achieves emotional domination, forcing the men to do even what they don’t want to do.
Dramatic sequence or alternative realities?
Does this help us decide which scenarios are more truthful than others? Unfortunately not, except in the all-important point, that the fight scene in the first version is untrue, and the last version is typical of real fights. Focusing instead on the woman, we see that she becomes increasingly dominant over the men, by her emotional tactics, from one version to the next. This implies that it is really the screen-writer and director-- i.e. Kurosawa-- who has developed the sequence in this way. He does it for dramatic considerations, in order to make the film build up towards a climax.
In effect, Kurosawa is running through a series of permutations on what can happen in a sexual triangle following a rape: who blames who, and who gets killed. Strictly speaking, there is no sequence; it could have been run in any order.* This would be maximally relativistic-- maximal Rashomon effect-- but it would not be as dramatic.
*In fact one could produce as many as 36 different versions of the film, differing only in the order of the 4 witnesses’ accounts. This relativistic device was used, 35 years later in Milorad Pavic’s novel, Dictionary of the Khazars, published in two versions (male and female) identical except for one key passage. More recently, the device was used to structure Orhan Pamuk’s My Name is Red.
The fourth version, as noted, is not in the original story Kurosawa used for his script. The original three episodes were not long enough for a full-length movie, so Kurosawa wrote a fourth episode, the woodcutter’s story. This introduces the realistic version of the sword fight, and it also gives the most complex psychology of the woman’s role.
The first three episodes were enough to establish the Rashomon effect-- multiple realities that are all equally plausible; and that is what the original writer (Ryunosuke Akutagawa) appears to have intended. But by adding a fourth episode, and making it into a banal, unheroic fight, Kurosawa shifted the emphasis: the fourth version ends up feeling more real than the others. In the concluding scene of the film, however, one of the listeners to the narrations at the Rashomon gate, declares that the woodcutter is lying too: he omitted to say that he was the one who took the pearl-inlaid dagger from the murder scene. Yes, this makes the woodcutter a liar, but only about that particular detail; what he saw and reported of the rape and its aftermath, including the incompetently-performed duel, is true.
Kurosawa clings to the Rashomon effect, although adding the realistic fourth version undermines his philosophical statement. His autobiography says that he intended a multiple-reality effect all the way through conceiving and making the film. It was a one-shot trial. None of his other major films use multiple realities.
Would a micro-sociological conclusion still be a great film?
Would the film be any better if it explicitly said the ignominious fourth version is the true one? Obviously not; the whole metaphysical Rashomon-effect would disappear, and it would turn into nothing but another mystery story solved at the end.
This raises a further question about the usefulness of micro-sociology in literary creativity. A thought experiment should convince us: omitting the woodcutter’s version would leave us feeling unsatisfied, even with the Rashomon-effect intact. Dramatically the film needs the fourth retelling in order to rise to the level of one of the great films.
This is the same conclusion reached in my analysis of the realistic violence in Camus' The Stranger; a micro-sociological insight is at the core of the plot, but the author can’t dwell on it, and has to stay on a philosophical level in order to keep up its serious message. The spoiler isn’t micro-sociology in general (most good fiction writers are good micro-sociological observers); it is the micro-sociology of violence in particular, the dirty secret of how ugly and disgusting people look in committing real violence. The aesthetic fact is, real violence is just too unpalatable to get much space in a narrative that people will want to view. The writer’s dilemma is this: nothing makes a plot more dramatic than violence; but the more realistically violence is depicted, the more it has to be covered over by aesthetic distractions.
Rashomon does a lot to soften the violence. The bandit, Toshiro Mifune, is a rapist and professional murderer; but he is made into something of an anti-hero. In part, by his good looks and handsome physique-- he comes across as low-class only because he is scruffy and badly groomed; and he shocks the Japanese stereotype by lolling around in undignified postures, grimacing and slapping at mosquitoes. This was Toshiro Mifune’s first major film, and he and Kurosawa rose to stardom on the same vehicle. Yes, he is the villain of the plot, but he is irresistible to watch on screen; ebullient and spontaneous, laughing boastfully and childishly, but also overcome by fits of puzzlement. He grows more human through the sequence of retellings, alternatively in love with the woman he has raped and sympathetic with the samurai he has humiliated. By the fourth episode, we feel he is not a bad guy through and through... and our search for the bad guy widens to everybody.
The woman is beautiful and delicate in classic Japanese style, but the film makes her more than a mere victim. As the film goes on, she becomes the scene-stealer, the psychological center of attention. Against these two, the samurai is the straight man in every sense of the term, with his limited range of facial expressions, few speaking lines, his prim look: his slicked-back hair contrasting with Mifune’s wild hairiness.
Without the fourth viewpoint, Rashomon would have been a near-great try at a great film. It still would have the beautiful cinematography, perhaps the very best of the black-and-white era, with its rhythmic camera movement synchronized with the thrusting tom-tom of the music, and its psychologically revealing close-ups. It would have missed greatness, though, because the visual rhythms, the music, the shifting emotions of the actors, and the mounting philosophical doubt surrounding the whole thing are so tightly interwoven. A three-act version could have been just as beautiful, but it would have missed its climax. The three scenarios are too idealized, each artificial in its own way; the fourth, realistic scenario was needed to shift the mood and tie everything to the real world.
The four-part Rashomon is a greater film than the straight Rashomon-effect of a three-part version. The fourth alternative, anchored in the micro-sociology of violence, undermines the easy relativism of the Rashomon-effect. That dose of aesthetic tension makes it one of the greatest movies ever made.
Tragedy, the most serious form of literature, and action-adventure, one of the most popular forms, both depend on episodes of violence; but they cannot show violence as it really is. What does this mean for the rules of creative success? Future posts will take this further.
Akira Kurosawa. 1982. Something Like an Autobiography.
Eiko Ikegami. 1995. The Taming of the Samurai.
Randall Collins. 2008. Violence: A Micro-sociological Theory.