Ozu's reactionary cinema:
suggestions about the socio-economic determination of signifying practices

by Marc Holthof

from Jump Cut, no. 18, August 1978, pp. 20-22
copyright Jump Cut: A Review of Contemporary Media, 1978, 2005

Karl Marx wrote that Asiatic society had no real history but was socio-economically a monolith in which there seemed to be little or no internal change.(1) One can perhaps argue that Japan in the past had no history in the Marxist sense. That, however, is no longer the case since Japan is now one of the most advanced capitalist nations in the world. In other words, it now has a history.

Japan was the only Asiatic country capable of internalizing imperialism. It did not succumb to an external power but developed its own kind of capitalism. The Tokugawa type of feudalism, which reigned in Japan, contained in itself the germs of mercantilism and then capitalism. But this capitalism bears residual marks of feudal structures. For example, the dependence of workers on the factory has its roots in the feudal system. Feudalism serves the dominant ideology by masking class struggle even today.

The work of Japanese director Ozu clearly reflects this conflict between feudal structures and newly developing capitalism. Some critics(2) have argued that Ozu's formal devices are potentially progressive because they deviate from the classic rules of traditional Western filmmaking (eyeline match, camera placing, the use of space, narrative, etc.).

It is my intention, by means of the simple Marxist base sketched above, to disprove that evaluation. My knowledge of Japanese culture and society is limited. However, I think it is possible on the basis of (however sketchy) a Marxist analysis of Japanese history to prove that Ozu's films are not progressive but actually reactionary. That they are so does not diminish their artistic quality or indeed my liking them. It is, however, something we should keep in mind about the (hidden) ideology of these films.

Obviously the decoding of Japanese films presents us with some problems: What do we call progressive or reactionary in a cultural tradition we know very little about?

I think it is an indication of the political implications of Ozu's work (within Japanese culture and for Western audiences) that young Japanese directors dismiss it as reactionary. That Ozu is a reactionary cineaste is a conclusion I want to base neither on a very extensive knowledge of Japanese cinema nor on conclusions about the social content of the films themselves. Instead, I will concentrate my investigation on the ideology of those deviations that will appear fairly obvious to someone schooled in Hollywood cinema.

Ozu's way of addressing images to the public is conservative and reactionary. Every instance of his filmic system is ideologically determined. This system transmits an ideology to the spectator. Every image we see forces us into a way of thinking that is socio-ideologically determined. Obviously we are far from aware of this as spectators: unconscious processes transmit ideology, and psychoanalysis plays an important part in discovering them.(3) As is the case with all unconscious processes, the ideological kernels that underpin any text can be made conscious through a kind of psychoanalysis—be it criticism, be it other processes of self-reflection on the part of the spectator (all the easier when he/she detects ideological views contrary to his/her own).

In A HEN IN THE WIND (1948), high food prices in postwar Japan force a woman to prostitute herself in order to pay for medicine for her sick child. When her husband returns from captivity, he discovers her infidelity. After a fight, she falls down the stairs of her house. However, later on the husband forgives her and they are reconciled.

In BANSHUN (LATE SPRING), Noriko, a young woman in her twenties lives with her father. She is still unmarried and actually disapproves of getting married. Her father worries about it. Noriko's aunt tells her that her father wants to remarry. The idea shocks Noriko. After a trip to Kyoto she decides to marry herself. Her father, who did not have the intention at all to remarry, remains behind.

Ozu made BANSHUN immediately after A HEN IN THE WIND, and both films have certain stylistic features in common. These include

  • the general low angle of the camera (Ozu's celebrated tatami-height);
  • the predominantly static camera (although both films contain tracking shots);
  • the straight 90-degree angles that present characters and interiors;
  • the complete lack of fades;
  • the use of object shots, unmotivated by the narrative and used as transition.

In the latter film, for example, object-shots of the gas factory nearby and the stairs of the house of the woman, where most of the action takes place, occur between the different sequences.

The films even commit some explicit faults against Hollywood film school grammar. The films disregard space continuity rules. They jump from one side of the room to the other and thus abruptly change the position of two characters sitting fact to face (enough to confuse a Western spectator). They disregard eyeline matches. For example, BANSHUN shows Noriko and her father both looking to the right in different shots (in a Hollywood film, on the other hand, they would be looking away from each other). They are, however, facing each other. The formal rigor of this style gives Ozu's cinema a dignified atmosphere that combines great formal construction with humanistic content.

A lot of Ozu's films are contemporary sentimental comedies or low-key melocomedies; in other words, they mix comedy with melodrama. More than with any other Japanese filmmaker I know, the (classic Japanese) family is the predominant scene of the action. To a Western eye it is a pleasure to be in those quiet serene surroundings (though it bores some people after a time). This happens to such an extent that it comes as a shock when Ozu shows us a modern house, as happens in BANSHUN to characterize Noriko's friend. I am not going to discuss the use of other scenery—though Ozu's use of it is significant. I only want to use this example as a starting point for an ideological interpretation.

The love which emanates from the static images of the rooms of the houses in the classic Japanese style speaks for itself, indicating Ozu's position in the well-known Japanese conflict: the modernization, the Westernization of the old Japanese traditions, which is in fact simply the subjection of a pre-capitalist culture and economy to the laws of capitalism. The brochure, "Japanese History through Cinema,"(4) says Ozu's main theme from the early thirties to the early sixties was "the fear that the age-old family traditions should be lost forever." And indeed the family is the main value that Ozu uses in his filmmaking. His films rest on the age-old and possibly cathartic narrative device of sequencing balance/ imbalance/ return to balance (which Vladimir Propp's analysis of Russian folktales revealed). The family is the center for the film's "action"—which seems very subdued in comparison to the noisy modern and Western counterparts.

Bearing this in mind, it is quite easy to make ideological readings of the films. It is possible to explain the particular "deviations" in Ozu's filmmaking and to give them meaning in an ideological framework, even if on the surface level they appear to be "meaningless" formal devices.(5) The stability of the family and the house, which is the main signifier for the family, mean nothing less than the stability of the traditional signifying system. The apparent contradiction then is that Ozu works in a new medium (not existent in the traditional Japanese sign system) and has to work out new devices correspondent to the old ways of signifying. These old ways do, however, sometimes survive, as they do in the quite unfilmic (to Western eyes) rendering in more than one film of Noh or Kabuki plays, or in the filming of the Zen gardens in Kyoto (which play a similar nostalgic role in Mishima's novel Forbidden Colors). Such a decor in BANSHUN actually brings father and daughter together again, under the traditional values.

Even if the destruction of traditional Japanese values by capitalism may imply that we should reverse the traditional meaning of the word "progressive," we should keep in mind that it is a profoundly feudal system that progress destroys. Marxism certainly does not suggest that we should go back to pre-capitalist historical situations but to post-capitalist ones. Moreover, the family situations present in Ozu's cinema do not reflect a pre-capitalistic romantic paradise. On the contrary, the repressive dishonest moral of A HEN IN THE WIND is clearly recognizable as reactionary to Western audiences. Here, the female "sinner" forced by modernism, i.e., capitalism, into prostitution returns to the dominating male who after some time finally "forgives" her.

Without being exhaustive but merely by making a few suggestions, the application of which would vary from film to film, I want to explore the different stylistic devices Ozu uses. Especially I want to deal with what seems the crucial one: the use of "the look" in these films, from which it seems to me that we can derive the other deviations. In order to do that it is necessary first of all to describe Hollywood's norm, against which Ozu's cinema stands out.

Hollywood cinema uses narrative techniques that seem peculiar in comparison to Ozu's system. If we compare the Hollywood paradigm with Ozu's use of space, we notice the following points:

1. Hollywood uses space only as a background for the characters. Not only are there almost no shots without characters, but when characterless shots do occur, the narrative recuperates them as place indicators (for example, establishing shots), lyrical intermezzos dependent on what has just happened in the narrative, a clue in a detective story, etc.. Hollywood cinema only permits shots with characters or actions, with which the characters are involved. In other words, Hollywood cinema forces us to identify with characters.

2. In order not to confuse the spectator, Hollywood cinema has developed a fictionalized space in which the characters evolve. This leads to rules like continuity space editing: the laying out of an imaginary space behind the characters. Focusing on the characters while keeping the background out of focus also serves the same purpose.

3. 180-degree and 30-degree rules are necessary to prevent disorientation of the spectator: to keep him/her believing in the false space.

4. Hollywood also uses the moving camera—real movement or movement through a succession of shots—to establish such spatial illusion.

5. The reverse shot is another typical instance of Hollywood's use of space: the spectator travels, jumps really, like an invisible voyeur through the rooms around the characters. With the reverse shot, it is possible to establish that where the viewer should be, there is in fact more fiction. The reverse shot forces the spectator out of the fixed position he or she has in the theater and makes him/her part of the fiction.

6. The same kind of fictionalization or of incorporating everything into the narrative is at work in the stock dialogue scenes of classical Hollywood cinema. Traditionally such scenes work along a diagonal axis. This is a characteristic technique of Western cinema, a consequence of the "don't look in the camera" rule—which means, "Don't break the illusion." But not must the actor keep his or her eyes away from the camera lens, the actor also has to play past the camera (while playing for it). The point is to make believe that the camera (= the spectator) isn't there, without forgetting of course that the only reason the actors perform is that there is a camera, a spectator, an audience.

7. Over-the-shoulder shots and eyeline-match cutting assure that the actors look at each other. Through the exchange of looks they preserve the apparently closed world of the fiction film. What we seldom find in the classic Hollywood paradigm is a strictly horizontal or vertical setup that presents the characters in profile. Or as Godard does so brilliantly in VIVRE SA VIE, one can align the characters with the camera behind, forcing the camera to move sideways, back and forth. Brian de Palma in OBSESSION moves his camera between Cliff Robertson and John Lithgow more or less on the rhythm of the passersby behind the window of the Florentine Hotel. But still characteristic of commercial cinema, de Palma moves his camera and does not keep the two actors framed together: he still plays on the alternation of identification and the look.

The spectator cannot identify with the two characters at the same time. Moreover, if a film uses a horizontal setup, the game no longer grips the spectator. In the traditional dialogue scene the camera has to change position to see what the other person will say, what his/her expression will be. In other words, the film acts as if the spectators have a role in it, a certain activity, as if they were an invisible actor themselves. If the scene places the characters horizontally, it excludes the spectator, who can only look on whatever "private" game the actors play (hence it has a potential use as a distanciation device in modern films as used by Antonioni and Kubrick, etc.)

8. A last element of Hollywood filming is the fade out or dissolve separating the fictional scenes. The fade out permits a smooth ellipsis of elements unnecessary to the fiction, but at the same time it suggests that the narrative has left out some "reality."

All of these elements have relevance to the theory of the "look" as recent film criticism as developed such a theory.(6) Indeed, what is the purpose of these different techniques but to keep spectators from realizing that they are looking at a film, not real people? In other words, these cinematic elements function to keep the spectator in the narrative fiction.

Paul Willemen's article on Stephen Dwoskin nicely sums up the three kinds of "look" at work in the cinema: 1. the camera recording the profilmic event; 2. the spectator looking at the screen; 3. the looks the characters exchange in the fiction. It is clear that the use of space, diagonal setups, continuity space editing, "Don't look in the camera" rules, eyeline match, and even fadeouts and dissolves to separate pieces of fiction derive from the exclusive attention of Hollywood cinema to the looks characters exchange in the fiction. In order to keep us in a fictionalized world, the film requires us to restrict our attention to the fiction on the screen, accepting it by a temporary suspension of disbelief (we know that it is a film after all).

Voyeuristic or identification patterns (a form of narcissism) develop through these techniques, governing the desire that keeps the spectator in the narrative, and the audience in the cinema. A complex interplay of voyeurism and identification keeps the narrative going: otherwise we fall out of the fiction. Hence the need for characters to identify with, the need for a recognizable space, the need for clear transitions. Voyeurism provides not only the pretty women that are supposed to interest us in a patriarchal society but also slick surroundings, glorious color or widescreen techniques (or just nice black and white photography), music, etc.

The play of lack and fulfillment which qualifies narrative, the tempting use of voyeurism, in other words, the exploitation of the desire mechanism of the spectator forms the base of all entertainment,(7) and hence of the commodity value, the salability of the film. By not letting the spectators feel that they are looking at a film, but by keeping them in the narrative fiction, a film keeps the spectators in the movie theater. The better the entertainment, the better the film's seductive powers. In other words, the better equipped the narrative, the stars, the color, the screen ratio, the sets and the craftsmanship are to handle the spectator's identification and voyeuristic mechanisms (including the use of space and the other elements outlined above), the more dollars come in (providing there is good advertising, etc.).

The whole mechanism, the entertainment system and the particular use of narrative with its concrete rules, are capitalistically determined. Hollywood cinema makes money, creates surplus value with these structures of cinematic fiction. It is, however, a misconception to state that a cinema which does not capitalize on the characters' looks but makes the camera's look or the spectator's looks visible (in other words, which makes the material mechanisms of cinema visible) is automatically a progressive or Marxist-oriented cinema. One has to make the materialist foundations of society visible. Ozu's case, in which his films diverge in terms of all eight narrative elements from Hollywood cinema, provides an example that clarifies this important political distinction.

In terms of the formal qualities of narrative (space in function of character, diagonal setups of scenes, rules of fictionalized space, use of fade outs, etc.), Ozu's cinema is usually divergent. But, as I have said, these elements are characteristic of a capitalistic signifying practice, designed to provide the pleasures of identification and voyeurism by means of films structured and produced to make a profit. So too does Ozu's cinema, but his audience lives in a capitalist society which has a large residue of feudal structures.(8) And capitalism uses these to support itself, to preserve the existent production relations. The signifying system Ozu works in (insofar as it is feudal, or, to be more precise, insofar as it presents itself as being feudal for well defined capitalistic reasons) does not spring from profit making, psychological "entertainment" processes but derives from a more religious concept of the sign, the best description of which Roland Barthes gives in L'Empire des signes.

Here, as in the theater, a horizontal positioning towards the spectator is dominant. Hence each gesture receives, as in theater, mythical resonations. The film can place the spectator outside of the action (and he/she can be partly conscious of being in such a position). Instead of being caught up in the fiction through the interplay of voyeurism and identification, the spectator is not part of it but contemplates the action, the sign that is before him/her. He/she does not follow the fiction on its inevitable drive/ desire through identification and voyeurism to the ultimate gratification, the end" (in which positively or negatively all wishes reach fulfillment, all loose ends connected, all needs or lacks filled or definitively prevented), but the fiction centers more directly on the immediate consumption of fixed sign appearances.

Most of Ozu's "deviations" are still based in Japanese signifying practice. They certainly differ from Hollywood film conventions but they are also embedded in a system set up to produce profit, just one at a different stage of capitalist development. Since cinema did not exist in the feudal or semi-feudal eras of the Western world, the distinction here between pure capitalist or capitalist mixed with feudal ideological systems may be more important as a critical tool than the usual East/ West dualism by means of which critics evaluate Japanese cinema and Ozu in particular. Signifying practices in European feudalism were quite different from capitalist ones and, in fact, were probably closer to what we now call Eastern culture. Ozu's originality lies in the way he has developed a film technique which combines an understanding of a Western (capitalistic) medium with an appeal to a semi-feudal audience (or one nostalgic for feudalism).

Without giving detailed discussions of each of the deviations (which the Screen articles amply provide), it is quite clear that Ozu's address system constantly uses the tatami height of the camera to accentuate the quiet semi-detachment and contemplation that is the filmmaker's trademark. For the same reason, he uses many low angle shots and profile setups, especially in the dialogue. Ozu's actors do not usually look in the camera either, and instead of the external exchange of looks in Western cinema, Ozu's camera presents the actors as "objects" for contemplation. This explains the frequency of "profile" situations. This theatrical, horizontal positioning brings into play the second look—the public's looking at the screen—but it does so in a mythologized way. The public does not materially realize its situation, that is, think constantly that this is a fiction on the screen. Rather the film fictionalizes the public itself. The public's look becomes part of the fictionalized world. This fictionalizing partakes of a particular kind of idealism that is typical of feudal and religious signifying systems.

Ozu's use of "inserts" very clearly demonstrates this idealizing. In Ozu's cinema, inserted stills of objects function more on a symbolic level than they do as a component of the narrative to more the story forward. These inserts play an important part in Ozu's narrative technique (see Edward Branigan's analysis of Ozu's inserts in a Screen article) and let him avoid fades, which give way to hierarchical transitions through the use of this kind of object shot.

The symbolic use of inserts places the spectator in a more direct (but not less imaginary, idealistic, or anti-materialistic) connection with the sign. Or to refine that. Instead of the dynamic (Western capitalistic) desire (awakened and sustained throughout the narrative by an interpolation of lacks and partial fulfillments inherent in the entertainment system), Ozu's cinema condenses desire to single moments. Finally, Ozu's cinema itself is a hybrid economic and signifying form also belonging to the entertainment system, a capitalist one.

This static enjoyment of the sign as a totality (Barthes cites some more examples of this as it appears in Japanese culture) seems typical of a feudal signifying system. According to Marx, feudalism as a socio-economic system depends upon personal dependence, hierarchy, the place of God on top of that hierarchy as the ultimate signifier, and a religion utilizing fetishes. Such a system reflects the kind of economic realities under which Ozu works: a capitalist system which still recruits its customers from feudal society and which uses the feudal ideology as support for the emerging capitalist economy.

A HEN IN THE WIND organizes space in a number of straight angle shots. These shots hardly overlap, if at all, in the space they show. But since most of the action takes place in one house the spectator can organize its space after a while. For transition shots, Ozu repeats a shot of the wooden stairs (or rather shots, since the light varies according to the hour of the day in the fiction). These seemingly unaccounted for shots get symbolic and narrative resonances, however, when the woman falls down these very stairs after the brawl with her husband. The stairs serve for the moment of catharsis; they are the place which has put its traces over the film from its beginning. The previous shots of the stairs get an a posteriori justification. This does not alter the attitude the viewer adopts towards those shots. At the film's end, the spectator contemplates this "fatal" location, brimming as it now is, after the spectator has seen it so many times, with "symbolism." Though this kind of device can also occur in Hollywood cinema, a "normal" narrative (in other words, Hollywood style) makes these stairs unimportant in order to highlight the action itself. The Hollywood film invites not contemplation of the "locus" but identification with the victim.

Tatami-height of the camera, low and profile angles, object shots, sequences organized in parallel fashion, sequences deviant from the eyeline-match rule, lack of fades, etc. —  all derive from one basic mechanism. They serve the fictionalized inscription of the viewer's look in the fiction itself and the utilization of screen objects as observable objects—not in a voyeuristic/ identification sense nor in a truly material sense (which would make visible all three looks that are present in cinema) but in a contemplative attitude.

A film like LUDWIG, REQUIEM FOR A VIRGIN KING by the young German filmmaker H.J. Syberberg with its theatrical camera-viewpoint (static scenes played before projected backgrounds), shows that the kind of alienation which contemplation provides is not sufficient, but in fact it keeps everything on a mythical level. This "objective" setup of the spectator towards the narrative space does not really reveal the spectator's look (even less the camera's look). It merely disentangles objects from the look of the characters. Hence the voyeurism of the spectator does not lessen but increases (replacing identification processes). Now, however, this voyeurism emphasizes objects, which produces a kind of mythical fetishistic pleasure.

The prime object of this fetishism/voyeurism is not really those objects which Ozu inserts into his narrative, but the stability of the main signifier for which they stand: the family. At the center of each scene are family relations, often given concrete form in dialogue scenes, the whole situated in the framework of the traditional Japanese house.

In Ozu's cinema (the setting in motion of desire) arises from a threatening of the fixed, closed family signifier. That threat impels the motion of the film to its end, where the structure is closed again. In this way Ozu reconciles static (Japanese) and dynamic (cinematic) signifying practices. The threatening of the central signifier occurs through Western, capitalistic signifiers unbalancing the family. For example, the post-war situation (high food prices) forces the women into a case of prostitution in A HEN IN THE WIND. BANSHUN deals with the conflict in Noriko between the family (her father) and traditional values, and her friend and more Western values—the whole is actualized in a marriage/non-marriage dilemma.

Ozu's cinema is thus a hybrid. It echoes the mixture of feudal ideology and capitalist economy in which Ozu produces his films. Ozu's cinema chronicles the transformation of a static contemplation of the family signifier into a more Western, more capitalistic signifier. By exploitation of this conflict for profit, Ozu further entrenches himself in a capitalist economy. But even without that capitalist dimension, Ozu's nostalgia is reactionary. His cinema is nothing more than nostalgia for a feudal signifying system and a feudal structure (the family).


1. See Shlomo Avineri, "Introduction." Karl Marx on Colonialism and Modernization (New York: Anchor Books, 1969).

2. Kristin Thompson and David Bordwell, "Space and Narrative in the Films of Ozu," and Edward Branigan, "The Space of Equinox Flower," both in Screen, 17, No. 2 (Summer), 1976.

3. I use the term psychoanalysis (and not psychology) in reference to recent film criticism (Screen, Communications) that deals with unconscious processes. The function of psychoanalysis in understanding ideology is obviously a difficult question, especially since the theory worked out mainly around the French magazine Tel Quel is not so much Marxist as a peculiar brand of materialist anarchism. A progressive text for these critics breaks with the traditions of capitalism, religion, and patriarchy. Tel Quel reacts more against pre-capitalist ideologies (used to support capitalism) than against capitalism itself. The group's limiting itself to a consideration of superstructures, its reading of even socio-economic realities as "texts," seriously constricts the social relevance of its findings. One does not make a revolution with Mallarmé, even less with the texts of Sollers, as Tel Quel seems to advocate. It seems essential to combine liberation from unconscious constraints with socio-economic liberation. Such constraints are those unconscious processes that place the subject (the spectator) in prefabricated moulds and thus provide a structure for transmitting ideology. These mental processes are not separable from economic processes, which change according to the historical moment and are characteristic for certain kinds of societies.

4. "Japan: History through Cinema," Japan Film Library Council, Tokyo

5. Although Ben Brewster gives a good ideological reading of SHOSHUN in Monthly Film Bulletin, August, 1976, he notes the film does not seem to offer very useful material for the demonstration of how ideology was imposed. I think it more valuable to offer an ideological reading of the so-called "modernist" deviations. The Screen position on Ozu (Foreword, Summer and Autumn issues, 1976, and Stephen Heath's elaborate article, Autumn, 1976) contests the degree of deviation and questions whether or not Ozu's cinematic structures differ much from Hollywood's. As I demonstrate, the differences are regressive rather than modernist. Thus Ozu does not seem to be the right starting point for those seeking to found an emerging progressive cinema culture.

6. Paul Willemen, "Voyeurism, the Look, and Dwoskin," Afterimage, No. 6. The article contains a valuable assessment of Laura Mulvey's important "Visual Pleasure and Narrative Cinema," Screen, 6, No. 3 (Autumn) 1975.

7. Richard Dyer discusses this term central to Hollywood film practice in "Entertainment and Utopia," British Film Institute. See also his "Light Entertainment," BFI Television Monograph.

8. I borrow the terms residual/ dominant/ emerging culture from Raymond Williams, "Base and Superstructure in Marxist Cultural Theory," New Left Review, No. 82 (November-December), 1973. Marx saw history as a succession of socioeconomic stages: (European) feudalism changes itself through internal evolution (the birth of the bourgeois class) into capitalism via intermediary stages like mercantilism. This does not happen at once, and parts of the socio-economic system, geographically and historically, can still be in an earlier stage of development. It is, however, a characteristic of capitalism that it seeks to expand itself ceaselessly, through the search for new surplus values. Ernst Mandel discusses the relation between pre-capitalist ideologies and capitalism that absorbs them in Spätkapitalismus (Late Capitalism), Frankfurt am Main, 1972, as does Avineri. Seeing a similar evolution for culture, Williams distinguishes between residual, dominant, and emerging cultures or ideologies. Ideology seems to follow the same pattern as economic history; pre-capitalist ideologies fall subject to capitalist laws (for example, the mass media absorb art, or Japanese classical culture gets cinematic form, as in Ozu's case). The dominant ideology and economy absorb residual ideologies and use them for their own purpose.