Pricked by Poetry: The Films of Hiroshi Shimizu

In this mammoth appraisal of the films of Hiroshi Shimizu, David Phelps digs into films from across his vast body of work, stretching from Shimizu’s silent days through the war years and finally to his challenging, largely unloved work of the 1950s.

Outskirts Online / Get the Print Issue / Back to Homepage

By David Phelps

The Masseurs and a Woman (1938)

Many months after watching, there are moments from The Masseurs and a Woman (Anma to onna, 1938) that I cannot shake even from my own forever-tottering mind. They are found in the climax of a film that has, for the past hour, wriggled its way out of any kind of rising action, as we’ve followed a few blind masseurs in a resort town as they compete to outwalk the tourists and hear of some shadowy figure, somewhere in the background of what feels like another film, robbing the visitors. Like Frederick Wiseman in Blind (1987), Shimizu films the masseurs from a series of characteristic reverse dollies that prevent our seeing what’s in front of them any more than they can. We, too, have barely seen the actual drama of this film.

And then, in three of the most beautiful sequences of any film, we are confronted with the unspoken story the film has refused to articulate – but this, too, elliptically. A woman with a fan watches the droplets of rain on a river; a blind masseur tells her, desperately, “even though I am blind, I have been watching you,” collapsing to the ground when he realizes she is not the thief he was hoping to save; nobody to save him, no need to save her; and the camera, in one of the rare forward tracking shots of Shimizu’s films, chases a carriage as it carries the woman away.

I don’t think there are words for the power of these shots. The drama is, strangely, that the characters themselves are denied the kind of cathartic drama any other film of the era would have given them: the robber is never identified, never seen, and in place of salvation, the rain falls on the river, the parasol twirls, and everyone goes home. There is a drama that might have tied their fates and changed their lives. But that drama is not Shimizu’s interest. The drama is that everyone must leave this film in solitude because that drama was nowhere to be found.

II.

Children of the Beehive: What Happened Next (1951)

What I mean is that much of the drama of the mundane is its refusal to give us the drama we need: the drama that would let us transcend as heroes or villains, to find salvation in death or love. Indeed, all of these familiar storylines are almost – almost – completely absent Shimizu’s films, short of recurring boss and bully figures who serve as forcing-functions for our protagonists to form makeshift worlds of private rituals on the margins of polite society.

In other words, much of the drama of the mundane is that characters have to create it for themselves. Part of the immense power of Shimizu’s films, I think, is how often they suggest plotlines of romances and criminal intrigues that none of the characters dare consummate, preferring, like Shimizu, to escape the trappings of melodramas back home to try, for a moment of respite, to play small games with each other instead. It’s the fact that we have to imagine these plotlines, which both the film and characters are happy to tease without ever bringing themselves to enact, that makes the film so much more moving, so much more devastating, than if it had fulfilled its humming, subsurface promises.

In place of the usual genre drama of classical cinema, Shimizu’s films are filled with these moments of bored characters inventing competitions to gamify – to drama-ify – the world around them: the walking contests in The Masseurs and a Woman and funny-walks in Tokyo Profile (Tokai no Yokogao, 1953), the waking boys betting which of the elderly snoozers will snore more in Ornamental Hairpin (Kanzashi, 1941), the boys in Children in the Wind (Kaze no naka no kodomo, 1937) staging a nationally-broadcast swimming match as one freecrawls across their living room rug and the other offers sports commentary on a chair.

Shimizu is famous, if at all, for his many films of children, but I suspect that’s partly because children have always been the most combustible vantage point for interrogating the mundane – for interrogating the ways humans learn and question social rituals (note the sheer number of Shimizu films about schools’ indoctrination), construct and deconstruct drama as simultaneous directors and actors each time they play (note the sheer number of Shimizu films about kids hanging out with nothing to do), and come to embody and resist state mandates (note the sheer number of Shimizu films about boys facing the humiliation and liberation of nightly bedwetting).

In all these ways, children are ambiguous figures, but not least because they let us see the mundane as both the soul-crushing emptiness of a post-industrial era where humans are still expected to behave as cogs as well as a kind of emancipatory space to reconstruct our world in play. And this notion helps us challenge the popular conception of neorealism as a movement born of bombs, as though the industrial genocides of the 1940s had exploded all the innocent, swashbuckling fantasies of another era. It helps us remember that in fact, neorealism was not only a far more morally complicated movement, but one that started with those supposed harbingers of innocence, the children of Manoel de Oliveira’s Aniki-Bóbó (1942) and Shimizu’s Four Seasons of Children (Kodomo no shiki, 1939)—and, fine, Children in the Wind.

The ambiguity of what children represent can help us understand why, across these three films, children more or less invented neorealism – not because they’re incapable of acting, but because they’re incapable of not acting. Four Seasons of Children in particular is an encyclopedia of universal childhood gestures: running too fast one way, stopping too fast in the middle of running too fast one way, carrying each other on their backs, fiddling with objects as dad spirals into a self-centered soliloquy, climbing bridges and trees and anything in sight, fake-crying, real-crying, imitating adult niceties (clapping) with zero comprehension of their reason or purpose, forgetting anything important because there’s an animal to play with instead. It helps us remember that neorealism, like Shimizu’s films generally, is the art of documentary performance, of bodies getting to play, to play themselves.

The adults’ world, meanwhile, rhymes with the children’s as a series of intrigues and games, but with one difference: the adults are far too serious to know the game is a game, that the roles they play are just that. There’s a simple reason for that, of course: the adults, unlike the children, have to think about money. Money plagues them, empties them, divides them.

Shimizu’s films are obsessed with what it might mean to carve true solidarity across the intractable class divides of a profit-seeking world. His answer – children – is a fairly bleak one when you remember what they must grow into. But it is also a measure of hope, registered in the cyclicality of seasons and Shimizu’s own fairly radial plot construction, so grounded in documentary that it can feel like an epiphany.

III.

Children in the Wind (1937) / Acorns (1941) / Record of a Woman Doctor (1941)

It’s understandable that Shimizu’s most beloved films are his bucolic idylls of life in the Izu Peninsula documenting the ceremonies of everyday life for transients – tourists and children – who learn the ways of the locals as they pass through space and time. But that also means that critics have struggled to try to place the other half of Shimizu’s entire filmography, potboilers and programmers which follow none of these conventions, in any kind of coherent framework. The bucolic idylls feel gentle, light, and comic. Surely these others are, by contrast, tragedies?

The problem is that comedy and tragedy always intermix in Shimizu’s films, often rugpulling the genre out from under you in the final moments. A more precise designation, I think, is that for at least the first half of his career, Shimizu largely alternated between two different types of films: open and closed. The Open Films indulge the foibles of its playful protagonists as they tease each other across yawning fields and rolling rivulets. It’s in these Open Films that the characters often seem to be play-acting their way to a new kind of realism. But even Shimizu’s camerawork feels different in these films.

A tracking shot in a Mizoguchi film is an operatic thing, sweeping our characters into all the bristling bustle of high melodrama. A tracking shot in Shimizu – well, a tracking shot in Shimizu is far more of a bumbling affair, working overtime to catch up to its protagonists, to get the action fully in view, to delineate a world that pours out to the depths and hidden crevices of the frame and then beyond. And this is why Ozu and Mizoguchi have commanded so much scholarly love while the far more documentary, fly-buzzing-off-the-wall, find-things-burbling-forth-and-film-them approach of Shimizu has gone neglected for so long. Mizoguchi and Ozu use every cinematic device to cocreate closed worlds that they seem to be assembling piece-by-piece with every shot and frame. Shimizu’s worlds are quite literally open formally. And that makes them messier things.

At some level, his great Open Films seem to operate like attempts to see how much of the world they can contain in a torn-off newspaper fragment of people in a place. The late 30s works in particular represent radical experiments in duration and temporality: Mr Thank You (Arigato-san, 1936) functions almost like a real-time bus tour of Japan while the 2.5 hour Four Seasons of Children, despite being one of a half-dozen films Shimizu shot in 1939, feels to have been filmed across all four seasons, Bruegel-like, to document kids at play as their relationships shift through the adult drama of a year.

Neither time nor space nor bodies can be orchestrated or controlled in these films, but can only be surrendered to in slabs of space-time, glimpses of lives lived well beyond the physical and temporal confines of the shot. It’s a wholly familiar post-classical view, familiar from the school of Kiarostami, Hou, Yang, and Somai (a major Shimizu fan), all of whom also innovated late-century neorealism through films about children. But this was 40 years too early, shaped by a filmmaker whose hands trained on the gestures and melodramatic plot devices of silent film. And it means there’s nothing else in cinema quite like Shimizu’s films – not least because of the way duration distends melodrama into something it’s never really been: relatable.

For example, if I told you the plot of Children in the Wind – two boys losing their friends after their dad is falsely accused of embezzlement – I’m sure you would immediately understand this story at its desolate core. And yet, on the surface, Shimizu does not seem to understand it himself. Instead, he tells this tale of abject solitude, rejection, and false accusations in a universe of catatonic adults slunk somewhere in the corner of the screen – Shimizu tells this story entirely through long scenes of the boys playing, so that they take the place of the filmmaker in staging scenes purely for their own pleasure.

And here we start to understand why, formally, it took children to reshape cinema: because the height of realism is pure performance. The easiest way to make docufiction, to tease the boundary of fact and fiction, is to make a movie about kids. Kids’ play itself becomes a kind of stand-in for Shimizu’s own direction as the children stage their own mise-en-scène with household tools. Emerging organically out of the characters’ own boredom, their actions overtake the drama of the movie itself: they represent both the escapism of genre film as well as the impossibility of escaping from it at all.

Letting the kids take over the movie is, in some ways, an acknowledgement that Shimizu’s game is the same as theirs, papering pain with play to trick ourselves that what we’re watching is fun, is joyous, is not what it actually is: some errant act of truancy, two kids-in-all-senses acting out in the shadow of a familial fall. They are reimagining themselves in anger and isolation from social disgrace, of course. And so at some point they run away from home to catch a falling star and make it as far as the end of the block on their dusty, empty street. Their response to being trapped, to having nowhere to go, is, in some ways, to make a movie.

IV.

Forget Love for Now (1937)

The Closed Films are not exactly loved: they’re narratively programmatic, formally boxed-in, works of creaky machinery that subject the characters to a fatalist plot and mise-en-scène that can feel especially frustrating from a filmmaker who would, like his characters, break these in his serendipitous outdoor films. Above all, though, I suspect these are challenging films because they’re so slow; characters mostly seem to be sitting and staring in the sounds of the film reel itself. But that’s also a reminder, then, that Shimizu wasn’t just a godfather of neorealism but of durational cinema more broadly.

In the definitive Closed films – entries like A Woman Crying in Spring (Nakinureta haru no onna yo, 1933), Forget Love for Now (Koi mo wasurete, 1937), and Notes of an Itinerant Performer (Utajo oboegaki, 1941) – the self-sacrificing heroines feel trapped by Shimizu’s suddenly boxed-in interiors to perform socially-mandated roles. The play-acting is no longer emancipatory but fatalistic, as if their sense of self is little more than a record of its own oblation.

In all three films, a poor woman is trapped in a building of the wealthy to whom she must devote her life and body. In A Woman Crying in Spring, it is that quintessentially 1930s setting, the inn: the ground zero for a decade without heroes, with only straggling masses plodding from bed to table, where an ensemble enacts a fully three-dimensional theatre through a proxy audience watching downstairs as intrigues unfurl in the bedrooms above.

And yet this, too, is an outdoor film, shot on location in Hokkaido. At every moment the snow provides a kind of visual musical accompaniment as it flutters, cries, crashes, and entombs the scenes in dread. Every fight scene is as unnecessarily, beautifully artsy as William A. Wellman’s Track of the Cat (1954), an abstraction of half-seen bodies dancing in shadow against a white blanket of ice. The formal plays of early sound are often heartbreaking: watching one character’s face as another speaks, decades before Éric Rohmer, or a mother calling her lost child’s name against a cavalcade of abandoned spaces where the child is nowhere to be found. And yet it is the gestures that do the emoting: a character pausing at the top of the stairs and lowering her head before descending, or gathering her laundry off the line absent-mindedly at a moment of crisis.

Likewise, Forget Love For Now registers as a kind of formal masterclass in how to ground extravagant melodrama in the banal routine of everyday life (echoes of Jeanne Dielman). The film seems to circle the same four sets, repeating sequences in subtle permutations as the characters return to them day by day. As usual with Shimizu, the bleak world of the adults is juxtaposed with the bleaker one of the children as they each deal with the bullying of a boss, the empty camaraderie of their peers, and the embrace of foreigners (dangerously for the mother solicited by rich tourists, fortuitously for the son welcomed by poor immigrants).

Here, Shimizu’s interest in the relationship between temporal duration and spatial distance reaches its apotheosis in an unending series of fixed deep-focus establishing shots that let the action emerge from the background until it erupts into the foreground. But these are broken, at pivotal moments, by a few sparing close-ups in the peak of melodrama, and a few of the most beautiful tracking shots I have ever seen in a movie: the camera gliding across an abandoned dance floor towards our protagonist as she stands immobile and askance; the camera then switching 180 degrees to track her as she sashays onto the empty floor; the camera moving into an antagonist as he comes at the camera to reveal us in the middle of a fist-fight. It’s as though Shimizu wants to respond to the movements of his characters with stillness, and to respond to their stillness with his movement, but even that is not quite right. The bigger point is the feeling these gestures evoke in an otherwise icy film; that even at his coldest, Shimizu and his camera literally can’t be unmoved.

Still, the Closed Films are formal tests of all the usual platitudes about what “A Shimizu Film” actually is, and no test is greater than Notes of an Itinerant Performer. All in one go and for one of the only times in his filmography, Shimizu tackles a period piece, a tale of the ultra-wealthy, a romance, a political allegory, and maybe most notably, a theatrical picaresque in which the histrionic acting by stage legend Yaeko Mizutani seems like the film itself to be utterly out-of-time from Shimizu’s usual naturalism.

Here, the planes of the interiors isolate the characters from each other as they attempt again and again to reconnect across the space of the frame while only showcasing their own theatricality in the proscenium-like spaces of Shimizu’s homes. (If there is one film to understand all the possibilities of deep focus, I’m not sure it’s Citizen Kane; I think it might be this.) But these shots, constantly rupturing the spatial layers Shimizu seems to be opposing, also clue us into what Shimizu is doing. He is making the ultimate Closed film, but he is also making the ultimate anti-Closed film.

Its plot, which true to the title seems to be made up scene-by-scene based on our heroine’s latest proclivities, concerns a working-class woman who wants to leave her measly theatre troupe for a position in a prestigious household, where she gradually takes over the affairs of the business in the absence of a patriarch, presumably standing in for all the ‘40s men who had gone to war, and then decides that she wants to leave that too because it’s too “stuffy,” with no room to drink or smoke. It is, in other words, too Closed, just as the theatre itself at the beginning was Closed – a series of roles to perform that left no room for self-expression. In a suspicious happy ending, our heroine prepares to marry her savior-prince, a man we’ve only known from his glowering closeups and his beating her to submission a scene earlier, by pulling up her hair and applying her makeup in a mirror. The show must go on.

Presence isn’t an option in the endless mandate to represent one’s family and country and roles in Shimizu’s Closed films. After all, listen to her complaints one more time. What she’s articulating is the political dream of every Shimizu hero young and old. She doesn’t want to be the hero of some film or family or stage. She wants to hang out and chill and just be.

She wants to inhabit the Open film.

A Woman Crying in Spring (1933)

The Closed Films are essential to understanding Shimizu, I think, because they give us the flipside of the children’s mock-playacting and playact-mocking from the Open films. Instead, it is the adult women here who all act in arch theatrics, as they find themselves beaten-down into roles to both subserviently please and strategically maneuver the men around them. Other Shimizu films concern the travelers who pass through other people’s lives. But in the Closed Films, we chronicle these other people’s lives as everyone they meet passes them by.

They remain stuck in the places that everyone else will flee.

VI.

Mr. Thank You (1936)

There are three types of films, I sometimes think: films that take place behind closed doors, films that take place entirely outdoors, and then, that secret, complex third thing, films that take place in open doors.

Somewhere between Open and Closed lie Shimizu’s three most beloved films: Mr. Thank You, The Masseurs and a Woman, and Ornamental Hairpin. All three crosshatch society at some stopoff point where the wealthy and the working class will meet; all three chronicle characters as passing ships, finding connection in the only place it could exist – and the only place it cannot last. And all three films take place in an outdoors that is also indoors, a vagrant site where strangers congregate to hang out in a moment of leisure, an outside as situated in a bus, a resort town, and a proper resort. We are between the Open and Closed films physically and narratively: the adults here operate in the kind of holiday space that destabilizes their own authority, leveling them to forge relationships with one another like new kids at the school. The fact that their confinement is temporary is what makes the films comic and tragic alike.

In these Open-Closed films, plot arises organically only in the final stretch of the film as a kind of byproduct of the characters watching and studying each other in criss-crossing interactions across the first two thirds of the movie. As usual in Shimizu, those interactions are enabled not only by a series of lateral tracking shots and deep focus compositions that feel unable to contain the characters they marshall from space to space, but by a transitory site that enables the characters to play at something other than their job in the outside world.

What am I supposed to write in my diary? the kid in Ornamental Hairpin asks. I do the same things everyday: I eat, I wash, I sleep. Well, says the adult, don’t just think about what you do. Think about what you observe: the old men outsnoring each other, as if in competition; the hobbled man crossing the bridge; the people horrified by others having a good time. It’s that potential for emancipation – for the characters to be something other than what they do, to become stand-in audience members observing each other with love – that can turn Shimizu characters into ghosts of a sort, their bodies premature elegies for the lives they have led. For what are Shimizu’s transitory sites for people of all classes to intertwine and watch the lives of others, if not miniature cinemas?

The lower the stakes, the greater the resonance and relatability. The old man’s message is, of course, a manifesto for Shimizu’s own diaristic practice, to tell the story of romance and war through the banalities of people hanging out, to understand that that’s what it means to hang on the precipice of love and death.

In the hands of pretty much any other writer-director, Mr. Thank You’s 78 minutes in a bus would probably teach us the backstories of each passenger, their motivations for going to Tokyo, their political oppositions that would pit them against each other in a climactic brawl seismically heaving the bus against the cliffs. The snobbish, upper-class mustachioed would stand in as the specter of repressive norms, taunting the young girl for her position as a courtesan. At some point, she’d call out the moral hypocrisies of all the passengers and break down in tears, as we wonder what side we’re on.

But that is not Shimizu’s Mr. Thank You. In fact, that is the opposite of Shimizu’s Mr. Thank You, in which we find out almost nothing about the characters’ past or present, infer the most major dramas through their silence, and discover drama in its most primal form: the character’s decisions about where to place their body, who to align themselves with as seatmates. The mustachioed man is mainly teased by the young woman, not vice-versa, as Shimizu builds drama from documentary – from the characters’ offerings of sweets and alcohol to pass the time, from the shifting countryside caught by enacting this entire film on site, by the real-life workers Shimizu met along the way who counterpoint the more upwardly mobile figures on the bus. Voyage to Italy would not arrive for another 20 years.

Mr Thank You ends with an elision as radical as the finale of Ozu’s An Autumn Afternoon (Sanma no aji, 1962): an entire wedding, conveyed only through the presence of the wife within a shot on a bus ride home from a city where she was supposed to stay. But the removal of the central drama everywhere in the movie is crucial to understanding the difference between Ozu and Shimizu. For Ozu, the drama is so important that it must be experienced through its consequences: the empty rooms of a home our protagonist has abandoned for matrimony. For Shimizu, however, the drama is so totally unimportant that what matters is not the staid ceremony of characters dutifully playing their role in the wedding itself, but the relationship in its smallest hang-out moments that let us witness the characters negotiating their own personality in real-time.

And this is key, here: the characters are constantly renegotiating the roles they think they should play with each other to break formalities for little sparks of feeling. The bus, we realize, is one of the few democratizing political spaces, like a rave might be today – a place where social and class difference dissipate in the face of sharing chocolates or a drink. It is a political place where, for a few hours, collectives can genuinely arise.

Mr. Thank You, our friendly busdriver, stands in for Shimizu himself, gently nodding his appreciation at all of life’s foibles and failures, collating these characters for a few hours to see what happens when you create a collective from a broken world. He is not so far from the titular protagonists of The Masseurs and a Woman, who in watching and steering the action come to take the place of the director and audience alike. The difference is simply where Shimizu tips his genre in the final minutes – comedy in Mr. Thank You (a wedding), and a kind of tragedy in The Masseurs and a Woman that’s far more familiar than any classical tragedy. The tragedy of being alone, living your life, with nobody to take care of you at all.

Shimizu would invert Mr. Thank You after the war. In Tomorrow There Will Be Fine Weather (Asu wa nipponbare, 1948), an opening 20 minutes of silent movie gags on a bus turns increasingly declamatory as the gaggle of bus passengers – who of course all happened to know each other years ago – turn against each other and themselves after the bus breaks down in the countryside. Shimizu writes and directs us into familiar terrain: a ragtag crew of physically disabled outcasts converge in a single location for a couple hours to confront their hopes and fears. Instead of a film about passengers in the bus, we now get one about passengers out of the bus.

Like most of Shimizu’s postwar films, the movie is more dramatic for a simple reason: the characters are angrier. In various ways, nearly every character rages at the ways they submitted to the inevitable compromises of the war, serving, not serving, sleeping with men who could help them dodge the draft to save their family. No real reconciliation proves possible in this ghost diorama – only small moments of grace from the usual Shimizu chorus of passengers on the sidelines, watching and intuiting the drama on our behalf as they recite poetry, take each other’s arms, and dissipate into shadows in the midday sun. It is Shimizu’s most Bruegelian film, a panorama of human emotions in a faintly unfeeling world. And it introduces a new note here, unfamiliar from his earlier work, that would come to inflect that output that followed. Grief.

VII.

Nobuko (1940)

I have been making the case for Shimizu as one of the Great Major Filmmakers, but I would be remiss not to make the case for him, William-Wellman-like, as one of the Great Minor Filmmakers too. Much like Wellman, for every small genre-defying masterpiece Shimizu would toss off here and there, he would immediately retreat to the most generic storylines he could find as a kind of skeleton that he could flesh out with formal experimentation alone. These genre films are, in a way, Shimizu’s sketches, as you watch him trying out new devices before taking on a sort of real-life painting a year or two later.

Still, there’s no way around this: Shimizu would undoubtedly live a lavish legacy in the annals of academic film journals if he simply hadn’t made most of the films he did. Frankly, without his intermittent piles of dreck so programmatic that he was dubiously renowned for telling his ‘50s crews to roll camera, no instructions on acting or cinematography worth his time, well, I would also have an easier time writing this piece. But just as form and content dissolve in his best films, they often diverge wildly in his worst, as formal experiments to vainly make fairly uninteresting material interesting.

In impressive muck like Home Diary (Katei nikki, 1938) – a bit like watching your favorite DJ remix your least favorite songs – you find astonishing formal daring-do: sweeping camera cross-cuts and zooms. Even a more classically conventional film like Nobuko (1940) will excuse a dazzling lateral shot tracking the protagonist in the background of another building from the camera as she flits through a crowd toward the foreground, just as Shimizu’s ethnographic verité musical Sayon’s Bell (Sayon no kane, 1943) ends with pigs and ducks waddling through a funeral march as the camera glides through a graveyard to catch them from afar. These are far from personal films, exactly: Nobuko’s classic comedy of a rural schoolteacher trying to instruct far more suave children could pretty easily be reset in the US with a Janet Gaynor-type lead, the script preserved word-for-word, while Sayon’s Bell remains an admirably-and-deeply problematic celebration of the indigenous population in Taiwan gamely assimilating to Japanese culture. But they are also both, in exact opposite ways, formally perfect films.

In particular, Shimizu’s propaganda films like Sayon’s Bell are kind of fascinating things because they substitute the requisite plot and character development – things which Shimizu prefers to approach obliquely in his major films, like a clothesline on which to hang pet movements, gestures, and jokes – with some morally hollow messaging. This means, of course, that we lose the emotional core of his best work for some empty preaching even he clearly doesn’t believe. And yet the assignment to hand us a message leaves Shimizu liberated formally to deliver it however he likes, a testament to the way that minor cinema lets directors do audaciously experimental shit with camera and sound: extended real-time singing sequences in unbroken tracking shots through mountain paths in Sayon’s Bell, or a legendary 100-shot crescendo of soldiers on the march in the dimensional whiplash of A Star Athlete (Hanagata senshu, 1937).

For all of Shimizu’s long lateral takes pulling bodies across intersecting planes in an almost slow-cinema-ish durational deep space choreography, A Star Athlete reminds us that he was one of the masters of cutting dynamically across characters on the move. That film’s infamous Noel-Burch-approval-stamped montage of the soldier’s march, for example, counterpoints a far less-remembered 8-minute stroll of two bickering lovers near the opening of The Golden Demon (Konjiki Yasha), from the same year. Protagonists wandering down mountain paths are a Shimizu staple, usually in single tracking shots drawing them forward, but in The Golden Demon, Shimizu’s camerawork operates in a kind of start-and-stop call-and-response with the action, pausing and flipping 180 degrees when the characters halt, precipitating them into confrontation with sparing close-ups, and then pulling them back into a harmonious lockstep with a slow-moving dolly.

If I ever got to teach a class on decoupage, I’d love to show these two sequences from A Star Athlete and The Golden Demon as exemplars of how to show characters walking. It’s as though the characters’ maneuvers are puppeteered by the camera’s own in a criss-cross ricochet of stasis and scuttling. Both are a tour de force of the ways camera motion and stillness can not only underline action but catalyze it.

VIII.

Children of the Great Buddha (1952)

Almost uncharacteristically, Nobuko practically initiates a cycle of films that feels the most important to Shimizu personally: the School Films. Nobuko itself is in some ways, I guess, a great work of studio fluff. But for the rest of his career, Shimizu would make occasional breaks from such for-hire throwaways to make his real high-order labors of love, his docufictions on schools. Three of these – Children of the Beehive (Hachi no su no kodomotachi, 1948), Children of the Beehive: What Happened Next (Sono ato no hachi no su no kodomotachi, 1951), and Children of the Great Buddha (Daibutsu sama to kodomotachi, 1952) – are traditionally linked as a trilogy starring the children of the orphanage Shimizu himself established after the war. But it’s a cycle that could extend far earlier and later as well. For there really has never been a filmmaker outside of Frederick Wiseman or Abbas Kiarostami as obsessed as Shimizu with schools: with the ways they represent imperial indoctrination and education-as-an-instrument-of-freedom simultaneously, the ways they corral social order out of the chaos of childhood at the frontiers of civilization, and the ways they place adults as government vessels, empathetic humanitarians, hierarchical punishers, and students of their children all at once.

Or rather – with the ways that schools themselves are movies in miniature, sites that force children as untrained assets of all that is raw and real to play a role for the adults around them.

As Kiarostami and Godard understood much later, any school film is documentary and fiction at once, or rather, a kind of documentary of its own staged fiction, a record of children learning to act and perform their assigned parts. I have personally struggled with some of Shimizu’s late-period work, which, like a lot of ‘50s films by great ‘30s filmmakers, seems to have mastered the unfortunate art of genre competency. Who does Shimizu think he is proficiently eyeline matching all these shots-reverse-shots? Howard Hawks? But at some point I had to accept that his late-period game is just interested in different things. Or rather, he’s interested in the ways that the most basic point-and-shooting can give space for characters to present themselves in their own terms.

Shimizu’s earlier films are dazzling choreographies, but in a way he had to escape his own formal mastery for the films to become cocreations with his young actors. The School Films are as much their films as his, as Shimizu remains one of the few filmmakers genuinely interested in leaving his camera running to learn from them. Shimizu’s style, here, isn’t reduced to convention out of the usual late-stage soullessness. It’s reduced to convention because he no longer can force the action of the film to obey his formal precepts; it’s reduced to convention because he’s leaving room for this fiction to molt its skin and emerge as documentary. He has to point and shoot as simply as possible to let the action go where it needs.

Only a quarter of Shimizu’s filmography survives, but my sense is that The School Films kick off properly with Introspection Tower (Mikaheri no tô), released in January of 1941, the same year as Ornamental Hairpin, Notes of an Itinerant Performer, Record of a Woman Doctor (Joi no kiroku), Dawn Chorus (Akatsuki no gasshô), and Acorns (Donguri to shiinomi). The documentary trappings here that would come to dominate Shimizu’s work, including an extended opening tour of the actual school, are both a kind of light agitprop touch of state-sanctioned filmmaking, as well as an abdication of authorial clarity. Like many of Shimizu’s attempts to humanise propaganda by making the figures more conflicted about their duty, this turns out to be a frustratingly ambiguous film: a paean to the collective labor of, um, uh, children, as well as a record of institutional indoctrination of kids that all too often seems to be indoctrinating us as viewers into its view of state apparatuses as kindly places to homogenize the marginalized masses.

And yet the Shimizu style of making a movie – map out an enclosed institution in the middle of nowhere (bus, spa, school), pack it with characters from all walks of life, and see how they negotiate their space with one another, how drama can erupt from documentary as the professors start hanging out on playground swings – feels radical enough in the 1940s to completely destabilize this kind of work of ambient propaganda, and make you wonder whether this is truly smirched by the gunk of imperial sanctimoniousness or in fact a trenchant criticism of schools as fascism’s ground zero around the world.

The uneasily gentle but not quite uncritical view of authority continues in Record of a Woman Doctor, starring Shimizu’s ex-wife Kinuyo Tanaka. In a way, Woman Doctor inverts Nobuko from the previous year: instead of a rural bumpkin going to the big city to learn a lesson from her students in an ultra-classical comedy, we now follow an urban doctor going to a small village to learn a lesson from the children in what amounts to a kind of documentary PSA. And yet both serve as a kind of generic Rosetta Stone to the tropes he would reorient across another half-dozen films of the institutional figurehead trying and failing to play a role that hides all sense of self.

One decent heuristic for Shimizu is that his best films tend to be two things: plotless and crowded. Indeed, Record of a Woman Doctor is both at once, a scenic portrait of rural mobs: throngs of villagers pressed hard against the doctors to witness unseen operations; packs of schoolchildren served tea in class by a teacher who changes diapers and bemoans the lack of day care. This teacher is the closest thing in a Shimizu film to a hero: a man who gave up his dreams to attend to the dull necessities of daily life in the provinces.

Given the vaguely propagandistic focus here – country folks learning the importance of modernization through that familiar domestic soldier archetype, The Good Doctor – it’s easy to miss that this is also a School Film, posing the key questions of all his great films: how to separate the dancer from the dance, or rather, the bureaucrat from the bureaucracy, the worker from the work. If nothing else, schools are perfect environments in which to test humans’ own commitment to their institutional role. This being Shimizu, that commitment will of course show its cracks, as townspeople and doctor learn to love each other—a sign of humanist classicism that’s harder to find in the post-classical School Films of Kiarostami or Wiseman, but one whose sentimentality was probably as radical a gesture as Shimizu was permitted in the context of a medical ad he worked to elevate to semi-neorealist greatness.

And there’s a bigger point here:

Propaganda proved a formal turning point for Shimizu, a forcing function to extol the kind of adult authority he had always despised by dealing with it as blankly as possible through the alibi of an impartial documentary. If Shimizu’s later films take a more ambiguous view of institutional authority, it’s because propaganda had trained him to undermine the authority figures he was supposed to eulogize by making them haplessly human – rendering figures like the soldiers in A Star Athlete as lovably incompetent. It was a humanist’s way to both subvert and problematically magnify propaganda at once, contouring the Imperial Army like a frat party, but it would transform Shimizu’s desire and ability to film institutions with grace. What propaganda gave him, I mean, is a kind of language to chart postwar Japan through its schools.

In what remains Shimizu’s most canonically important film, Children of the Beehive, Shimizu gets closer to documentary than he ever could in the studio era, chronicling the cities and towns of postwar Japan through the eyes of the non-actor kids of his real-life orphanage. Weirdly, though, what makes this a work of classic Neorealism, is not simply his dedication to ‘40s verité, but to using ‘40s verité to make his most blatantly hamfisted melodrama since the silent era – shrink-wrapped in the kind of tirelessly prestigious score popular in postwar documentary, filmed in shaky shot-reverse-shots that upend the deep-focus durational mise-en-scène of his earlier work. But the documentary and melodrama aren’t really in tension: both the disingenuous form of “life caught unawares” and the depiction of working class poverty are just evidence of how raw this movie is, for better and worse the work of a filmmaker completely relearning his craft, like some kind of student filmmaker making the best of a MacGyver toolkit of a couple kids, a camera, and a devastated nation.

So look: what I personally think is pretty insignificant against the tide of critical consensus. But for me, Children of the Beehive: What Happened Next is his greatest school film, a miraculous assemblage of absolutely mundane real-life interactions between the kids as they discuss accents, catching raccoons, peeing the bed, and even, in an extraordinary 4th wall break, the filmic contrivances of the first movie. (Characters in the School Films have a compulsive tendency to comment about watching earlier entries like Introspection Tower and Children of the Beehive to emphasize that those weren’t real life.) All of this banality is necessary not only as a record of lived-in history back in the countryside – a direction Shimizu had been pursuing in his propaganda years – but to ground the melodrama in the fullest credibility when it comes.

And when it does finally come, in a tight sequence of a boy failing to meet the mother who has abandoned him, we see that Shimizu was there all along. The culmination of this sequence is two shots – a long shot as he runs in a distant plane to the ocean, a speck within the shot, and then a shot not of the boy but of the ocean waves themselves as they overtake the camera – that in all their gentle, devastating violence, feel impossible for any other filmmaker to have conceived.

It’s not until the final installment, Children of the Great Buddha, that Shimizu goes full documentary in a semi-send-off to his career. His last independent film, it’s a movie that Shimizu clearly relished as his great passion project, spending a year researching Buddhism to put everything he learned into a film, like an audiophile self-financing a film about their record collection. Plot is mercifully inconceivable here. The kids themselves take over the documentary, narrating a series of tours of the Buddhist temple in Nara, often spuriously, correcting each other as they dream of finding home, or at least a night of sleep in the Buddha’s arms. And the sequences of the boys circling the statues, as if at any moment the past, what little is left of this glorious past in bombed-out Japan, might come alive, well – the sequences of the boys circling the statues represent some of the most haunting postwar filming this side of Resnais. In a career trying to film documentaries with the tools of fiction, this is probably the closest he came, and a film nobody else could have made. It’s boring, beautiful, bewildering, and one of the apotheoses of one the greatest filmmakers.

Shimizu’s final School Films would attempt to intertwine all the strands of his career. In theory, The Shiinomi School (Shiinomi gakuen, 1955) should mark some kind of perfect intersection of Shimizu’s films on children and the physically disabled: the boy with the broken leg in Four Seasons of Children, the hobbled soldier with the wounded foot in Ornamental Hairpin, the sick and out-of-shape soldier in A Star Athlete. But in this late film about disabled children who have been bullied by the outside world, the landscape has become both more sentimental and more hollowed-out, more bleak. In those three earlier films, the resolution would always remain the same, pulled directly from the opening of Shimizu’s first extant feature (The Blacksmith of the Forest / Mori no kajiya, 1929): another character carries the hobbled child on their back. Individual struggles are, in those more sanguine Shimizu films, a test of collective solidarity to overcome. In their own ways, those films would all detail the socially downtrodden developing requisite skills and networks far surpassing those of very lonely adults whose own quests for power destroy their ability to collectivize.

By contrast, at the end of Shimizu’s career, The Shiinomi School faces us with its mawkish tale of do-gooder adults selflessly saving a pen of fairly anonymous disabled children. Gone is the ambiguity of Introspection Tower about the role of education in standardizing the quirks of kids’ behavior in service of some imperial notion of normality. And yet, with all of Shimizu’s great focus on restoring fiction to documentary, this remains almost certainly the best document of the methods of schooling disabled youth until Wiseman’s series on the Alabama Institute for the Deaf and Blind in the 1980s. The end of the film – the children struggle phonetically to read their notes to their dead friend, real children reading the script of the film itself, their letters rippling in the wind – attains a kind of sadness unique to Shimizu: a devastation begotten out of the banality of the awkwardness of everyday life.

That devastation overtakes Shimizu’s final School Film, Girl’s Reform School (Naze kanojora wa sô natta ka, 1956). Still, it might sound strange to say about a movie that covers self-induced abortions, sex trafficking, and child abuse over the course of 80 minutes and about half a dozen docufiction vignettes, but in a sense Girl’s Reform School is not only Shimizu’s cruelest film, but the closest he ever got to filming utopia. For what he gives us is a kind of marriage of heaven (the school as a site of salvation) and hell (the outside world as a crucible for all the girls’ bad behavior).

In so many ways, it is the culmination of decades of work and nearly a dozen films on schools. As usual, we find ourselves in a familiar rural getaway for outcasts who will form a microcosm of outside society as they intersect across slow tracking shots wafting in and out of buildings and pulling our protagonists down hallways in slow-motion pirouettes. But more fully than in his other work, Shimizu dares to imagine his most progressive ideals actually working, if only for a few brief moments. What if all the cruelty of Japan’s post-war existential desert internalized by a population of dispossessed preteens… well what if that cruelty could be stared down by kindness, crafts, and lullabies?

The movie never pretends these will solve the ills of making money to feed a hungry family. But it does imagine they’re the basis for a semi-anarchist space where kids can be given sovereignty instead of punishment. As often in Shimizu, the point is not simply how easily we could give each other grace, but how hard, how impossible, it is for all of us to receive it.

IX.

Mr. Shosuke Ohara (1949)

I have no great interest in biography, and neither did Shimizu: there are no great heroes in his film, no transcendent love affairs, no depictions of the tortured depths of the human psyche. There are only a few occasional Shimizu characters obsessed with writing their story – the money lenders in The Golden Demon and Four Seasons of Children – and they are all villains. The characters Shimizu loves are the ones, like him, who crave coexistence with the world, who lead their lives as rituals cocreated with their social and natural landscape. In a way, he reminds me of another filmmaker who was born extraordinarily wealthy and spent her life making semi-documentary performance pieces of the poor, Shirley Clarke.

And yet, in one of Shimizu’s last great masterpieces, Mr. Shosuke Ohara (Ohara Shôsuke-san, 1949), we get something akin to an allegorical autobiography. This is not exactly apparent. This is a film and filmmaker, after all, that chooses to depict two men sending off a donkey by following the donkey instead of the men: if there are heroes here, the camera is decidedly uninterested in them. And Mr. Shosuke Ohara is full of these meandering touches of the incidental that few other filmmakers would show, then or now: men slumped over drunk in the corners of shots, a woman clipping her toenails during conversation, a series of tracking shots of a decaying ceiling when the protagonist talks of all he’s lost. In one of the many dazzling lateral tracking shots, the camera follows the trajectory of two characters through an adjoining room of seamstresses on the other side of the wall from where they’re ostensibly walking, and the effect of Shimizu’s filmwork here – with its deep focus vision of multiple rooms, even multiple stories in every shot – is to track the wealth of objects and roles as they’re slowly hollowed out to nothing by the end.

But you see where I’m going. It’s hard not to wonder if this story of a charismatic loaf “born with a silver spoon” who claims to spend all his money preserving his privilege, but actually, we suspect, is wasting it all in a revolt against his inheritance so he can spend time with vagabonds and thieves instead, is something of Shimizu’s self-portrait. For what’s this story, if not that of a director who can no longer direct his world?

But in that sense it is also, I guess, something of a post-war manifesto for Shimizu’s own artistic practice in the wake of a barbaric era.

Image of a Mother (1959)

I’ll say it again. ‘50s Shimizu is not remotely the same filmmaker as ‘30s or ‘40s Shimizu. To accept him means accepting a filmmaker who feels defeated by the classical conventions of his era, both in terms of the society he depicts and the grammar he employs to do so. By the end of his career, the long-held long shots with several planes of intersecting action, the lateral tracking shot, the metered returns to the same shots with shifting perspective—all of these staples of his earlier work would dissipate into a kind of blandly literate regimen of shots and countershots, establishing shots, and close-ups. These late films will occasionally come alive for a few moments in supple tracking shots of a singer commanding the camera to follow her like an amulet (The Sentimental Idiot / Ninjo Baka 1956) or of street life in the stalls (Dancing Girl / Odoriko, 1957) or of gliding shot across an empty house (Tale of Jiro / Jirô monogatari, 1955), like a cinephile startled out of his stupor to glimpse, for a second, the light from the machine.

The mechanistic frigidity of 1950s filmmaking was clearly very good for formalists like Hitchcock and Ozu; for freewheelers like Shimizu, so much was lost. There is another essay to be written here on exactly why the most rambunctious visionaries of the ‘30s found themselves bound to the cogs of melodrama machinations in the ‘50s, a period in which movie producers seem to have figured out just a little too well what worked. But there is also, I suppose, some case to be made for any Late Style generally that feels haunted by phantoms precisely because the filmmaker has erased the life from their work. In Sound in the Mist (Kiri no oto, 1956), Shimizu does an exceedingly good impersonation of Naruse via Sirk, as bodies careen into foreground in a kind of formal, split diometer exaggeration of space that exaggerates the melodrama in turn; I’d like to wishfully think of A Mother’s Life (Bojô, 1951), in which an impoverished mother tries to offload her children to her family, as some kind of allegory for Shimizu abandoning his own practice.

And yet, there are wonders here. In some alternate universe run by a less eurocentric critical establishment, Tokyo Profile would surely be listed along with Bicycle Thieves (Ladri di biciclette, 1948) and the quite similar Little Fugitive (1953) as the defining works of urban neorealism. As often is the case in a neorealist city, a child is lost, though here she serves something of the same role as the earrings of Madame De or the STDs of Le Plaisir (1952) or the gun of Winchester ‘73 (1950) or the debt of The Inside Story (1948): a midcentury hyperlink, a circulated structuring device to teach us about characters by the ways they respond to her. But unlike in any of those films, which condense time and characterization to a series of vignette crystallizations, Shimizu’s focus is characteristically duration: what the characters do when they have nothing to do, how they fill their free time telling stories and performing silly walks, how these archetypes of contemporary society intersect in near-real time and real locations across a series of tracking shots mapping their movements against each other.

Like so many of Shimizu’s late movies, Tokyo Profile is another cruel film – stories of theft, poverty, cheating, and exploitation all structured around a lost child – told like a musical comedy, as the characters soft shoe their way through streets and cantinas and classes and music halls and towers as if at any moment they could break into dance. Shimizu feels at every moment on the verge of inventing a whole new genre that it would take Chantal Akerman to perfect decades later: musical realism.

But the full bleakness of Shimizu’s late years on fully flourishes – or is it wilts? – in his final film. It feels like one of those jokes of the film gods that the first film of Japan’s great bad boy Nagisa Oshima and the final film of its gentle good boy Hiroshi Shimizu would each come out in 1959 as studies of the cynicism of pigeon-raising youth looking to the sky in a grimy Tokyo. Then again, I suppose you could argue that Oshima – with his return to pre-classical 90 degree cuts, focus on children betrayed by an adult world, and experiments in documentary fiction – picks up strands of Shimizu across his work. For sure, the world of ghosts in Image of a Mother (Haha no omokage), here the remnants of a child’s dead mother and the life she left behind, is a world that brims like Oshima’s with latent anger, with eruptions of wild physical violence out of a deadened landscape of slouched bodies and slowly dripping taps.

And so, when you accept this hollowed-out filmmaker, when you see the anger at his core at what this prosperous new world built on an atomic bomb entailed, you can see what it means for him to return to his favorite subjects, the cruelty and anger of children, and to set them in a nearly-plotless rumination of childhood anger – to set them in the land of the dead. The place where Shimizu’s filmmaking ends, in 1959, in the last proper year of classical cinema, is the place where filmmaking ever since begins.

September-December, 2024

With special thanks to Edo Choi, Alexander Fee, The Museum of the Moving Image, The Japan Society, and all the projectionists, print handlers, restorationists, and interns – and I was once one myself at MoMI – who made the Hiroshi Shimizu retrospective possible.

Originally published in Outskirts Nº3 in July 2025. The print issue can be ordered at our online shop.