Continuity editing was one of the great collective inventions of filmmakers. In the ten years after it crystallized in Hollywood around 1917, it was adopted around the world. The technique has hung on a surprisingly long time, rather like geometrical perspective in pictorial art. It’s so powerful that it’s hard to escape.
It’s powerful partly because it’s adaptable to a lot of narrative situations. It provides filmmakers what we can call a set of stylistic schemas, or routine patterns, that can be adjusted in various ways. Analytical cuts that take us into or back from the action, shot/reverse shots and eyeline matches, over-the-shoulder framings (OTS), and slight camera movements that set up reestablishing shots: these straightforward schemas can be used in an indefinitely large number of ways.
I‘ve been noticing this flexibility while writing (still!) my book on narrative strategies in 1940s Hollywood. In particular, I was looking at fantasy tales and the peculiar problems they pose for filmmakers. First, how do we represent ghosts, angels, and other visitors from the Afterlife? And how do we make sure that audiences understand that what we see isn’t necessarily what some of the characters onscreen are seeing?
Our Town (1940).
On the first problem: Today we have CGI resources that allow Afterlifers to move freely among the living characters. These effects were much harder to achieve back then. The supernatural-fantasy genre developed its own conventions (yep, schemas). Most films resorted to presenting the Afterlifer in double exposure.
But superimposed characters can’t move easily among the clutter of furniture in a set. A supered ghost can’t go behind a sofa; the sofa will always be visible through it. You might resort to a traveling matte shot, as William Cameron Menzies did in Our Town (1940), when Emily revisits her family after her death. The shot (just above) is particularly flashy because the younger Emily is also in it.
Or you could pull off the remarkable trick in Earthbound (1940). Here a ghostly Warner Baxter settles comfortably into the middle ground and strides around behind furniture and other actors.
He can even give up his seat to an old lady and shift to another.
These effects were achieved by a technical feat that I don’t fully understand. (See the codicil.) But the trick wasn’t widely adopted, and most filmmakers opted for a simple expedient. Typically the Afterlifer shows up as a superimposition, but he or she gradually materializes and becomes a solid presence like all the other actors. This is from The Canterville Ghost (1945).
Then comes the second problem. Sometimes the living can see the spectre, but sometimes not. In Here Comes Mr. Jordan, the angel Jordan and Joe the prizefighter arrive from the Beyond and watch the murderous couple from behind. They aren’t visible to the living, but they’re just as tangible.
If a living character seems oblivious to the otherworldly guest, we’re to assume that the guest is invisible. Each film has to inform us of who can see what, and most films do—redundantly. The spook or divine messenger will explain that the living can’t see them, or that only certain characters can. (Sometimes children and animals can see them while grownup humans can’t.)
These conventions can get tweaked. Once we know the Afterlifers aren’t visible to the living, they can comment on the action from the sidelines, as when the dead pilot in A Guy Named Joe (1944) slips in wisecracks while his girlfriend is wooed by a young aviator. In the comic The Man in the Trunk (1942), an all-too-tangible ghost can’t follow a character leaving a room. He explains, “I failed my examination on how to walk through walls.” In The Remarkable Andrew (1941), the ghosts of US founding fathers can ransack offices for evidence in an utterly unconstitutional search and seizure. Most ghosts can pick up objects, but the ones in The Cockeyed Miracle (1946) can’t, a fact redundantly explained to us. This generates suspense when they try to grab a fallen bank check. (Incidentally, this is a long, skillfully directed scene, and it does have recourse to a matte shot when one ghost tries. unsuccessfully, to hide the check by standing on it.)
All of these tweaks rely on continuity editing. But in the course of the 1940s, I’ve noticed, filmmakers played a little more ambitiously with the Afterlife conventions, and continuity style allowed them to do it. The results are sometimes provocative.
Ghosts coming and going
Consider the arrival/departure of the Afterlifer. The default is the special-effects twinkling that lets him or her materialize into the scene and fade out of it. But some filmmakers tried more natural options. In Alias Nick Beale, the title character, aka Satan, strolls in from offscreen, or, thanks to John Cromwell’s staging, is masked by other players before being revealed on the scene. He’s presented to all the characters as a real person, but the staging gives him arrivals of relaxed abruptness.
Joseph Mankiewicz’s The Ghost and Mrs. Muir (1947) finds other ways to avoid the hugger-mugger of transparency and melting departures. Captain Gregg assures Mrs. Muir that to her he is “like a blasted lantern slide,” but for us he just steps into the frame and stays on as a solid presence. He disappears just as simply: Mrs. Muir turns away, then a new shot shows that Gregg has gone. Good old shot/reverse shot does the trick.
It’s neat that the second reverse shot on Mrs. Muir reaffirms Gregg’s absence by making her seem more isolated than the earlier mid-shot does. This cut-back to show her isolation after his departure is stressed even more in a later scene, when a medium shot shows her turned slightly away from him; in that interval, he disappears again.
Only after the captain has decided to leave her forever does he resort to the standard spook trick of dissolving away. But in this melancholy context, the familiar device takes on a forbidding finality. He leaves her sleeping and has magically made her forget all about their year together.
In the epilogue, when the elderly Mrs. Muir dies, the captain returns, sturdy as ever, and again the film avoids the cliché. The default schema is to let the dead person’s spirit float up in superimposition from the corpse. Instead, Mankiewicz presents Gregg standing over the lady in a tight shot and simply lifting the now eternally young Mrs. Muir into the frame. Again, all we need is shot/reverse shot, this time with the extra intimacy of an OTS.
These uncommon options take us a little by surprise, refreshing the genre conventions, while also suggesting that the films drive a little deeper into the heart than the usual spook story. The simplicity of presentation helps us take their spectral affair more seriously.
In The Bishop’s Wife (1947) the angel Dudley, like Gregg, comes and goes via offscreen space. A pan follows Henry the bishop, who hears a noise outside his study. The shot pays off with a reverse-angle cut to Dudley, now magically present at the fireplace where Henry was.
At other points, as in The Ghost and Mrs. Muir, the framing excludes Dudley, and when we cut to the long shot Dudley is gone.
Soon, though, there’s a gag on the device. Henry, furious with Dudley, turns away and prays that Dudley will leave him. As earlier, the camera tracks in.
At first it seems that the prayer has been answered, when the camera tracks back from Henry and we see his slightly surprised expression, implying that Dudley has vanished.
But it’s then revealed to us, before Henry knows, that Dudley has changed position and is still in the room.
This quiet revision of the schema is in keeping with the film’s other jabs at Afterlifer conventions. At another point, Henry demands that Dudley execute a miracle. He locks the study door, evidently expecting Dudley to stalk through it in the usual phantom fashion. Instead, Dudley just twists the knob, magically opens it, and exits, closing it behind him and leaving Henry to bang against the relocked door.
Re-turning the screw
In general, supernatural romances play down the magical side of the Afterlifers’ visits. This is partly because they rely on a certain ambiguity about who can see what.
Henry James’ brilliant tale “The Turn of the Screw” provides an instance that people have argued about for generations. The unnamed governess, sent to take care of little Miles and Flora, starts to see dead people: the servant Peter Quint and the governess Miss Jessel. Most of her encounters with the ghosts take place when she’s alone, or with the children. Since the story is narrated from her viewpoint, and in the first person, we have no other testimony about the apparitions. The children seem to spot the spooks, but we can’t be sure. And we can’t be sure that they aren’t simply figments of the governess’s imagination. In the one scene that brings in another witness, the housekeeper Mrs. Grose declares that she can’t see Quint or Miss Jessup. Is the governess mad, or are the children really haunted by the corrupt couple?
The tension exemplifies what narrative theories Tzvetan Todorov called the “fantastic,” the tale of uncertain explanation. The fantastic hovers between scientific, or at least real-world explanations, and supernatural ones. Either the ghosts exist, as in most ghost stories, or they can be explained psychologically, as the narrator’s hallucinations.
Actually, in film, and I think in “The Turn of the Screw,” there’s a third possibility: that the Afterlifers exist as beings who can be seen only by the select few. In “The Turn of the Screw,” Flora definitely seems to see Miss Jessel on one occasion, so perhaps we can posit that the governess sees the ghosts when Mrs. Grose can’t.
This possibility is of course the premise for The Sixth Sense, and we find it as well in The Ghost and Mrs. Muir, in which Gregg is visible only to Mrs. M. The situation is more equivocal in Portrait of Jennie (1949). In this film, David O. Selznick inflated the fantasy-romance genre as he had pumped up the historical drama (Gone with the Wind), the home-front film (Since You Went Away), and the western (Duel in the Sun).
During the winter and spring of a single year, the luckless painter Eben Adams encounters Jennie at different points in her life: as a little girl, a teenager, a college student, and a mature woman. He falls in love with her. Their penultimate encounter takes place when she comes to his apartment and allows him to paint her portrait. Then she disappears.
Is Jennie a time traveler or a ghost or an illusion, or even a supernatural muse? A bit of each. After each brief visit she withdraws, leaving Eben yearning for her. The art dealer Miss Spinney suggests that in order to paint well, Eben must love someone, so perhaps he has conjured up Jennie to inspire his art. It’s true that he sketches, draws, and paints several Jennies. (The final version, the portrait, won’t be shown us until the film’s last image, in blazing Technicolor.) And it’s true that, in a scene much like that featuring Mrs. Grose in James’ tale, we get a hint that Eben is hallucinating Jennie.
He has just spoken with the adolescent Jennie in Central Park when Spinney comes up to him. He watches Jennie go off, in a standard passage of continuity editing.
But then, thanks again to continuity eyeline technique, Spinney doesn’t see her.
Normally this cutting pattern would suggest that Jennie is purely Eben’s vision. (For a modern example, see Johnnie To’s The Mad Detective.) Yet we will soon learn that Jennie did exist. Playing detective, Eben discovers that she was orphaned, taken care of in a convent, and left for college—all things her apparition told him. Worse, he learns how she died: on a rocky seacoast he has already depicted in some paintings.
Selznick was concerned to make Jennie neither a real person nor a pure product of Eben’s imagination. In an early story conference he remarked that Jennie is both in Eben’s mind and in some really existing realm. “We must convince the audience that this story may be strange and odd, but it’s true. All the other characters may say, “Poor Adams, he must be nuts,’ but we know it is true.” The fact that only Eben can see Jennie doesn’t make her a figment of his imagination. Eben may have conjured her up, but she also chose to visit him, predicting that some day he will paint her portrait. The man has, somehow, met a woman from a spiritual world who has been seeking him. Eurydice-like, she will be pulled back into it.
Too few fancies, one powerful friend
Recognizing that spirits can become selectively visible to the living helps explain the delicate power of another supernatural fantasy of the period. The Curse of the Cat People (1944) uses hyper-judicious framing and editing to create perhaps the most mysterious 1940s Afterlifer.
Irena, the woman who can change into a panther, has died in The Cat People (1943). Her widowed husband Oliver has married Alice, and their child Amy, dreamy and unpopular, wishes for a friend. Near the start of the film we’re led to think that Irena is that imaginary friend, wholly a product of Amy’s mind. Irena comes to her in the garb of a traditional princess or fairy godmother, a bit like the fairy of Pinocchio (1940), so we might take her as Amy’s imagining. And the teacher Miss Callahan, who might seem to be playing the raisonneur, says that the little girl has “too many fancies, too few friends.” But that doesn’t seem to me quite right. I don’t think that ultimately Irena is Amy’s projection. Nor are we exactly in the realm of Todorov’s fantastic, hesitating between subjective and objective explanations.
Consider the progression in the film’s depiction of Irena. At first, Amy is shown playing in the garden, purportedly with her friend but alone in the frame.
Later she says that her friend taught her a song, but she can’t recall the words. This suggests that that encounter, kept offscreen, is a vague one. That night, as Amy awakes from a nightmare, she is soothed back to sleep by her friend, whom we hear softly singing but see as only a shadow. Her lullaby continues as Amy sleeps.
Still later, Irena appears to Amy in the garden only after Amy sees a picture of her. Irena appears un-magically, entering the shot as casually as Dudley or Captain Gregg and holding the ball that Amy has tossed out of frame.
Amy asks her who she is. “I’m your friend.” The fact that Amy doesn’t recognize her from earlier friend-encounters, suggests that the friend wasn’t yet defined in bodily form. She had magical powers, such as darkening or lightening the garden or teaching Amy a song, but she assumed no particular shape. We, however, saw her female silhouette while Amy slept. Now Irena is fully embodied, and we can see her along with Amy.
Apart from the Irena scenes, we do get into Amy’s mind. But the techniques used in those scenes are ones that are never applied to Irena. When a deranged neighbor lady recites the tale of the Headless Horseman, we get exaggerated optical POV shots from Amy’s perspective and the subjective sound of wind and horses’ hooves.
The sounds are repeated as auditory flashbacks in Amy’s nightmare and while she is hiding in the forest during the climactic snowstorm. Later, Amy will calm the old woman’s homicidal daughter by envisioning her as Irena (in a superimposition) and embracing her as “My friend.”
From a storytelling standpoint, such images and sounds are sharply distinguished from Irena’s scenes with Amy. Those are treated in quite a neutral, objective manner.
There’s more. At the midpoint of the plot, in a crucial scene, Oliver learns that Amy considers the dead Irena as her playmate. He’s convinced she’s imagining it all. Anxious and angry, he takes her out into the garden and demands to know if Irena is there. Amy sees Irena under the tree, but Oliver doesn’t. As in Portrait of Jennie, reverse-angle cutting conveys each character’s vision.
In cinema, we assume objective (fictional) reality to be the default value. So the incompatible viewpoints of Amy and Oliver in the garden might suggest that Irena exists only in Amy’s mind. But in The Ghost and Mrs. Muir and Portrait of Jennie, the protagonist can see the ghost when no one else can. Nothing in this garden encounter denies the possibility that Irena is a ghost visible only to Amy and us.
We get some immediate backing for this. When Amy looks at Irena a second time, at Oliver’s insistence, Irena puts her finger to her lips, as if urging Amy not to acknowledge her.
This, to put it stiffly, is the action of an independent agent. Amy is unlikely to conjure up an imaginary friend who warns her to keep quiet. And by telling Oliver that of course she sees her friend, she disobeys just as briskly as if Irena were of flesh and blood.
We have a clincher at the very end of the scene. Oliver and Amy go in, turning away. Neither sees the garden, but we get two shots of Irena watching them and reacting unhappily.
Again, it’s hard to reconcile this image with Irena being a pure projection. She’s behaving like an ectoplasmic free agent. In addition, Alice and Oliver, despite their skepticism about Amy’s friend, don’t rule out supernatural goings-on completely. At various points both mention they sense Irena’s presence in the house. Irena has become in the course of the action a full-fledged ghost, but one with benefits.
The epilogue confirms Irena’s otherworldly mission. As Oliver takes Amy into the house, he asks if she can see Irena. She looks and sees her, smiling.
Can Oliver see Irena? Now that he’s decided to trust Amy, he says yes—without even looking.
As in the earlier garden scene, once father and daughter have gone inside, we’re treated to an independent shot of Irena under the tree. And now she melts away, in the conventional disappearing act of an Afterlifer. As with Captain Gregg’s withdrawal from Mrs. Muir, by saving this well-worn option for this moment the filmmakers invest it with an air of permanent departure.
The Curse of the Cat People cleverly invokes the possibility that Irena is purely imaginary, then dispels it. For once, nobody redundantly explains to us who can and who can’t see the ghost; we have to figure things out. Instead, the vague role of imaginary playmate gets gradually defined as Irena, the ghost. Continuity editing is recruited to suggest that Irena is coalescing into the friend Amy wishes for. At first only an atmosphere, then a shadow, and finally a properly solid spirit, Irena is a shape-shifter, more elusive than other apparitions of the period. The plot creates a sympathetic spirit seeking to console a lonely child and correct her father’s harsh, plodding common sense. Perhaps making amends for her effort to destroy Oliver’s romance with Alice in the previous film, the dead cat-woman fills out the role of imaginary playmate and saves a family.
There are other ingenious ways that the conventions of 1940s supernatural films are tweaked by the resources of continuity style, and I’ll be considering them in the book. The general point is that even schemas as commonplace as eyeline matching, or conventions as hoary as having a ghost dissolve out of the scene, can take on new force when filmmakers practice their craft with discreet intelligence.
Earthbound‘s ghost effect derives from a prism set in front of the camera. Warner Baxter is located in an adjacent set, and one face of the prism was silvered to reflect him onto the primary set. The best description I’ve found is here, thanks to good old Lantern. But we need more information. How can Baxter move so precisely around “our” set if he’s offscreen? Presumably, he’s in either a black set or one with furnishings laid out like ours. In that case, the furniture would need to be draped in black, so parts of his body will be blocked to the right extent. But in either case, we need to explain how he manages to synchronize his movements so exactly with the actors in front of the camera. If you know more, please correspond!
My quotation from Selznick comes from “Portrait of Jennie Conference notes (1/20/47),” David O. Selznick collection, Box 1123, file 11, Harry Ransom Research Center. Thanks to Emilio Banda and to Steve Wilson, Curator of Film at the Harry Ransom Center.
You probably know that “The Turn of the Screw” was filmed as The Innocents (1961). It’s also the basis of a strong Benjamin Britten opera.
A vast survey of Afterlifers on screen is provided in James Robert Parish, Ghosts and Angels in Hollywood Films (McFarland, 1994).
For more on my still-in-progress book on 1940s Hollywood, go here and here. If you’re keen on the Forties generally, you might be interested in The Rhapsodes, my study of film criticism of the period, to be published in March by the University of Chicago Press.