3 Hollywood Theorists: Edward Dmytryk and Walter Murch

Editing theories have been advanced not just by academics, but also by Hollywood editors, working professionals whose immersion in the art and craft of editing led them to write significant tracts offering their insights on how editing “works.” Edward Dmytryk and Walter Murch are two such theorists. Between them, they have edited 46 feature length films, television films or shorts. Dmytryk went on to direct 55 features, including Crossfire (Dmytryk, 1954), which was nominated for an Academy Award. Murch has won three Academy Awards, one for Best Editing The English Patient (Minghella, 1996) and two for Best Sound, The English Patient and Apocalypse Now (Coppola, 1979). Their immersion in the practice of editing gives their writings credence, and they focus on specific practices gleaned from their years in the profession.

Edward Dmytryk

Born in Canada, the son of immigrants from Ukraine, Edward moved with his father to Los Angeles after his mother died. His father beat him, and Edward frequently ran away from home, so the local juvenile authorities found him a job working as a messenger at Famous Players-Lasky and allowed him to live on his own by age 15. He later worked as a projectionist and moved to the editorial department in 1929. He edited a number of unremarkable features, but by the late forties had moved to directing significant features like The Caine Mutiny (Dmytryk, 1949). He was a leftist and joined the Communist Party U.S.A. during World War II, and was later called before the House Un-American Activities Committee (HUAC). First refusing to cooperate, Dmytryk was jailed as one of the “Hollywood 10,” but after spending a few months in jail, reversed his decision, went before HUAC and identified people he said were communists. Dmytryk’s action changed the course of his career, as many in Hollywood never forgave him for “naming names.” He moved to an academic position at University of Texas Austin from 1976–1981, teaching film theory and filmmaking, and later was the chair of the filmmaking program at USC from 1981– 1997. He authored several books on film, including On film editing: an introduction to the art of film construction (Boston: Focal Press, 1984).

Given his career track, it is not surprising that On film editing is a very practical work, and offers insight on industry practices of his era, like the differences of cutting film on Moviolas versus flatbed editors. Dmytryk talks about the special relationship that develops between director and editor, and how the editor should negotiate the editorial demands of talented directors and incompetent hacks alike. He writes about the need for smooth, invisible continuity cutting, and how directors work on set to provide the material to make this possible, often by using master takes, i.e., filming a single long take of an entire section of a script before moving in to film closer views. On the other hand, Dmytryk rails against “perhaps the greatest sinner of all. . . the ‘clever’ director who ‘cuts in camera.’ This phrase. . . usually signifies that the director, in any particular ‘take,’ shoots only a portion of the scene which he expects to use as one complete cut.”¹ Dmytryk views this approach as self-defeating, since it limits the range of solutions in the edit room and forces actors to work superficially, since they do not have sufficient time to get into the particular scene at hand.

Dmytryk argues that, “The usual theatrical films, excluding art films, film vérité, and so forth, are meant to appeal to the largest possible audience, and sound theories of filmmaking, including cutting, are based on this fact.”² Grounded in the film industry, his practical theorizing comes in the form of “editing rules” that will support the film’s audience appeal, and he offers his insight in the form of six general rules for editing.

Six Rules for The Ideal Cut

Rule 1: Never Make a Cut Without a Positive Reason

Before addressing Rule 1 directly, Dmytryk begins by asserting, as many editors have, that “the proper cut can only be made at a single point.”³ He goes further, stating that the issue is not one of duration of the shot on either side of that proper point. Dmytryk says that making a cut 3 or 4 or 5 frames on either side of that “single point” would be off (short or long) by 1/6, 1/8 and roughly 1/5 of a second respectively. He grants that since the normal blink of a human eye is 1/5 of a second,⁴ that duration is insignificant, and not one human beings would normally be bothered by. But to frame the problem differently, if someone asks “How can 5 frames, roughly 1/5th of a second make a difference in a cut, since that’s literally the blink of an eye?” Dmytryk’s answer looking at duration misses the point. He admits the durations involved are short, but that the issue is those frames will “spoil the match,” to use his precise term.

“Match” is an editing term that applies to many situations and, taken out of context, is virtually meaningless. In its broad sense, the term refers to a cut where there is duplication in the profilmic material of the outgoing shot A and the incoming shot B. Here, Dmytryk’s use of the term “match” refers to continuing an orderly sense of space and time over the cut, so that the flow of a story is not disrupted by the removal of slices of time or character positions that change abruptly. Although “match” can refer to other things like the graphic qualities of two shots, Dmytryk is using it here as it normally applies in narrative Hollywood films: that is, “match” is short for “match on action,” a term that means that the incoming shot B continues an action begun in shot A. And though he doesn’t specify further what “spoils the match,” the continuity system he embraces assumes that when we cut on a character’s action, the position, velocity and screen direction, etc. will continue from one shot to the next. So his point here is that, given an outgoing shot A and an incoming shot B, there is a proper cut point for the smoothest match. Therefore to focus on the fact that shifting the cut point 3, 4 or 5 frames is an insignificant change in duration misses the point that one proper point carries the action across the cut better – it creates the illusion of matching action better than other edit points do. The problem for the beginning editor is trying to uncover what the proper edit point is. Some of the rules below will help in suggesting where that point is, but often the process involves simply working with a range of frame choices and using trial and error to find the proper point. That search can be time consuming as the editor shifts out points and in points on both sides of the edit and previews the edit, looking for the right cut point. And yet to paraphrase Justice Potter Stewart’s famous saying about pornography – “I know it when I see it” – most accomplished editors know the smoothest match when they find it, it simply feels right visually.

For Dmytryk, the thrust of Rule 1 is simple: don’t cut just to cut, don’t change the shot because the current shot is running long. The editor should change the shot only if the change improves the scene – there must be a positive reason for the shot change. He summarizes, “As long as the scene is playing at its best in the selected angle, leave it alone!”⁵ In the contemporary media landscape, this advice is widely ignored, but the editor who seeks to clarify their craft practice would do well to consider this dictum, if for no other reason that it forces one to identify what is the specific reason behind a particular cut?

Rule 2: When Undecided About the Exact Frame to Cut On, Cut Longer Rather than Shorter

Dmytryk argues that the editor should always follow their instinct. “The rule that applies to school examinations also applies, logically enough, in cutting: the first immediate and instinctive cut is more likely than not to be the right one.”⁶ However, he argues that, other things being equal, if an editor is fretting or unsure about a particular edit point, it is better to leave the shot long than to cut it too short.

Here, Dmytryk’s advice is particularly apropos for the era in which he worked. When working in film, editors cut and spliced work print – an inexpensive print from the original camera negatives – so that the originals were protected and preserved to be the basis for fabricating a significant number of release prints that would be needed for projection in theaters. When the editing of the work print was complete, the work print was used as a precise map for the same edits made to conform the camera negative to the work print version, much like an offline video edit generates an edit decision list that is used to conform an online version of the video. Consequently, when working with work print using the slow process of physically cutting and tape splicing each edit, it was more efficient to be cautious, and leave a shot long rather than to cut it short, when later changes might require the editor to laboriously re-insert and splice three or four frames back into a work print.

Nevertheless, the advice is still relevant in the digital age, particularly for novice editors. If a cut is bothersome, rather than revising it endlessly trying to find the right solution, leave the edit long. By letting the edit “breathe” and be more “open” in the early stages, the editor can move forward, knowing that refinement and revision will polish the cut. This situation parallels writing. Working through a rough draft of a written document, it is often more important at the early stages to get the material out on the page from beginning to end, rather than to focus too much on specific issues like tense, word choice or sentence structure. Just as in writing, taking an edit through a number of “drafts” that ultimately leads to “picture lock,” the editor should move forward to assemble the scene leaving shots long, and assuming that subsequent cuts will clarify, simplify and resolve those passages that are difficult in the first pass.⁷

Rule 3: Whenever Possible, Cut "In Movement"

Rule 3 contains some of the most valuable practical guidance that Dmytryk offers beginning editors, and most of that advice relates to the “cut on action,” or as he calls it the cut “in movement.” Dmytryk points out that cutting on action is the norm, but that there are some cuts in a film that do not involve looking for a movement to cut on: the beginning or ends of sequences, self-contained shots and dialogue or action that do not require an insert edit. Beyond those, the editor is generally looking for some character movement to motivate a cut, and “A broad action will offer the easier cut, but even a slight movement of some part of the player’s body can serve to initiate a cut which will be ‘smooth’ or invisible.”⁸ He gives as an example a wide shot and a close shot of a character sitting down in a chair, noting that the editor’s preferred choice would be to find a point in action of “sitting down” to make the edit, rather than allowing the action to take place fully in either the wide or the close shot.

For the action of sitting down – really for any action – the question for the editor remains, “Where in the action to cut?” For Dmytryk, the answer hinges on a hypothesis: he asserts that those moments where we can motivate the viewer’s eyes to move or blink will provide the smoothest cut. Dmytryk gives as an example a typical case of where screen direction is maintained across a cut: the action of a character exiting frame left and entering frame right (Figure 3.1).

Figure 3.1 An exit/entrance cut According to Edward Dmytryk, the exit/entrance cut is an example of where the editor can motivate the viewer’s eyes to move to provide the smoothest cut. Once the eyes of the actor exit frame left, the eyes of the viewer habitually move to the center of the screen. Then the entrance of the actor in the incoming cut continues to carry the viewer’s eyes all the way to the right side of the screen.

He argues that once the eyes of the actor exit frame left, the eyes of the viewer begin to

Dmytryk is correct that when the eyes move they are “unfocused” as they move, but not literally correct. Research on reading has established that when we read, our eyes move intermittently, stopping to fixate on a word or words to bring new information to the brain, and then moving in short, rapid movements called saccades, during which vision is suppressed, so that no new information is taken in.¹⁰

Dmytryk’s theory seems intuitively correct. Presumably, having the actor re-enter frame right does assist the viewer’s eye reverting towards the middle of the screen, since one central function of our visual system is to detect change in the visual field. Dmytryk goes further to suggest that anytime we can cut within a period that the viewer’s eyes are distracted, that distraction tends to smooth the cut. One such trick that is commonly used is the editorial ploy to cut on a sharp sound in the track, such as an explosion or door slam, on the assumption that the viewer will instinctively blink, distracting the eye and smoothing the cut. (As we will see in the next section, Walter Murch is another editor who is greatly interested in the correlation of “the blink” and the cut.)

In the two cases under consideration here – the character sitting down and the character exiting and re-entering – Dmytryk adds one final consideration for smoothly cutting on action: the action overlap. Here we are not referring to the short jumps backwards in time to replay, say, the moment of impact of bullets on a windshield as in Mission Impossible 2 (See Chapter 2, p. 60). Rather, Dmytryk (like many editors) is convinced that whenever a shot changes, it takes three to five frames for the viewer’s eyesight to register the change. Or to put it differently, every time a shot changes, the first frame that actually registers in the viewer’s eyesight is frame 3 or 4 or 5 of the incoming shot. So if the goal is to match a person sitting in their chair, the editor should not search for the exact spatial/position match between the last frame of the outgoing shot and the first frame of the incoming shot. Instead, the edit point for the incoming shot should commence, say, five frames before the exact spatial match: that is, the last five frames of the outgoing is overlapped by five frames of the incoming shot. Dmytryk theorizes that those five frames will not visually register with the viewer, but frame 6 of the incoming shot will, and frame 6 is the exact spatial match of the last outgoing frame (Figure 3.2).

It is difficult to determine with any precision how widely this practice has been adopted. However a study published in 2014 by Shimamura et al., “Perceiving Movement Across Film Edits: A Psycho Cinematic Analysis” asked viewers (n = 48) to judge smooth movement across such edits, and found

These findings held even when a pattern mask was placed between the shots. Dmytryk’s notion that film edits disturb the viewer’s cognitive processing such that they must be given a short repetition of the action on the incoming shot in order to perceive smooth movement is supported by this research.

According to Dmytryk, what is equally important to produce a smooth cut from one shot of a character walking to another shot of that character walking is maintaining the rhythmic cadence of the movement. Dmytryk says

Figure 3.2 Overlapping cut for smooth continuity. An exact spatial match from the outgoing shot to the incoming shot often produces a hard cut. That’s because when the shot changes, the viewer does not really “see” the first 3–5 frames of the incoming shot. Adding “overlap frames” to the head of the incoming shot – C’, D’, E’ frames that repeat some of the action in the outgoing shot – creates the perception in the viewer that the cut was a spatial match from F to G.

Presumably, this cadence is maintained even as the editor factors in the “overlap” frames that are visually suppressed while the eyes travel back across the screen: what is important is that the hit on the ground appears to be smooth rather than be a technically correct spatial match.

Perhaps Dmytryk’s “3 to 5 frame overlap” is too long for some cuts, suggesting a certain level of imprecision or “wiggle room” in his guidelines about smooth matching. By contrast, the Shimamura findings were quite specific: “When asked to select the clip with the smoothest movement, participants chose clips that contained a brief overlap in action (125ms).”¹³ Less than the blink of an eye, 125ms is .125 seconds, or 2 frames in a 24-frame time base. Perhaps Dmytryk’s ’3 to 5’ frames is too long for some cuts, but as noted above, trial and error in shifting outgoing and incoming edit points will often uncover the smoothest match point.

Rule 4: The "Fresh" is Preferable to the "Stale"

Dmytryk says that many professional and almost all beginning editors handle the “exit-entrance” cut by letting the character exit the frame in the outgoing shot, holding the empty frame for a beat, then cutting to the next shot. This is not the best solution, according to Dmytryk, for three reasons: it misses the opportunity to use the natural eye movement to the center of the frame, it makes the scene (and ultimately the film) unnecessarily long, and most important, “it extends a scene which no longer has any meaning or interest for the viewer.”¹⁴ In short, the thrust of Rule 4 is don’t let the ends of outgoing shots run long, since, compared to incoming shots, which represent “fresh” information, outgoing shots are “stale,” their interest for the viewer is in decline. Consequently a better place to let a shot run long is the incoming shot.

Dmytryk asks us to consider a stagecoach moving through a gorgeous John Ford-like painted desert. If on an outgoing shot, the stagecoach clears the frame and the editor leaves the desert on screen, the beauty of the scenery begins to take the viewer’s interest, and they begin to think, “What an incredible landscape,” thereby taking them out of the film narrative. By contrast,

Dmytryk says the principle of placing extra footage in the “fresh” or incoming shot holds whether we are talking about footage that runs a few seconds or a few frames. Holding a scene after a character has exited is leaving the viewer with “cold coffee,” but even a few extra frames at the beginning of the incoming shot present the viewer with a new location to assimilate into the narrative.

Rule 5: All Scenes Should Begin and End with Continuing Action

Dmytryk argues that while the principle invoked in Rule 5 is well known, directors often overlook it. He is using the term “continuing action” here to mean footage with the actor already in character and acting, whether the script calls for the character to be sleeping in bed or dancing across the room. When opening a scene showing a character, the editor needs the character to be acting, not “preparing to act” and as the scene is ending, the editor needs the actor not to immediately “let down” or come out of character. The problem for the editor is being able to find a natural cut point at the beginning or end of the take where the actor is “alive,” and in character. When there is no extra footage at the beginning or end of a take that allows the actor to be in character, the editor’s job is made more difficult, because it can be hard to find a natural cut point, particularly if the scene requires “handles” to dissolve to or from the scene. Dmytryk suggests that there are few fixes when working with footage that does not support continuing action to begin or end a scene. When beginning a scene, one option is just to cut into the body of the scene, even if it means making a “j-cut:” a cut in which the sound of the incoming shot precedes the picture; an edit that “pre-laps” the audio of the incoming shot into the outgoing scene. When ending a scene where an actor “lets down” too soon, Dmytryk proposes cutting away to another character’s reaction, or to an object within the scene, or making an “l-cut:” a cut in which the sound of the outgoing shot continues after the picture ends; an edit that “post-laps” the audio of the outgoing shot into the incoming shot. The l-cut is not an ideal solution, but Dmytryk suggests it is better than ending the scene (picture and audio) before the “let down,” which often leaves the scene feeling short and chopped.

Rule 6: Cut for Proper Values Rather Than for Proper "Matches"

When cutting a dramatic film, Dmytryk argues that the first consideration is to go to the shot that carries the dramatic thrust of the scene, rather than one that is technically a better match cut. Why should the editor follow this rule? First, finding a technically better match point may move the edit point earlier or later to a point where it is illogical, and therefore feels like a “jump” or a hard cut in spite of the technical match. Second, such a match will damage the dramatic substance of the scene for the same reason: it feels illogical and forced.

Dmytryk says that in a situation where the dramatic needs of a scene force the editor to cut from a wide shot A to a closer shot B that contains a mismatch, the editor should first try the cut leaving the mismatch, and try to identify where within the frame the average viewer is looking from A to B. More than likely, if the viewer is looking at the actor’s face and eyes, then mismatches in other areas – the position of the hands or arms that seem so prominent to the editor – will probably not be seen by the audience. The editor might next consider cutting in to another closer shot where only the face will carry over from shot to shot, essentially “framing out” the mismatch in the incoming shot. If there is no such closer shot, the editor can consider enlarging the incoming shot so that the mismatch material is out of frame or partially out of frame. Dmytryk did these kinds of frame enlargements in the age of film, when blowing up the frame by reshooting the footage in an optical printer always added grain to the shot that in many cases was quite visible. When enlarging digitally, the process is much easier technically as it does not require another generation of the footage. Yet, the editor faces similar problems with “digital grain” or pixilation, where pixels are displayed at such a large size that individual pixels become visible. When the image is less than 110%, such issues are mitigated. And when we shoot in a larger resolution, like 4k video, and distribute in 1080 HD, we can often just choose what part of the 4k frame we want to work with, because the larger frame size allows us to reframe in postproduction without “loss of resolution.” One final work-around would be to cut to a closer shot of the person shown in shot A, before we go to shot B, thereby hiding the mismatch because the new outgoing shot is tight enough to handle the problem.

Dmytryk ends his practical theorizing with a forceful piece of advice that sums up how editors need to approach their work:

Walter Murch: Six Criteria for the Ideal Cut

Director Francis Ford Coppola calls Walter Murch “a study unto himself: a philosopher and a theoretician of film – a gifted director in his own right.”¹⁷ Close collaborators for over 40 years, Murch worked with Coppola on a number of films including The Conversation (1974) (sound montage), Apocalypse Now (1979) (co-editor/sound designer), The Godfather (1972) (post production consultant), The Godfather: Part II (sound montage), and The Godfather: Part III (1990) (editor). Murch won his first Oscars for sound montage in The Conversation (1974) and he won two Oscars for the same film: The English Patient (Anthony Minghella, 1996) for both sound mixing and editing.

Murch’s book on editing, In the blink of an eye: a perspective on film editing (1995) grew out of a lecture Murch gave in October 1988 in Sydney, Australia that was sponsored by the Australian Film Commission. As the title suggests, it is Murch’s singular perspective on the craft he has worked in for most of his life, and like Dmytryk’s work it is practical and centered on editing issues arising in narrative features. Working in the industry, Murch, like Dmytryk, experienced firsthand the dominance of Hollywood continuity as a controlling principle. Murch calls this principle “three-dimensional continuity,” and describes the magnitude of its reach:

Murch argues that while “three-dimensional continuity” – matching the spatial arrangement of the actors in the filmic space from shot to shot – is taught in film schools, it is actually the least important consideration in editing. Rather, the primary consideration for the editor to consider is “How do you want the audience to feel?” and this principle is seldom taught because it is the hardest to identify and to deal with.¹⁹ Murch argues that audiences will remember how a film makes them feel, rather than the style of the film or even the story. Consequently, if the editor has made the audience feel what they should feel throughout the film, the goal of the edit has been achieved. Notice that this dictum matches Dmytryk’s Rule 6: Cut for Proper Values Rather Than for Proper “Matches.”

If the editor is confronted with making a cut, what criteria does Murch suggest for evaluating the best possible choice for that cut? Murch, drawing on his decades of professional editing experience, offers “6 Criteria for an Ideal Cut” as his decision making guide. A cut is “ideal” for Murch if it satisfies all of the following at once:

1. It is true to the emotion of the moment.
2. It advances the story.
3. It occurs at a moment that is rhythmically interesting and “right”.
4. It acknowledges what you might call “eye-trace” – the concern with the location and movement of the audience’s focus of interest within the frame.
5. It respects “planarity” – the grammar of three dimensions transposed by photography to two (the questions of stage-line, [i.e., the 180° line] etc.).
6. It respects the three-dimensional continuity of the actual space (where people are in the room and in relation to one another).²⁰

Six criteria is a lot to ask of a single cut: editors routinely make cuts with less at stake, particularly in non-dramatic films, but Murch’s point is well taken. He is addressing the primary considerations that an editor working in fiction narrative would likely encounter, but certainly these factors will often be present in non-narrative works. And more important, these criteria are ranked: other things being equal, the first consideration for the editor should be “emotion of the moment,” the second consideration should be “advancing the story,” etc. Murch’s criteria are designed to upend the beginning editor’s attachment to cuts that “match,” which he places last, lower even than “questions of stage-line,” his term for the 180° line. In fact, Murch goes on to give a tongue-in-cheek “weight” to each of these criteria, a percentage they should occupy in the decision making process:

1. Emotion – 51%
2. Story – 23%
3. Rhythm – 10%
4. Eye-trace – 7%
5. Two-dimensional plane of screen [i.e., the 180° rule] – 5%
6. Three-dimensional space of action – 4%

Here then is a specific and useful decision making process for the editor to consider, offering six criteria to consider simultaneously in reference to a single cut, with the aim of identifying the best alternative, based on Murch’s values and years of tacit knowledge he has gained from his professional life’s work. Given all of the potential factors we might consider in making a single cut, identifying only six factors eliminates the potential for “information overload,” and to some extent, “analysis paralysis,” where the editor begins to over analyze the situation to the point of inaction. Notice too that the Murch decision process aims high: the goal is the “ideal cut,” not just one that is satisfactory.

Murch and Blink Theory

Perhaps Murch’s best known theory revolves around the observation that when a person blinks, it is not simply a point at which they need to moisten their eyeballs, because if that were true, people would blink at a constant rate in a given environment. Rather, Murch believes that blinking is a good indicator of internal emotions and the frequency or intensity of internal thoughts. And although Murch never uses the word “gestalt,” his understanding of visual thinking is grounded in that model of perception. Murch asserts that blinking is a fundamental method of making sense of the world, of reducing the chaotic world of the sensory data that reality confronts us with into useable “chunks:”

In other words, shot changes in film mimic the human blink. Murch acknowledges the original inspiration for his blink theory the Hollywood director John Huston, subject of a 1973 interview in The Christian Science Monitor,

Murch takes this idea and expands it, by observing people blinking, an activity that he has unique access to as a film editor looking at recorded takes. He notes that there is a large amount of variation in the frequency of a person’s blinking, from minutes between blinks, to seconds or fractions of a second between blinks. Given this wide variation, Murch rightfully asks, “What is it that makes people blink?” Murch says the blink either facilitates a moment of separation between thoughts or marks that moment with an involuntary reflex, and he points to psychological research from 1987 that supports that conclusion.²⁴ More recent research that further supports Murch’s theory placed subjects (n=10) in a functional MRI while watching episodes of the British television show Mr. Bean. This 2013 study concluded “eyeblinks are actively involved in the process of attentional disengagement during a cognitive behavior by momentarily activating the default-mode network while deactivating the dorsal attention network.”²⁵ That is, blinking marks the point where viewers “take a breather” during periods of focused attention. When we blink we move from a state of attention (i.e., the dorsal attention network), to a moment of “wakeful rest” (i.e., the default-mode network). Contemporary Hollywood editors seem to support this position. Edgar Burcksen, A.C.E writes,

So how can this information be useful to an editor? Murch claims that it can be used two ways. First, when observing a conversation we can look at the listener’s blinks as the precise moment where the thought being expressed by the speaker has been completely understood, “And that blink will occur where a cut could have happened, had the conversation been filmed. Not a frame earlier or later.”²⁷

For Murch, “listener blinks” suggest potential cut points in a dialogue footage, because they mark for the editor when that listener – and presumably the audience as “listeners” – has completely understood the thought behind a section of dialogue. If the blink marks where we “take a breather” during periods of focused attention, then it is a good place for an editor to look when seeking a logical cut point that matches the “mental rhythm” of the filmed conversation.

Second, as we noted earlier, the cut is a medium dependent technique that mimics the human blink, providing a visual separation that punctuates a film text. For the editor, the cut is a way of concluding one idea and starting another. When properly placed, the cut cleaves a film text in two – bringing one idea to an end and starting another, and “the more extreme the visual discontinuity – from dark interior to bright exterior, for instance – the more thorough the effect of punctuation will be.”²⁸

Summarizing thus far, Murch’s argument is grounded in Huston’s notion that the experience of watching a film is “More like thought than anything else,”²⁹ and so the blink has two functions for Murch. In any take the editor considers, the blink suggests natural cut points to end one thought and begin the next. Placed well, those very cuts visually structure the film text with blink-like moments of “stimulus change.” But in practical terms, how does the editor bring this blink theory to bear on editing a film? Murch gives the specific example of Gene Hackman playing the paranoid surveillance technician Harry Caul in The Conversation (Francis Ford Coppolla, 1974) to answer this question. Murch says that because Hackman was so immersed in the character, his on camera thought patterns mirrored Harry’s, leading Hackman to blink in rhythm with Harry’s thoughts. Over time, as he worked with the material, Murch says he began absorbing the rhythms Hackman provided and “my cut points were naturally aligning themselves with his ‘blink points.’ In a sense, I had re-routed my neural circuitry so that the semi-involuntary command to blink caused me instead to hit the stop button on the editing machine.”³⁰ In other words, Murch is picking his out point on the fly, based on where he thinks he’s about to blink. He argues that a good editor must sensitize him or herself to the rhythms that the actor provides and then find

Urging the editor to get “in sync” with an actor’s performance and try to imbue the entire film with the actor’s thought patterns and pace is solid artistic advice. And while trying to “notice unconsciously” where an actor blinks may on the surface be self contradictory advice to follow, Murch firmly believes that observing the blink can inform cutting that accentuates and amplifies an actor’s performance rhythm.

Conclusion: The Tradition of Dymtryk and Murch

Today, editors work in an environment of rapid change brought on by laptop technology and software like Final Cut Pro – first introduced at the National Association of Broadcasters in 1999. This technology has democratized the editing process. Many film schools and media programs large and small now teach the basics of editing without much reference to Hollywood industry traditions, to the hierarchy of the guild system, to the older system of mentorship and on the job training. In the current phase of internet development, most of the worldwide community of editors at all levels are learning the craft outside “the industry,” using resources and the aggregate knowledge provided by software instruction sites like Lynda.com, educational sites like Moviola. com and peer to peer problem solving via the web.

Dmytryk and Murch are editors cut from a different cloth, both grounded in Hollywood feature editing. Their professional work lives are bound to the routines and practices of an industry that developed the continuity style of editing as a “collective invention,”³² to use David Bordwell’s term. Before film schools began their proliferation in the 1960s, the studio system had no formal system of instruction. Aspiring editors like Dmytryk were moved from apprenticeship to assistant editor to editor, using professional mentorship to develop skills and vision, tied to the experience of editors and directors who came before him. Murch, who came through the U.S.C. graduate film program in the late 1960s, represents a different era. In the 1960s and 70s, film school graduates were challenging what remained of the old studio system, transforming the old methods of finance, production and narrative forms into an “American New Wave” where the director drove the filmmaking process, and the studio’s role was greatly diminished. While the films Murch edited may have been more unconventional in subject matter and narrative logic than Dmytryk’s, both largely follow classical Hollywood norms.

As professionals within the industry, both of these editors are invested in a system of continuity editing that simultaneously is a restricted form of narration with enduring rules and conventions and an immensely rich and creative means of expression. Despite the fact that they worked in different eras, both are immersed in the standard style of the Hollywood feature, for which continuity editing’s clear articulation of space and time is a primary but “invisible” trait. The root of the tree that Dmytryk and Murch share is

The editorial wisdom imparted by Dymtryk’s and Murch’s simple rules grows from that lineage, editing as an incredibly effective narrative tool learned from countless iterations of cutting with a splicer, countless hours of replaying cuts on moviolas and flatbed editors within the studio walls. Their insights demonstrate why Hollywood editors remain key in a system whose goal is to unfold a coherent filmic text, constructed systematically with specific devices to creatively engage the audience in story construction while remaining more or less invisible. As we will see in the next chapter, David Bordwell’s work on how this larger system of narration functions reveals how fundamental editing is to the process of filmic narration.