![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
alibi_shopOn Directing Film, by David Mamet (1991)
It makes sense for cartoonists to read On Directing Film; at its core it's about visual storytelling, conveying meaning through juxtaposition, and interpreting a text that was written to be performed or illustrated rather than read as-is. Many comics, like most plays and movies, start with a script which is then interpreted by an artist or artists; the artist is in effect directing, and also acting. I say "many" and "most" because a comic, play, or movie can be a solo effort and/or unscripted; but even if the writer and illustrator/director/performer are the same person, visual storytelling can still be broken down as a collaborative process between the story-generating point of view and the story-rendering point of view. Mamet's main concern is how to communicate with an audience on an intuitive level.
I want to say right away that the book is often annoying and badly reasoned, but I don't want to waste too much time on that, so just to get it out of the way briefly: Mamet, at least the Mamet on display in this book, is not a good critic nor a good teacher. He's dogmatic (constantly repeating the rule that images should be "uninflected" without explaining why that should be so important, or even really defining the word); narrow-minded (he's only interested in one kind of movie: the plot-driven kind); arrogant (rather than just dismissing most other directors' work as being dramatically ineffective, he imputes character weaknesses to them, claiming that if someone tries to make their work aesthetically "interesting" it must mean they want people to like them, which is a no-no in his somewhat macho worldview); and even dishonest (he uses selective quotation to make it sound like Stanislavsky didn't approve of using any psychology in acting, which is hilariously untrue and not something Mamet could plausibly believe given his theater background). He's a crank. But cranks can still have interesting things to say, especially if you look at what they enjoy and appreciate more than what they hate and dismiss.
The first exercise in Mamet's lecture transcripts, discussing how to establish a particular scene through a series of shots, is the most straightforward part in terms of a purely technical approach to directing or illustrating— less about what the artist must do, more about what can be done. In one passage, the question is simply how to establish without tiresome exposition that a character is arriving early for a class. What does "arriving early" look like and how will the viewer understand it? The process of Mamet's students suggesting various clunky strategies and clinging to their first ideas ("Maybe a small clock?") is very familiar to me from my own creative efforts, and the principle of focusing on a very small set of core ideas in a scene is one I constantly struggle to remember.
A recurring theme throughout the book is that stories should be told in the transitions between images, rather than within the images themselves, whenever possible. This has an obvious resonance with comics theory as popularized by Scott McCloud, where the gutter between panels is thought of as the most important mechanism for harnessing the audience's imagination to bring a story to life. I think it's important to keep in mind that Mamet's claims here are 1. entirely focused on how this technique can be effective, rather than backed up by any evidence that other techniques are not effective (he simply asserts that if you try to show a complete action within a shot, you are "following the protagonist around," which he says is inherently bad); and 2. limited to the kinds of transitions that McCloud would call "action-to-action," "subject-to-subject," and "scene-to-scene," with a specific emphasis on driving a plot(*). But deep down the rationale is similar to McCloud's and makes sense to me: any human audience has a built-in desire for meaning, and an unconscious skill for creating meaning to fill any gap.
(* In an example of people enthusiastically trying to apply Mamet to comics in ways that he surely would not approve of, the critic Derik Badman decided that this was really about "aspect-to-aspect" transitions that establish a mood or a place through montage of outside-of-time images— even though Mamet specifically rejects the value of establishing a mood or a place. On the other hand, Tom Hart is able to appreciate Mamet's skill while acknowledging that his exclusive interest in "story" is not applicable to many kinds of art.)
Despite his desire for a literal-minded approach to directing and acting, Mamet respects and even loves the irrational soul of the viewer. He describes every film, no matter how realistic, as a dreamlike experience(**), and praises the role of the subconscious in artistic decisions. And in a rare gesture of humility, he urges directors not to underestimate the power of the viewer to generate unintended meaning: if you leave a gap that you didn't realize you left, or provide a confusing transition whose rationale you weren't sure of, the viewer won't just wait for an explanation, but will effortlessly fill that gap with something that may not work with what you were trying to do. In comics, this applies not only to panel-to-panel transitions but to composition in general, since the process of reading even a single panel involves many subtle hints about how to order the various parts of this still image into a miniature story. I also appreciate his remark about visual prop details being an "antiseptic" consideration: that is, if some part of the image is not supposed to communicate anything special, you need to choose its appearance carefully enough that the viewer won't get unwanted connotations from it. And another resonance with McCloud (though Mamet gets it from Bruno Bettelheim) is the idea that for primal and unconscious reasons an audience will identify most with a protagonist who's depicted in a minimal style, with little realistic characterization; I think that on this point both Mamet and McCloud are overgeneralizing and ignoring the many kinds of story that don't rely on audience identification, but it's still true that an artist should respect the power of a blank space and choose details carefully rather than piling them on in a quest for maximum realism.
(** Unsurprisingly given his eagerness to generalize about the human condition, Mamet goes too far in asserting that actual dreams are basically the same as Mamet movies, composed of "uninflected" images that only acquire meaning through transitions. In my own experience at least, dreams are chock full of images that carry an instant emotional charge, and "facts" that the dreamer just happens to know without being told or shown; and to me, this is why most attempts to depict dreams in movies— or in comics that use a purely cinematic approach— aren't convincing.)
Mamet wants screenwriters to stop writing "material that cannot be filmed," by which he means anything that doesn't have a one-to-one correspondence with on-screen actions. I agree that if I'm writing something to be performed or illustrated, I need to be aware that the audience can't read my mind, so anything that really is important will have to be acted on or represented somehow, beyond just existing in the script; and adopting a general principle of cutting everything to the bone forces me to think harder about which things I think are the important ones, rather than deferring all of those decisions to the director or illustrator. However, I think "material that cannot be filmed," as long as the writer understands that that's what it is, can still be valid and fruitful as a form of communication between creators (or within a single creator). Alan Moore's scripts are 90% composed of such material, but because Moore understands his collaborators, his conversational approach with a deliberate over-supply of background details produces comics that accurately represent his vision of the story— and I think that result would be less likely if he just told the artists exactly what to draw without saying why. Offering the director/illustrator a bundle of details, like a bouquet or a toolbox, with the understanding that only one or two of these might be needed, is a way of expressing the desired goal while still trusting in the collaborator's specialized expertise in selecting the best flower or the best tool for the job.
As a possible clue that Mamet does understand that the creative process can't always be reduced to mechanical steps and that a minimal text may not be enough, during a lecture on shot breakdowns he slips seamlessly from examining how to direct a specific script to asking the students to rewrite the script, introducing totally new goals for the characters so as to make the director's job easier. As a teaching strategy this is too much like rhetorical sleight of hand for me— leading the students to believe that his dogmatic rules have solved the problem, when actually he has changed the subject— but I can't argue with the overall lesson: if you're finding it very hard to convey the idea of a scene clearly, it might mean that you need a better idea.
It makes sense for cartoonists to read On Directing Film; at its core it's about visual storytelling, conveying meaning through juxtaposition, and interpreting a text that was written to be performed or illustrated rather than read as-is. Many comics, like most plays and movies, start with a script which is then interpreted by an artist or artists; the artist is in effect directing, and also acting. I say "many" and "most" because a comic, play, or movie can be a solo effort and/or unscripted; but even if the writer and illustrator/director/performer are the same person, visual storytelling can still be broken down as a collaborative process between the story-generating point of view and the story-rendering point of view. Mamet's main concern is how to communicate with an audience on an intuitive level.
I want to say right away that the book is often annoying and badly reasoned, but I don't want to waste too much time on that, so just to get it out of the way briefly: Mamet, at least the Mamet on display in this book, is not a good critic nor a good teacher. He's dogmatic (constantly repeating the rule that images should be "uninflected" without explaining why that should be so important, or even really defining the word); narrow-minded (he's only interested in one kind of movie: the plot-driven kind); arrogant (rather than just dismissing most other directors' work as being dramatically ineffective, he imputes character weaknesses to them, claiming that if someone tries to make their work aesthetically "interesting" it must mean they want people to like them, which is a no-no in his somewhat macho worldview); and even dishonest (he uses selective quotation to make it sound like Stanislavsky didn't approve of using any psychology in acting, which is hilariously untrue and not something Mamet could plausibly believe given his theater background). He's a crank. But cranks can still have interesting things to say, especially if you look at what they enjoy and appreciate more than what they hate and dismiss.
The first exercise in Mamet's lecture transcripts, discussing how to establish a particular scene through a series of shots, is the most straightforward part in terms of a purely technical approach to directing or illustrating— less about what the artist must do, more about what can be done. In one passage, the question is simply how to establish without tiresome exposition that a character is arriving early for a class. What does "arriving early" look like and how will the viewer understand it? The process of Mamet's students suggesting various clunky strategies and clinging to their first ideas ("Maybe a small clock?") is very familiar to me from my own creative efforts, and the principle of focusing on a very small set of core ideas in a scene is one I constantly struggle to remember.
A recurring theme throughout the book is that stories should be told in the transitions between images, rather than within the images themselves, whenever possible. This has an obvious resonance with comics theory as popularized by Scott McCloud, where the gutter between panels is thought of as the most important mechanism for harnessing the audience's imagination to bring a story to life. I think it's important to keep in mind that Mamet's claims here are 1. entirely focused on how this technique can be effective, rather than backed up by any evidence that other techniques are not effective (he simply asserts that if you try to show a complete action within a shot, you are "following the protagonist around," which he says is inherently bad); and 2. limited to the kinds of transitions that McCloud would call "action-to-action," "subject-to-subject," and "scene-to-scene," with a specific emphasis on driving a plot(*). But deep down the rationale is similar to McCloud's and makes sense to me: any human audience has a built-in desire for meaning, and an unconscious skill for creating meaning to fill any gap.
(* In an example of people enthusiastically trying to apply Mamet to comics in ways that he surely would not approve of, the critic Derik Badman decided that this was really about "aspect-to-aspect" transitions that establish a mood or a place through montage of outside-of-time images— even though Mamet specifically rejects the value of establishing a mood or a place. On the other hand, Tom Hart is able to appreciate Mamet's skill while acknowledging that his exclusive interest in "story" is not applicable to many kinds of art.)
Despite his desire for a literal-minded approach to directing and acting, Mamet respects and even loves the irrational soul of the viewer. He describes every film, no matter how realistic, as a dreamlike experience(**), and praises the role of the subconscious in artistic decisions. And in a rare gesture of humility, he urges directors not to underestimate the power of the viewer to generate unintended meaning: if you leave a gap that you didn't realize you left, or provide a confusing transition whose rationale you weren't sure of, the viewer won't just wait for an explanation, but will effortlessly fill that gap with something that may not work with what you were trying to do. In comics, this applies not only to panel-to-panel transitions but to composition in general, since the process of reading even a single panel involves many subtle hints about how to order the various parts of this still image into a miniature story. I also appreciate his remark about visual prop details being an "antiseptic" consideration: that is, if some part of the image is not supposed to communicate anything special, you need to choose its appearance carefully enough that the viewer won't get unwanted connotations from it. And another resonance with McCloud (though Mamet gets it from Bruno Bettelheim) is the idea that for primal and unconscious reasons an audience will identify most with a protagonist who's depicted in a minimal style, with little realistic characterization; I think that on this point both Mamet and McCloud are overgeneralizing and ignoring the many kinds of story that don't rely on audience identification, but it's still true that an artist should respect the power of a blank space and choose details carefully rather than piling them on in a quest for maximum realism.
(** Unsurprisingly given his eagerness to generalize about the human condition, Mamet goes too far in asserting that actual dreams are basically the same as Mamet movies, composed of "uninflected" images that only acquire meaning through transitions. In my own experience at least, dreams are chock full of images that carry an instant emotional charge, and "facts" that the dreamer just happens to know without being told or shown; and to me, this is why most attempts to depict dreams in movies— or in comics that use a purely cinematic approach— aren't convincing.)
Mamet wants screenwriters to stop writing "material that cannot be filmed," by which he means anything that doesn't have a one-to-one correspondence with on-screen actions. I agree that if I'm writing something to be performed or illustrated, I need to be aware that the audience can't read my mind, so anything that really is important will have to be acted on or represented somehow, beyond just existing in the script; and adopting a general principle of cutting everything to the bone forces me to think harder about which things I think are the important ones, rather than deferring all of those decisions to the director or illustrator. However, I think "material that cannot be filmed," as long as the writer understands that that's what it is, can still be valid and fruitful as a form of communication between creators (or within a single creator). Alan Moore's scripts are 90% composed of such material, but because Moore understands his collaborators, his conversational approach with a deliberate over-supply of background details produces comics that accurately represent his vision of the story— and I think that result would be less likely if he just told the artists exactly what to draw without saying why. Offering the director/illustrator a bundle of details, like a bouquet or a toolbox, with the understanding that only one or two of these might be needed, is a way of expressing the desired goal while still trusting in the collaborator's specialized expertise in selecting the best flower or the best tool for the job.
As a possible clue that Mamet does understand that the creative process can't always be reduced to mechanical steps and that a minimal text may not be enough, during a lecture on shot breakdowns he slips seamlessly from examining how to direct a specific script to asking the students to rewrite the script, introducing totally new goals for the characters so as to make the director's job easier. As a teaching strategy this is too much like rhetorical sleight of hand for me— leading the students to believe that his dogmatic rules have solved the problem, when actually he has changed the subject— but I can't argue with the overall lesson: if you're finding it very hard to convey the idea of a scene clearly, it might mean that you need a better idea.
