Thirteen Notes: A Poetics of Cinematic Randomization

By Nicholas Rombes

Download PDF

1. The digitization of films suggests a new way of thinking about a film as discrete images rather than as part of a flow.

2. “But delaying the image, extracting it from its narrative surroundings, also allows it to return to its context and to contribute something extra and unexpected, a deferred meaning, to the story’s narration.” [Laura Mulvey] (1)

3. It is not simply that films themselves can now be randomized (2) with ease, but that this potential randomization mirrors the decentered architecture of the Web.

4. What does a poetics of randomization—i.e., randomly selecting (or having a computer program randomly select) X number of images/frames from a film as the basis of inquiry—mean for interpretation? Is it possible (or desirable) to break just a little bit free from the traditions of film theory—grounded as they are in interpretation—and move towards a form of film writing that relies on chance?

5. “Every tradition has special ways of gaining followers.  […] Depending on the tradition adopted this way will look acceptable, laughable, rational, foolish, or will be pushed aside as ‘mere propaganda.’ Argument is propaganda for one observer, the essence of human discourse for another.” [Paul Feyerabend] (3)

6. What is the ideology of randomization when used as a critical tool? If it privileges disorder, this is only at the symbolic level, for randomization itself is made possible by highly ordered and functional technologies. It is already coded for something, but for what? It is more than a tool, as all tools are. Lev Manovich’s work with ImagePlot, (4) for instance, is a story not just about the development of a new set of tools, but also a story about the aesthetics of such software, which is beautiful and elegant in its visualization. 

7. What is the location of a film in the digital era? Netflix, for instance, streams its movies using content delivery network servers hosted by the company Level 3 which, in a section of their website entitled “Streaming,” (5) describe themselves this way:

Level 3 delivers a superior streaming media experience to your broadband users. We are uniquely positioned to support the rich-media delivery that users demand. Powered by one of the most connected content delivery networks (CDN) in the world, the Level 3 Streaming platform is designed to support the next generation of streaming for leading media formats. Using this network, combined with Level 3 Vyvx video broadcast backhaul services, we can be your single-source transport provider, from Content Creation to Consumption.

8. More than ever before, film has no body. Its presence is obscure and uncertain. The projection booth is haunted. The images we see, what is their physical origination point? No longer projected, they emanate, but from where? And yet, the disembodiment of cinema makes sense, since movies have always transported (“moved”) us to someplace else. 

9. Time-shifting—the act of recording a program to view at a later date, pioneered by Sony in the 1970s—created the foundations for randomization as a cultural prerogative.

10. In cinema studies, randomization and chance are the equivalent of pure research. 

11. “On the very last page of the postscript to The Varieties of Religious Experience, William James writes of people’s willingness to stake everything on the chance of salvation. Chance makes the difference, says James, between ‘a life of which the keynote is resignation and a life of which the keynote is hope.’” [Geoff Dyer] (6)

12. Before the increasing number of procedural, technical, and economic standardizations that accompanied the emergence of film as industry, the experience of the film spectator was subject to variables that resulted in sometimes drastically differing experiences of the same film. In some instances, the projectionist had a hand in editing the film through the selection of reel order. In Edwin S. Porter’s 1903 The Great Train Robbery, the placement of at least one shot—the bandit firing directly at the camera—has an element of chance or surprise depending on the local context of the film’s exhibition:

The spectators start out as railway passengers watching the passing countryside, but they are abruptly assaulted by a close-up of the outlaw Barnes firing his six-shooter directly into their midst. (This shot was shown either at the beginning or the end of the film. In a Hale’s Tours situation it would seem more effective at the beginning, in a vaudeville situation at the end as an apotheosis.) [Charles Musser] (7)

13. It may be more accurate to turn our thinking around and claim that it is not we who select this or that image or film to study, but rather than it is the film which selects us. “Randomization” thus refers not to the process of investigating the film in question, nor to the digital technologies that create the possibilities for chance-based retrieval of a film’s images, but rather to our own predilections, disguised as they may be within the discursive practices of academic inquiry. In other words, we are the ones selected—randomly it would seem—by the films we study.


(1) Laura Mulvey, Death 24x a Second (London: Reaktion Books, 2006), p. 151.

(2) My 10/40/70 project captures film frames from the 10, 40, and 70-minute points of films as the basis for criticism, in hopes that this constraint allows for an element of surprise. In addition, my year long Blue Velvet Project at Filmmaker Magazine and the ongoing DO NOT SCREEN project experiment with randomization and chance as potential methods of creative research.

(3) Paul Feyerabend, Against Method (London: Verso, 1993 [1975]), p. 226.

(4) ImagePlot is a free software tool that visualizes collections of images and video of any size. Online at:

(6) Geoff Dyer, Zona (New York: Pantheon Books, 2012), p. 93.

(7) Charles Musser, Before the Nikelodeon: Edwin S. Porter and the Edison Manufacturing Company (Berkeley: University of California Press, 1991), p. 264.


Frames # 1 Film and Moving Image Studies Re-Born Digital?, 2012-07-02, this article © Nicholas Rombes