Aarhus University Seal

"The Collaborative Potential of Algorithmic Procedures (and the Algorithmic Potential of Collaboration)"

Ariel Avissar

Presentation Transcript

Hello everyone.

Today I’d like to speak about the generative and collaborative potential of algorithmic thinking in videographic criticism and pedagogy. By “algorithmic thinking” I mean, thinking about videographic creation as following an algorithm: a set of ordered steps required to accomplish a task, in a way that could be repeated by others. As part of this, I’ll distinguish between two types of algorithmic thinking: prescriptive and descriptive.

By “prescriptive” I mean predetermined algorithms, that is: a set of instructions or constraints given as a prompt for the creation of videographic works. They prescribe various formal, structural or technical constraints that dictate how videos should be constructed, leaving certain elements up to individual choice. Such prompts can consist of highly rigid parameters, or ones that are quite loose. At their best, these algorithmic prompts are highly effective in motivating videographic creation, their limitations and impositions serving as an impetus to start making: all you have to do is sit down, follow the algorithm and let inspiration follow.

I’ll name a few instances where such algorithmic prompts have been effectively used to encourage collaborative creation – and I imagine many of these will already be familiar to some of you; you can check out the links in the symposium website (and in the video description) for more information:

First, of course, the videographic exercises developed in Middlebury’s “Scholarship in Sound & Image” workshops (the pechakucha, the epigraph and others); these highly influential prompts have been successfully and ubiquitously employed, incorporated in workshops, in academic courses, and also as homework assignments for the Video Essay Podcast. I personally use them all the time in teaching videographic criticism to undergrads, to colleagues and to highschoolers, and have made many of them myself – as I’m sure many of us here have done as well. On a much smaller scale, I’d also mention the TV Dictionary, a collaborative videographic project I started last year, which invites participants to try and capture the essence of a particular TV series using a particular word and its dictionary definitions; and, as an example of a prompt that is far looser, the Once Upon a Screen collaborative video essay collection, started by Evelyn Kreutzer and myself, which invites participants to engage videographically with their childhood screen traumas. If any of these are new to you, you’re welcome to check them out.

Again, what these examples have in common is a pre-existing prompt that serves as a stimulant, an invitation to follow the algorithm and create. These kinds of prompts have proven very useful both for those looking to take their first steps in video essay creation, and for reinvigorating more experienced makers. The best and most productive prompts are the ones that provoke collaboration, encouraging many people to take part in the creation of an ever-expanding videographic archive.

For the remainder of this presentation, however, I’d like to focus on the other type of algorithmic thinking: the descriptive. In this case, rather than a pre-existing algorithm promoting collaboration, the opposite occurs, with videographic dialogue prompting algorithmic thinking, retroactively extracting an algorithm from a pre-existing video. Let’s examine a specific example (and again, there are links to all the videos I’ll be mentioning).

This past year, as part of an undergrad honors course I taught at the Steve Tisch School of Film and Television at Tel Aviv University, I gave my students a “videographic response” assignment: they had to choose an existing video essay, and make a video that responded to it in some way.

Here is one video that was chosen for response - LIQUID PERCEPTION by Catherine Grant. This, of course, is a videographic epigraph – combining images, sounds and on-screen text to a poetic and illuminating effect. If we wanted to describe its more specific functioning logic, as it were, it would go roughly like this: [onscreen text].

One of my students, Shira Havron, made this video in response. As you can see, Shira followed the basic format and style of Catherine’s video, with some variation: using the same text and the same music, but paired with images taken from a different film, and focusing on clips of people in space rather than underwater, thus highlighting insightful similarities between these two, non-earthly modes of being. If we wanted to consider these two videos together and extrapolate an underlying algorithm governing both, it would go something like this: [onscreen text].

Next, another student of mine, Ido Harambam, made his own response to Katie’s video. As you can see, Ido’s video once more focuses on underwater scenes, but this time set to a different music track, and, more crucially, examining not one but two films, affectively reframing the “two movements” mentioned in the quote as intertextual movements, a dialogue between two films that this piece suggestively merges together. In this case, the underlying algorithm would be: [onscreen text]. And if we were to describe the underlying algorithm of all three videos, it would be: [onscreen text]. These are only two examples out of many possible variations, each achieving a different result through its application of the underlying algorithm derived from Catherine’s video. I should note that in my description I’ve left out other potential parameters from that original video, namely the deceleration and colorization of the image – as these particular elements were not utilized by the videos my two students made. One can imagine other videos that do make use of those parameters, or that suggest new parameters. For me these examples are illuminating because they illustrate how one can pick and choose which parameters to activate: which elements in the original video should be treated as constants, and which should be treated as variables; which parameters to adhere by and which to mess around with. In this way, as opposed to prescriptive algorithms, many different, hypothetical algorithms can be derived from a single video.

So, instead of a pre-existing prompt leading to the creation of videos, we have an existing video leading to the creation of response videos; considered together, these encourage algorithmic thinking, which in turn can suggest hypothetical permutations, and lead to the creation of even more videos. This response-based creation has been very fruitful when working with students – and I myself have made several response videos of my own.

One of my favorite examples of this mode of thinking, which I will only mention briefly here, is “Payne’s Constraint” – a highly-parameterized set of constraints derived from Matt Payne’s video, “Who Ever Heard…?”. The name and parameters of “Payne’s Constraint” were described by Alan O’Leary in his reverse-engineering of Matt’s original video for the Notes on Videographic Criticism newsletter. This descriptive algorithm was then offered as a prompt that others can take up and use to create their own versions, each adhering to and deviating from the given parameters in different ways.

So, just as prescriptive algorithms can breed collaboration and encourage creation, so can existing works start a conversation, inviting response and analysis, and breed descriptive algorithms that then lead to further participation and more collaboration. Working with constraints and thinking in algorithms thus hold great generative potential, and can stimulate videographic creation and community, among peers, inside the classroom and elsewhere.

Thank you.