Subroutines for Humans: Algorithmic Performance & Documentation
Studio Project – Computational & Performance Art
Overview
IMAGE PLACEHOLDER – Header / project image
In this project, you will create an artwork that exists in two layers:
- A procedural instruction-piece (an algorithm or “subroutine” for humans).
- A digital documentation work (how most people will know and interpret your piece).
You may work alone or in a small group (2–4 students). The computer must be involved not only as a tool, but as part of the structure of the work (through algorithms, platforms, or systems).
Epistemology, Performance, and Documentation
In conceptual art, a piece that provides instructions instead of an object raises epistemic questions: the work exists as an idea, and knowing the idea can become more important than seeing a finished form.
In performance art, when documentation replaces the live act, the audience’s knowledge of the work comes through photos, video, or text. The epistemic issue is that the work is known indirectly, through evidence rather than experience.
You are essentially making two works when you document performance:
- The live performance – like live music, it can only be experienced in real time, in a specific place or situation.
- The documentation artwork – this is how most people will encounter and understand your work afterwards.
Think ahead about documentation before you perform:
- Will someone be capturing it? Is that person part of the piece?
- Will you use photos? Photos with text? Video with text? A website? A PDF/book? A social media thread? A different structure?
Your task is to design both layers intentionally so that the documentation is not just evidence but a second artwork that may contain new meanings.
Inspiration: “20 Subroutines for Humans Made by a Computer”
IMAGE PLACEHOLDER – Screenshot/quote from “20 Subroutines for Humans Made by a Computer”
As a point of departure, look at this project:
20 Subroutines for humans made by a computer – Kazys Varnelis
Notice how the “subroutines” read like instructions written by a machine for human bodies and attention. They are procedural, repeatable, and open to interpretation. Use this as inspiration for your own human-readable “code.”
Phase 1 – Write Your Algorithm (Instruction Score)
IMAGE PLACEHOLDER – Sketch / notes for an instruction-piece
Write a short, repeatable procedure for humans that can be executed like code. This is your instruction-piece or subroutine.
Your instruction must:
- Be clear enough that someone else could follow it without you present.
- Include at least one constraint (time, place, materials, data, platform, etc.).
- Include at least one element of indeterminacy (chance, choice, environment, algorithmic randomness, or open parameters).
Possible structures (adapt, don’t copy):
- “Repeat the following steps until X happens …”
- “If you encounter Y, then do Z; otherwise, do W.”
- “Every N minutes, perform this action, unless condition C is true.”
- “Use a random number generator to decide which of three actions to perform in each location.”
Think of this as writing a subroutine in human language: a small, reusable procedure that could be “run” many times by different people or in different contexts.
Phase 2 – Plan the Performance & Documentation
IMAGE PLACEHOLDER – Diagram: instructions → performance → documentation
Define the live performance:
- Where does it occur (studio, public space, online, hybrid)?
- How long does it last (minutes, hours, days)? Is it continuous, periodic, or triggered?
- Who is “the computer”? Just you? Your group? The audience? A platform or device?
Design the documentation as a second artwork:
- Photos with text overlays, timestamps, or data.
- Short video (1–4 minutes) with on-screen text of instructions, logs, or subtitles.
- A simple web page (HTML), blog, or social feed thread.
- A designed PDF / “manual” that combines screenshots, maps, chat logs, and images.
- Alternative format (dataset, code + screenshots, map-based narrative, etc.).
Decide on an epistemic stance:
- Is your documentation faithful or deliberately unreliable?
- Does it reveal what “really” happened, or does it withhold crucial information?
- Does it flatten the performance into “evidence,” or does it open up new interpretations?
Phase 3 – Use Technology as Part of the Work
IMAGE PLACEHOLDER – Screenshot of code / timer / mapping / AI tool
The computer (or network/platform) should not be just a neutral tool; it should be part of the logic of the piece.
Choose at least one structural use of technology:
-
Use a randomization tool (random.org, dice, a small script, a phone app) to select actions, locations, or timing.
-
Use a timer, metronome, or scheduler to trigger actions at specific intervals.
-
Use a mapping or locative tool (Google Maps, GPS, step counter) as input to your instructions.
-
Use AI or code to generate mutations or alternative versions of your subroutine, and decide which to follow.
-
Use a platform logic (Instagram grid, TikTok video length, email threads, collaborative docs) as a constraint that shapes how the work unfolds and is documented.
Ask yourself: what does the system “know,” and what do you (or your audience) know? How does this difference shape the work?
Phase 4 – Perform & Document
IMAGE PLACEHOLDER – Stills from performance / documentation layout
Run at least one full execution of your subroutine (more if you want to compare different runs).
Your documentation artwork should:
- Be understandable to someone who was not there live.
- Show the relationship between instructions, performance, and trace without collapsing them into one thing.
- Stand on its own as an interesting, considered digital object (not just raw evidence).
Phase 5 – Reflection on Knowledge and Indeterminacy
Write a brief reflection (300–500 words) addressing the following:
-
How is your work known differently by:
- You (author/performer).
- Any participants or witnesses.
- Someone who only encounters the documentation.
-
Where does indeterminacy enter the system (chance, environment, timing, human decision, software/platform behavior)?
-
In what ways does the computer/platform act like a co-author or constraint engine?
-
Does the documentation stabilize the work’s meaning, or open new meanings that were not present in the live act?
Solo vs Group Options
IMAGE PLACEHOLDER – Diagram of roles in a group (writer / performer / documentarian)
Solo option:
-
You write and execute the subroutine yourself, but you may still involve others as “unwitting participants” (public, online contacts, etc.).
-
Focus on the relationship between your internal decision-process, the algorithm, and the documentation format you choose.
Group option (2–4 students):
- Divide roles clearly: at minimum, have responsibility for writing, performing, and documenting/editing.
- Embrace the division of knowledge:
- Writers may know the whole system and intention.
- Performers know embodied, situational aspects.
- Documentarians control what future audiences can see and know.
-
Make this asymmetry visible: who knows what, and when, in your final documentation?
Deliverables
- 1. Instruction-piece (subroutine) – typed, 1–2 pages max.
- 2. Documentation artwork – video, website, PDF, image+text sequence, or other approved format.
- 3. Reflection – 300–500 words connecting your piece to epistemology, algorithms, and documentation.
- 4. (Group only) – short note on roles and collaboration.