Convivial Fabrication: Towards Relational Computational Tools For and From Craft Practices
This paper argues that current computational fabrication tools often fail craft communities by treating materials as passive, and proposes a framework of "convivial relations" across individual, community, and institutional levels to guide the design of tools that foster active material dialogue, collective knowledge, and autonomy.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
The Big Idea: Tools That Talk Back
Imagine you are cooking a meal.
- The Old Way (Current Tech): You follow a rigid, digital recipe on a screen. The computer tells you exactly how much salt to add and how long to cook the steak, assuming every steak is identical. If the steak is tough or the salt is wet, the computer doesn't know. It just keeps following the plan, often ruining the food because it ignored the reality of the ingredients.
- The New Way (This Paper's Goal): You taste the food as you cook. You adjust the heat when the pan gets too hot. You listen to the sizzle. The food "talks back" to you, and you talk back to it. This paper argues that our current computer tools for making things (like 3D printers, laser cutters, and design software) are too much like the rigid recipe. They treat materials (wood, yarn, metal) as passive objects that just sit there waiting to be shaped.
The authors want to build new tools that treat materials like partners in a conversation, not just raw stuff to be forced into a shape.
The Problem: The "Blueprint" Trap
The paper says most computer tools use what experts call a "Hylomorphic" model. That's a fancy word for "imposing a shape on a lump of clay."
- The Analogy: Imagine a sculptor who has a perfect statue in their head. They use a computer to carve a block of stone exactly to that shape. But what if the stone has a hidden crack? Or a weird grain that makes it split? The computer doesn't care. It just keeps carving. The result is a broken statue, or a piece of wood that warps because the computer didn't know the wood shrinks when it dries.
- The Result: Craftspersons (people who make things by hand) feel disconnected. They lose the "feel" of the material. They stop listening to the wood, the metal, or the yarn, and just start obeying the screen.
The Solution: "Convivial" Tools
The authors borrow a concept from a philosopher named Ivan Illich. He talked about "Convivial Tools."
- The Bicycle vs. The Car: A bicycle is a "convivial" tool. You pedal, you steer, you feel the road. You are in control, and the bike amplifies your movement. A car, in Illich's view, is less convivial because it requires a massive infrastructure (highways, gas stations) and you are just a passenger in a system.
- The Goal: The paper wants computer tools to be like bicycles. They should help you make things, but they shouldn't take over. They should let you stay in charge and keep your relationship with the materials strong.
The Three Levels of Connection
The researchers interviewed 23 expert crafters (woodworkers, fiber artists, metalworkers) and found that "making things" happens on three different levels, or "orders." Current tools mess up all three.
1. The Immediate Level (You + The Material)
- What it is: The moment your hands touch the wood or the needle touches the thread.
- The Problem: Computers create a wall between you and the material. You design a chair on a screen, but you can't feel if the wood grain is going to split when you cut it.
- The Fix: Tools should let you "listen" to the material. If the wood is hard, the tool should slow down. If the yarn is slippery, the tool should adjust. It should be a dance, not a march.
2. The Community Level (You + Other Makers)
- What it is: How crafters share secrets, tricks, and stories with each other.
- The Problem: Today, sharing happens on platforms like YouTube or Instagram. But these platforms often turn craft into a "spectacle." They show perfect, finished products and give you a "magic ring" kit that does the hard work for you. This kills the deep knowledge of how to handle the material. It's like watching a magic trick vs. learning how to do the sleight of hand.
- The Fix: Tools should help communities share the messy parts of making. Instead of just posting a photo of a finished sweater, share the story of how you fixed a mistake, or how you adapted the pattern because the yarn was weird.
3. The Extended Level (You + The World)
- What it is: Where the material came from, who grew it, and how it got to you.
- The Problem: Supply chains make materials look like magic. You buy a spool of "sustainable" cotton, but you don't know it was grown in a drought-stricken area or picked by underpaid workers. The computer tool just sees "Cotton: $5." It erases the history and the cost.
- The Fix: Tools should tell the story of the material. They should show you the journey of the wood or the fiber, so you know the true cost and can make ethical choices.
The 7 Rules for Better Tools
Based on what they learned, the authors suggest seven rules for designing these new, friendly tools:
- Embrace "Imperfect" Materials: Don't pretend every piece of wood is perfect. Let the computer know about the knots and cracks so you can design around them, not ignore them.
- Use "Fuzzy" Models: Don't demand a perfect plan before you start. Let the design be a rough sketch that changes as you work. If the material resists, let the tool change the plan with you.
- Slow Down to Natural Rhythms: Don't force machines to go as fast as possible. Let them move at the speed of the material (like the rhythm of weaving). Sometimes, pausing is part of the process.
- Support Growing Knowledge: Let communities share their "work-in-progress" and their mistakes, not just their perfect final products.
- Connect the Dots: Link the immediate action (cutting wood) to the community (how others cut wood) and the source (where the tree grew). Show the whole picture.
- Plan for Mistakes (Fault Tolerance): If a tool breaks or a material fails, the system shouldn't crash. It should suggest a new way to finish the job, just like a human would.
- Be Modular (Mix and Match): Don't lock people into one giant software program. Let them swap parts of the tool to fit their specific needs and share data between different tools.
The Bottom Line
The paper isn't saying "throw away computers." It's saying: Let's stop treating computers like bosses that tell materials what to do. Instead, let's build tools that help humans and materials have a conversation. This way, we can make things that are more creative, more honest, and more connected to the world around us.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.