Artist Sterling Crispin has been engaging in a remarkable project over the course of the year: designing software to simulate the “growth” of a plant in a 3D model, which can then be printed using standard 3D printing equipment. He was kind enough to explain to me how this project is going, what it means to “grow” something in simulation, and how it fits in with the work of other artists working with additive sculpture technologies.
Adam Rothstein: Could you describe your project to grow 3D plants, in your own words?
Sterling Crispin: Well it’s one part tool-unto-itself and one part interactive-artwork that I’m calling Neophyte. It’s a large software project in the C++ programing language that generates relatively life-like plants. My current focus has been on 3D printing the plants so I can use them in my sculptures as form and ornamentation.
The plants are generated in part using an algorithm called an L-system, which is a grammar that describes how single-celled organisms subdivide. This algorithm produces a kind of recursive “DNA” that my software then contextually re-writes and edits, and finally interprets into a volumetric 3D form. Neophyte was partially based on the first chapter of The Algorithmic Beauty of Plants by Prezmyslaw Prusinkiewicz and Aristid Lindenmayer. But it’s not just a 1/1 adaptation of their work, I’ve made lots of progress and filled in lots of gaps, made things up, cheated and faked it, to get Neophyte where I wanted it to be.
The deeper I’ve gotten into writing this software simulation, adding minute little details here and there that only I’ll ever notice, I realize that it’s really a game of approximations. If I really wanted to program a truly generative plant I’d have to start with simulating carbon atoms. But obviously that would take forever to compute, so I have generalize a lot.
I’ve spent so much time writing the Neophyte itself that it probably would have been faster to sculpt the plants by hand, but the forms would inevitably turn out differently, and I’m interested in the hand-of-the-machine and the kind of marks it makes.
For me the work is about entropy, the environment, and technology. We’re collectively creating this global-organism of networked information systems, but destroying the earth’s ecosystem to live like this. So there’s something kind of romantically futile about trying to re-create beautiful natural systems using high technology that’s partially responsible for its destruction in the first place.
AR: So there is a sort of abstraction going on, and necessarily so. A translation of biology into data? Or how else might you put it, if not like this?
SC: There’s a definite translation from biology to data, but it’s not direct. It’s the biological abstracted into strict logical rules, mixed with a lot of arbitrary subjective feelings, and memories, of “how nature is,” that’s guiding the software. I want these 3D printed parts to seem like they physically grew and were somehow made to be synthetic.
A frozen state of a living synthetic system that somehow emotes life while existing as decaying plastic.
It’s like the fable of the artist who drew sleeping pigs, the artist’s mentor kept asking, “why are you drawing all these dead pigs?” Finally with enough practice the artist was able to emote that feeling of a pig resting peacefully in the shade. That’s where I’m trying to get.
AR: Did you have to research actual biology much, or are you building on the algorithmic/programming side of things?
I’ve read through lots of research papers about how people are developing algorithms for simulating plant life, but most of my biology research involves being in nature and just silently observing the world. Or digging around in the freestyle garden my girlfriend and I planted in our yard. There’s something really satisfying about watching plants grow. It’s like hearing birds singing, you just feel like the world is a little more right, and everything is going to work out just fine.
AR: What are the challenges working with this digital/physical art form?
SC: I became completely obsessed with minute little details of the software. The saying “Art is never finished, only abandoned” goes for software as well. At some point you have to just walk away from it.
Digital art has the advantage of being so fluid and malleable, but industrial manufacturing methods have strict limits, 3D printing included. Working between those extremes can be tough, especially with custom software like this.
Also when you’re working with generative systems they spiral out of control, it takes a lot of tweaking to develop the right results. There are hundreds of parameters all working together in harmony. If I deleted the default settings and let someone else try to use the software they may not get it to do anything interesting. It’s kind of like a musical instrument.
Beyond all of that, C++ is not the friendliest programming language once you start doing advanced and sort of exotic things with it.
AR: What are the signs of success? In other words, as you improve upon your method, what strikes you as “getting it right’?
SC: Sometimes I’m working from photographs I shot of plants and trying re-create subtle and specific details, like how leaf-stems emerge from their branches. With those references and fresh memories of direct observation it’s easy to know if I’m getting it right or not.
But when I’m trying to make a new feature in the software that has a parameter space covering a wide range of plants, it’s way more subjective. For example, writing code that causes branch bifurcation and leaf placement that can equally emulate a grape vine, a sparse shrub, bamboo, and an oak tree. There’s a lot of feeling it out and seemingly arbitrary decisions made. I think when people who’ve never written code criticize digital art, they don’t take into account how many creative decisions are made. Every line of code is a creative act.
And in general I want it to be beautiful, and you know it when you see it.
AR: Has the work been shown yet?
SC: I had produced Entropy and Extropy as commissions for the art collector Stefan Simchowitz. The plants in those two sculptures were grown in the same software, but that was way back in March, so the software has evolved a lot since then. Now there are leaves, a physics simulation cycle, and a bunch of subtle changes about how branches rotate and merge to face one another.
Neither sculpture have been exhibited yet to my knowledge but partly that’s due to the rate I’ve been able to make new artwork. I’ve been totally obsessed with developing the software itself for the last 9 months, and it’s finally maturing to the point that I’m ready to just focus on creating physical sculptures with it and stop nitpicking the code.
AR: Other work you’ve done involves creating a technological setup, and then seeing what results— kind of performance, or time-based work, one might say. Do you see this work with generating plants as sculpture, performance, or both?
SC: I don’t think about it categorically like that while I’m working, it’s just art to me. It’s one total system with lots of inputs and outputs.
But I suppose it’s a tool to make sculptures, and a series of digital video artworks, it’s an infinitely generative system, and yeah its performative while I’m using it. But more of a performance like an instrument being played than an Abromovic piece.
AR: How do you think this project fits in with other 3D printed art work that is out there?
SC: Difficult to say, I guess I relate to the aesthetics of people like Daniel Widrig, Neri Oxman, Michael Hansymer with this work, but the politics of the Nefertiti bust leak, and Morehshin Allahyari’s work. in the sense of creating modern artifacts.
I’m interested in the shelf-life of these plastic parts after reading some archival nightmares about old Jeff Koonz sculptures and others. Thankfully I’ve got all the digital files, which is kind of the DNA of the work, the physical parts themselves can always be reprinted.