Conversation with Ellen Lupton, May 7, 1994. Unpublished.
Tell me about the Visual Language Workshop, which you direct at the MIT Media Lab.
My one-liner on the goals of the VLW: we’re exploring or researching new design characteristics or principles or vocabularies that will characterize what is essentially a new medium.
What is this new medium? In general its outstanding characteristics are dynamic in real time, interactive, incredibly malleable, some capability of learning and adapting to the user, or to information, or to some other set of relationships. Our goal is to make information into some form of communication, which information alone is not. Information by itself does not have the level of “filtering” that design brings to it.
I’m concerned with what’s the new definition of graphic design, and what role it plays with regard to information. If you take a book, in traditional terms, or a magazine, newspaper, or television news program, the object has been filtered through some technological constraints.
Graphic design can be called a filtering process. Materials come to you from an editor, writer, photographer, etc. The designer filters these existing elements. Perhaps the role you’ve developed for yourself in your [Ellen’s] work as a curator is a more appealing model, because you design everything.
Well, I’m still working with existing materials; I just choose all the materials and design the setting in which they will be seen and interpreted. I have more editorial control than designers normally do.
Anyway, in the traditional model, the designer tries to interpret what these elements are “supposed to do” together. So what happens with computers (beyond the primitive desk-top publishing model)? On the “information highway,” all sorts of things are up for grabs—authorship, how people read, how people gather and generate material for their own purposes.
This has to demand new processes and concerns for designers. How do you design structures and processes for containing information in this new environment?
We’re trying to build an environment here at VLW for testing ideas in this new medium. If the task is to gather news, for example, how do you create a system for giving people the info that they want? Do we design pathways for different kinds of people? Does the system adapt to users?
One of the professors in the Media Lab is a physicist, whose students have built a board that is quite unobtrusive, set into a table. Just gesturing over the board lets you move through “pages” of text. The board knows where the user is in space, in relation to the computer screen. Different departments of the Media Lab are getting those boards to experiment with. VLW is getting one of these boards. We’re experimenting with using this sensing device to determine distance of viewer: farther the viewer, the bigger the type.
I have always been frustrated and intrigued by technology. Jackie Casey and I both went to Mass College of Art in the late 40s. We were cashiers in the school store; we both eventually became bookkeepers—first Jackie and then me. We learned more in the store than we did in the school. In a way, I think of the school store as a model for the VLW. When the store would close in the afternoon, the students who worked there—about a dozen of us—had a studio to ourselves, our own little bin of paints and papers and materials.
Lots of good people were in Boston at the time. Gropius was at the Harvard GSD, Kepes at MIT. Lots of great lecturers came through town—Thornton Wilder. As students, we would work more with each other than with the faculty. About a dozen of the best and the most interesting students did a lot of work together at night when everyone was gone. We were very experimental.
After school, I went to New York and tried to get a job in advertising. It was very difficult for everyone then, in the early 50s. Paul Rand was one of the first people I met with in New York. I believe he was at Esquire then. He was very critical of my typography.
I had a mission: design was a way of life. That was influenced by Gropius, Kepes, Herbert Read, even Dewey, who was still around then. Plus Duchamp was very important, the French, Gertrude Stein. I was distinctly a modernist. I still am, I fear, but modernism is coming back.
I was a modernist, but I was an uneasy Swiss, if you know what I mean.
We had a discussion here at VLW recently about Barbie. Students name all the machines; I don’t interfere. A new machine came in and someone named it Barbie. They wanted to put a plastic doll on top. I freaked. I sent mail around saying, “Anything but Barbie. Please don’t.” This precipitated a slew of mail about Barbie as a symbol. For me, she and Ken are sex objects that have demeaned what women should be accomplishing. I got back this unbelievable range of responses—Barbie as an underground anti-establishment figure. Apparently there is a Barbie renaissance going on right now…
Symbols have shifted through the generations. Can we coexist with our different symbols? I am a moderate feminist here on the faculty, who is well known on the faculty for this. I’m not a rabid, heavily political figure, but I let it be known when women are being ignored in decision making… Now, I’m getting this way about age.
Finally, I decided it was okay to name the machine Barbie, but not to put a doll on top of it. David Small, one of the great student contributors here, is openly gay, lives with his partner. He got upset about a counter-name proposed for the machine—Sissy. He thought this was anti-gay. Anyway, it’s a constant adjustment to changing symbols.
How does your experience as a book designer relate to your experience here?
My accumulated activities over the years give me a foundation. I worry about being boxed in by my experience. But on the high side, I was always trying to push the medium of the book in new directions.
Such as Learning from Las Vegas?
Yes, I look back to that. I remember there being more intensity to that book than I could find going back to it more recently…. Jan Abrams asked me about that recently. Original cover for Learning from Las Vegas was bubble-wrap with flourescent dots underneath. Denise hated it. That book was done on an IBM composer. That was the machine that allowed you to experiment with typography. I tried many things—including interleaved narratives—that couldn’t finally be included in the book.
Another book, File Under Architecture, was Herbert Muschamp’s first book. This was my favorite book. Very innovative, in the mid-70s. It was done on brown wrapping paper and set on an IBM composer, which was a typewriter (designed by Eliot Noyes) that has a head with a type ball on it. The ball let you change typefaces. It was very tedious of course—change the ball to get bold type… I thought about books as being like a movie. Once I presented the Bauhaus book as a single-frame movie—showing all the pages in rapid succession. It was fantastic.
Can you tell me about VLW history?
The 70s was the period of alternative book art—Xerox machines and corner copy shops were beginning to spread out, becoming more available. I was at MIT Press. I got support from the director to look into other media, electronic media. I pushed to get computer typesetting in house, which would give me an opportunity to explore the medium. I pushed for an experimental arm of the press that would do smaller edition experimental books. I sat in on a seminar session taught by Nick Negreponte on “Computers in Design.” The course attracted a lot of architects and engineeers. It was the early 70s.
So I had a little support for this R&D unit at MIT Press. It was eventually shut down for financial reasons. We did some stuff with rubber stamping, cut and paste—it was the Whole Earth Catalogue era. There was a lot of Method Acting in what I was doing. In the case of Learning from Las Vegas, I had given them a monument, and what they wanted was a duck. Something more casual…
The first book I did with MIT Press (as a freelance) was View from the Road, by Kevin Lynch. Idea was to develop a graphical language for engineers to use when building roads, so that they would think about views and reference points. It was a long tall book, with a flip book on the side. The author hated it—it was out of the reach of the engineers. Too big, arty.
It was not really an experimental book.
More history of VLW?
Tom Norton, who had been at RISD, was working with a color copier. He turned up at my door to show me some of his work with color copies. We made a deal with the local 3M distributor to house a color copier at MIT Press, that Tom could use for his art work, we could use to work on covers, etc.
Nick put a couple of machines at the Press that had some minor graphic capabilities. We did some editing on this machine. It had crude graphic possibilities. Ron MacNeil came to me to learn about graphic design. He was a physicist, a photographer, a Minor White disciple. Minor White ran a photo workshop at MIT; he had come to MIT in the late 60s.
Weird things have always been happening at MIT. Kepes was an extraordinary influence. Look at the book Hidden Curriculum, by Ben Snyder about MIT and the underground activities here. Published about 15 years ago.
So Ron and I met. He was incredibly open, always moving into new domains. He was in alternative printmaking; he had a press. Minor White drew the line—take the press and go. So Ron took his press and set it up somewhere else at MIT, in an unwanted space in the architecture department. I had been teaching at Mass College of Art and really liked teaching. I thought, wouldn’t it be interesting to teach a class at MIT? Between Ron’s press and my machines at the Press, we established a course called “Messages and Means.”
I found that students at Mass College of Art freaked out when they had to actually confront printing technologies. “Messages and Means” was about directly confronting technology. We started producing “one-night prints,” by passing over stages of paste-up and films. Working with acetate, press type, Letraset, etc., to expose the plate directly.
Now, we’re working with design principles for the new medium. Also, we’re working on ways to do design automatically using AI principles, as in Louis’s project.
Henry Lieberman, in MIT’s AI Lab (not part of Media Lab) helped pioneer the field of “object-oriented programming” and “programming by example.” Henry became a resident here at VLW 6 or 7 years ago. Ron MacNeil began to get deeply involved in AI techniques; he brought a visual attitude to it.
Could you explain the conjunction of AI and design?
I would like to see enough intelligence with enough rich graphical vocabulary that a designer could interact with technology in an empowered way. The designer should be free from the technical details in order to work with the more interesting aspects of the medium—finding pathways, etc. The machine can be a substitute intern or assistant. The machine can also take on a teaching role, and the designer becomes the intern. For the expert, the machine could be an intern. For the novice, the machine could be a teacher.
Please explain “programming by demonstration.”
You demonstrate something to the system, and it generalizes something from that and creates a program. “Machine learning” is a broader field of which this is a part—but programming by demonstration is its own field. Teaching the system something that it didn’t know before. “Macros” are a dumb version of this field, where you combine a bunch of commands that are called up by one keystroke. A template… Louis’s stuff is really different from that. Here’s an example that might help explain how the graphical part of the lab feeds into our AI concerns. Ron has a project that’s a map of Boston with a subway map superimposed. He taught the system to make a generalized subway map in relation to a geographical map of the city. If you tell the machine how to do Boston, then it can generate a subway map of Atlanta. You’ve given the machine an understanding of the problem. Ron used “case-based design,” where you give the system a number of “good cases.” The system looks for comparable examples in the new situation, modifying them where necessary. This is the overlap between AI and design.
There are two parts of the lab: hi-quality visual presentation and programming. It’s important that the flashy visual stuff feeds into AI. The explorations in 3D typography have led us to characterize what it is about 3D type that is special—what makes it a medium. A general theory is emerging about the nature of 3D typography. Then you could make systems that generate 3D typography in an intelligent way in more complex circumstances.
“Typographic Constructs” [the 3D type grid shown near beginning of TED demo] was designed by myself and David Small. Idea here was to create a vocabulary of “geometric primitives” that are part of this theory of/language of 3D typography.
Other considerations are how do particular kinds of info behave in 3D: numbers? text? video?
We’ve found that in 3D space, there are many of the solutions to problems we’d been trying to solve in 2D, such as the “piling” and obscuring of layers of information. Earl’s “Filtering the News on Internet” project is 3D typography, but only in parallel layers of space [one-point perspective]. It’s a project in self-organizing data—it has no actual “intelligence.” This is an example of a project that could become the basis for an AI project. AI could solve some of the legibility problems by setting up rules for how type presents itself, depending on length, etc.