Muriel Cooper and Typography That Moves
The Visible Language Workshop, Information Landscapes in 1994, and a research agenda that arrived about twenty years before the hardware could carry it.
Date
Category
Writer
Seven vertical bars. That is the whole of the MIT Press colophon: seven black strokes of varying width, which read as book spines on a shelf and simultaneously as the letters m, i, t, p. Paul Rand, who had already done IBM, recommended Muriel Cooper for the job. MIT's own news office says she designed it in 1963. In 2023 the Museum of Modern Art took it into the permanent collection, the only publisher's logo to be acquired that way.
It is a perfect small object and it is the least interesting thing she did. Cooper spent the last twenty years of her life on a question that the colophon cannot even be asked: what happens to typography when the surface it sits on is not fixed, not flat, and not still. She got a long way into answering it. Then she died, at the end of May 1994, a few months after showing the answer in public for the first time, and the answer was largely forgotten for two decades.
What she actually did at MIT Press
Cooper was born in Brookline, Massachusetts in 1925, took a BFA in design at the Massachusetts College of Art in 1948 and a BSEd in education there in 1951, and came to MIT's newly formed Office of Publications in 1952. She returned to MIT Press as design director in 1967. Across her career she is credited with the design of more than five hundred books.
Two of them are famous. The MIT Press Bauhaus book appeared in 1969. Learning from Las Vegas, by Venturi, Scott Brown and Izenour, came out in its first edition in 1972, also her design.
The important one is neither. It is File Under Architecture, 1974, and the reason it matters is a question of who controls production. This is also the first point where the sources I could reach disagree with each other. The English Wikipedia entry describes it as the first book she typeset directly on a computer. Frieze describes it as an innovative use of the IBM Selectric typewriter, which for the first time let designers rather than typesetting firms control the layout. Those are not the same machine or the same claim, and I have not been able to establish which is correct. What both accounts agree on is the consequence: the setting of the type moved out of the trade shop and onto the designer's desk. Every argument Cooper made for the next twenty years follows from having experienced that shift personally in 1974.
A Macintosh 128K, 1984. The Visible Language Workshop was already asking what type would do in three dimensions on machines roughly this size. Photograph by Sailko, CC BY 3.0, via Wikimedia Commons.
The Visible Language Workshop
In 1974 she left the full-time MIT Press post and co-founded the Visible Language Workshop with Ron MacNeil, in the Department of Architecture. That is the year Wikipedia, MoMA and the MIT Media Lab all give. Frieze dates the beginning to 1975 at MIT's Center for Advanced Visual Studies. Both may be true of different moments in the same founding; I flag it because it is the kind of detail that gets copied without checking, and I have not resolved it.
The VLW's research programme was, from the start, about separating what a text says from how it is set. In 1978 Cooper worked with Nicholas Negroponte on "soft copy", a representation of text independent of its typographic formatting, and in the same year Cooper, Negroponte and Richard Bolt submitted what has been described as the first virtual book proposal to the National Science Foundation. That is more than a decade before the web and about sixteen years before anyone was arguing about whether presentation belonged in a separate file from content.
When the Media Lab was founded in 1985, the VLW was folded in alongside the Architecture Machine Group and the Center for Advanced Visual Studies, and became one of its first research groups. Cooper became, by the accounts of both the Media Lab and Wikipedia, the first woman granted tenure there, and at her death was the only one.
I guess I'm never sure that print is truly linear.
That sentence of hers, quoted by Frieze, is the whole research programme compressed. It is not a claim about screens. It is a claim that the linearity of the printed book was always a property of the manufacturing process rather than of reading, and that once the manufacturing constraint lifted, the question of how information should be arranged reopened completely.
What Information Landscapes actually was
Information Landscapes was presented at the TED5 conference in Monterey, California, in 1994. I have not been able to confirm the precise date within that year from a source I trust; Frieze places it months before her death on 26 May 1994, which narrows it to the first part of the year.
MoMA holds it, and the catalogue record is the most reliable description available.
MoMA describes the work as a graphical interface for creating dynamic three-dimensional visualisations of complex real-time financial, statistical and spatial information, and as demonstrating radical possibilities for graphic design in the digital realm. Note what that is not. It is not an animation. It is not a rendering of a design. It is an interface, running live, in which type exists in a navigable three-dimensional space and the viewer's position in that space is part of what the information means.
Four things were demonstrated that had not been demonstrated together before. Type was placed in three dimensions, so that depth, scale and viewpoint carried meaning rather than decoration. Navigation was continuous and under the reader's control, so a document became something closer to a place than a sequence. Content and form were held separately, with the layout generated at the moment of viewing rather than fixed in advance. And filtering was algorithmic, with software selecting and organising from real-time data streams while the designer's job was to build the interpretive frame rather than to place each element by hand.
Cooper had already stated the design position five years earlier, in "Computers and Design", Design Quarterly no. 142, 1989. She named the main challenge as "the use of graphic design as a filter for this complexity of information", a way of making it both meaningful and expressive, and argued that designers had to appropriate the algorithms rather than be displaced by them. Her better-known line, quoted by MIT, is blunter: "Information is only useful when it can be understood."
Video. Douglas Engelbart and team, 1968. Published by SRI.
What the sources disagree about
Because so much of the Cooper literature is written from other secondary accounts, it is worth setting the disputes out explicitly rather than smoothing them over.
Question | One account | The other account |
|---|---|---|
Year of the MIT Press colophon | 1963, per MIT News | 1965, per the MIT Media Lab and Wikipedia |
Founding of the Visible Language Workshop | 1974, per Wikipedia, MoMA and the Media Lab | 1975 at CAVS, per Frieze |
Technique of File Under Architecture | Typeset directly on a computer, per Wikipedia | Set on an IBM Selectric typewriter, per Frieze |
I would take MIT News on the colophon and 1974 on the workshop. On File Under Architecture I have no basis for choosing and will not pretend otherwise.
Twenty years early, and what that means
The $250,000 machine is the fact that carries the argument. In 1994 a demonstration of live three-dimensional typographic navigation required a Silicon Graphics Reality Engine, a quarter of a million dollars of hardware in a university research lab. That is not a product. It is a proof that the thing is possible at all, and a statement of how far the world was from being able to do it.
The gap is measurable in a second way. Cooper's separation of content from typographic form dates to 1978. The idea had to wait for the browser and the stylesheet to become something a working designer used daily. Her filtering-from-live-streams work dates to the early 1990s and is now the ordinary condition of almost every screen anyone looks at. The work was roughly twenty years ahead of the machines and the infrastructure it needed.
The cost of being that early is brutal and specific. The original projects will not run on a contemporary computer. What survives is the 27-minute video, a handful of interviews, and the memories of the people who were in the room. The research was largely forgotten for close to twenty years, until the 2014 exhibition Messages and Means at the Arthur Ross Architecture Gallery brought it back into view. The clearest surviving descendant is oblique: Processing, released in 2002 by Casey Reas and Benjamin Fry, students of John Maeda, who succeeded Cooper.
Here is where I want to be careful, because there is a sentimental version of this story that I do not believe. The sentimental version says she was a visionary and the world caught up. That is flattering and useless, because it implies the only thing wrong was timing, which would mean there is nothing to learn.
The non-sentimental version is that design research produces two different kinds of output, and we habitually judge it on the wrong one. The first output is the artefact: a running system, a demo, a published thing. Artefacts decay. A 1994 artefact built on one $250,000 machine was never going to survive, and everyone involved knew it. The second output is the question, stated precisely enough that other people can work on it. Questions do not decay.
Cooper's questions were: what is a typographic hierarchy when the reader controls the viewpoint; what is a layout when it is generated at the moment of reading; what is a designer's job when an algorithm does the selecting. Those are still the live questions in every interface anyone is building today, thirty years on, and they were correctly posed before any of the enabling technology existed. That is the achievement. The Reality Engine demo was the evidence that the questions were real, not the point of the exercise.
Which implies a standard. Design research should be judged on whether the question it posed turned out to be the right question, assessed at a distance of decades, and not on whether its demo shipped. By that standard the Visible Language Workshop is among the most successful design research programmes anyone has run, and the fact that almost nothing of it executes today is close to irrelevant.
It also implies an obligation that the field has not met. If the question is the durable output, then the question has to be written down in a form that survives the hardware. Cooper's were preserved largely by accident: a video someone kept, a donation by two former students, a museum department that took it. Lisa Strausfeld's remark about being taught by her, quoted by the Media Lab, is lovely and slightly terrifying in this light: "She taught in a very unusual way. I didn't even know that I was being taught." That is how the best research supervision feels from inside, and it is also how a body of work disappears when the supervisor dies suddenly.
The sentence of hers that I keep coming back to is the one Frieze quotes as a statement of the new condition: "This new world demands a new kind of designer who will provide opportunities, pathways and processes for a more independent user." Written before the web was a consumer thing. Every word of it is still an open brief.
Sources
The Museum of Modern Art. Collection record: Muriel Cooper, David Small, Suguru Ishizaki, Earl Rennison, Robert Silvers, Lisa Strausfeld, Jeffrey Ventrella, Yin Yin Wong. Information Landscapes, 1994. Multimedia video, 27 minutes. Gift of Lisa Strausfeld and David Small. Accession 1078.2015, Department of Architecture and Design. moma.org/collection/works/199050
Maudet, Nolwenn. "Muriel Cooper, Information Landscapes." Back Office, issue 01, "Making Do, Making With". Details the Silicon Graphics Reality Engine purchased for USD 250,000, the seven subprojects, the VLW team, the 1978 soft copy work with Negroponte, the 1989 Design Quarterly position, the inaccessibility of the original projects, the 2014 Messages and Means exhibition, and Processing (2002) by Casey Reas and Benjamin Fry. revue-backoffice.com/en/issues/01-making-do-making-with/nolwenn-maudet-muriel-cooper-information-landscapes
MIT News. "A symbol with a story: the MIT Press colophon." May 2023. Gives 1963 as the design year, Paul Rand's recommendation, the seven bars reading as m-i-t-p and as book spines, the 2023 MoMA acquisition, and David Reinfurt's comment on the drafts. news.mit.edu/2023/symbol-story-mit-press-colophon-0512
MIT Media Lab. "Muriel Cooper's Lasting Imprint." Published on the MIT Media Lab channel, 19 October 2017. Gives 1952, the 1965 colophon date, the 1974 founding of the Visible Language Workshop with Ron MacNeil, the 1985 Media Lab merger and tenure, the TED5 presentation in 1994, more than 500 books, and quotations from Nicholas Negroponte, Lisa Strausfeld and Tod Machover. medium.com/mit-media-lab/muriel-coopers-lasting-imprint-b7a8b5d9d19a
Wikipedia. "Muriel Cooper." Born 1925 in Brookline; BFA 1948 and BSEd 1951 at Massachusetts College of Art; Office of Publications 1952; design director of MIT Press 1967; VLW co-founded 1974 with Ron MacNeil; Media Lab 1985; Bauhaus 1969; Learning from Las Vegas 1972; File Under Architecture 1974; "Computers and Design", Design Quarterly no. 142, 1989; TED5 1994; died 26 May 1994 in Boston. en.wikipedia.org/wiki/Muriel_Cooper
"This New World." Frieze. Source for the 1975 CAVS dating, the IBM Selectric account of File Under Architecture, the 1978 NSF virtual book proposal with Negroponte and Richard Bolt, the TED5 appearance alongside Bill Gates, and the quotations "I guess I'm never sure that print is truly linear" and "This new world demands a new kind of designer who will provide opportunities, pathways and processes for a more independent user." frieze.com, article titled "This New World"
Related reading
Latest Articles.
© Moc Mallen Studio
Thoughts, ideas, and perspectives on design, simplicity, and creative process.
.jpg){kind=link}