TYPE HomeLogo for TYPE Magazine

Join TYPE
Matthew Carter peering through a reducing glass.

Profile, New type

On a role

Profile, New type

Not content with making Miller, Galliard, Snell Roundhand, Big Caslon, Mantinia—and the ubiquitous Georgia—Matthew Carter is ushering in a contemporary superfamily, Role. His best yet?

Carter at his home in Cambridge. Photographs by Alessandro Vecchi

Matthew Carter is the rock star of type, and he’s just released a new album entitled “Role.” It is his biggest effort yet, a four-record set, including Serif, Sans, Slab, and Soft, with a total of 200 songs. Well, styles. And like his aging (but younger) countryman Paul McCartney, he’s still filling the stadium. Asked how this huge project started, Carter replies in one word: “Morisawa.”

“I have a long and personal association with Morisawa, going back to 1993, judging their type competition,” he says. “I hadn’t had a commission from them before, but in 2014, completely out of the blue, they approached me with this monster project.”

Takeshi Morisawa, executive director of the company, got to know Carter over these 25 years. Both enjoy hard work . . . and the occasional glass of sake and great sushi. Takeshi had contacted Matthew to talk about expanding Morisawa’s Latin library.

“They have a dominant position in Japan,” Carter explained. “They’re well represented in Asia at large, but as they expand internationally, they have to have Latin type.”

Over the time of the project, Morisawa began to make some interesting moves in the West, including an investment in Occupant Fonts, designer Cyrus Highsmith’s foundry. This year they launched a new retail site, Fontelier.

Takeshi Morisawa’s idea was to create a centerpiece for this strategy, a superfamily designed by a great designer. “They wanted to put a superfont in the middle of it, as a sort of a flagship,” Matthew explains. Superfamilies—a combination of compatible serif, sans, and sometimes slab-serif, rounded, or script styles—have become increasingly desirable as graphic designers look for fonts that work together for large applications, like the visual branding for a big company. Successful examples are Stone by Sumer Stone, Lucida by Kris Holmes and Charles Bigelow, and Officina by Ole Schäfer and Erik Spiekermann.

“So, we made a contract, and I did a presentation to the board.” Carter says. “It was going to be serif, sans, slab and round; many weights, and three optical sizes. They knew what they wanted.” The result, four years after Carter signed up for the job, is Role—with 200 styles.

Carter reviews paper proofs of the typeface design in progress, with Shotaro Nakano.

The team

Since leaving Bitstream in 1991, the first digital type foundry he helped start ten years previously, Carter has worked largely on his own. Supported by partner Cherie Cone, who takes care of the business side and manages production for the foundry, Carter & Cone, he has quietly worked at his desk turning out one typeface after another. He uses the classic desktop type design app, Fontographer, and has for nearly 30 years.

Realizing that a superfamily is a big challenge even for Matthew Carter, Takeshi had an idea: Bring in some talented type designers to help him. Along the way, they could gain invaluable experience in the design of Latin type.

“They chose three, two of them Morisawa staffers, Sakura Taruno and Shotaro Nakano,” Carter says. “The other is a very rare person in Japan, a self-employed type designer, Kunihiko Okano. He had been through the Type and Media program at the Royal Academy in the Hague, but he was the only one of the three who had knowledge of Latin type design—very good knowledge of Latin type design, which was a great help.”

And they not only helped, they moved to Cambridge, Massachusetts. “They came over on a six-month work visa, each of them,” Carter says. “We found them apartments in a building in Cambridge, a 15-minute walk from my place.

They were there from June through November. And we went right at it. At first, I worked with them every day, and when I wasn’t there, I was working on Role on my own.” Of course, the typeface didn’t have a name, yet.

Carter explains how they learned to work together: “It was sort of shakedown cruise,” he said, “because the design methodology at Morisawa is a bit more structured than I’ve gotten used to. So, we had some conversations early on. And once they understood that I wanted to them to work with me, not for me, everything became much easier. I let them individually decide which parts of the family they wanted to work on. Shotaro chose the Sans. Sakura chose the serif. Kunihiko chose the slab. And we all set to work.”

“For them,” Carter says, “it was a novel experience to work on Latin. For me, while I had collaborated with people before, of course, I never had a dedicated team of people working on a specific project—all the way through.”

“By the time they went home, they were really confident enough in what they were doing that they didn’t have to ask me every day, ‘Is this right?’” Carter says, “I thought that the more initiative they had, the more they would learn, and the quicker they would learn. Rather than my saying, ‘I’m the boss, you’re going to execute this the way I tell you,’ I said, ‘Here is my initial sketch, let’s talk about this, see if we think it’s right. . . . Now you take it. We’ve done the Roman, you do the Italics.’”

Sakura Taruno begins design with pencil sketches.

Like all collaborative design projects Role began to have its own logic, he says. One designer may try something, but you can see if it fits the logic. Or the other team members will point it out.

The resulting fonts speak for themselves, but the experience gained was not just on the side of the younger designers. “It’s funny because when you have been doing something for a very long time, there are a lot of things you do without analyzing quite why do you them,” Carter says. “But when somebody asks you why you do that you say, ‘Um, um, I can’t remember. I’m not sure. . . . We’d better look at that. . . .’”

“I found myself having to reexamine all sorts of things that I thought I knew!”

The designers from Japan helped Carter focus on the exact methodology he was teaching them. And the work became a team effort, building around Carter’s experience and talent.

The back story

There is no better-known type designer in the Western world than Matthew Carter. Learning punchcutting when he was 18, his work spans the eras of metal type, phototype, digital type for main-frame and mini computer computers, then for PCs and then for the web and mobile apps. He spent decades converting whole libraries to new platforms, which may be how he picked up so much knowledge about the essential design elements of Latin type.

Everyone calls him Matthew. He’s 82 this October, but his look reminds you that he is still a rock star in the type world. For many his steady hand on type design through the constant change in technology has seemed clairvoyant. But Matthew says the alphabet is a code, and what makes type readable does not actually change. That’s why we can read stuff that was printed 200 years ago.

It was his father, Harry Carter, typographer and Oxford historian (A View of Early Typography: Up to about 1600), who suggested an apprenticeship at the historic printing house and type foundry in Holland, Enschedé. There he fell in love with type design—and the process that made the font-cutting steel punches with an actual-size letter on the point, by hand. Instead of going to Oxford himself, as planned, he went to work drawing letters, as a freelancer, then at Crosfield Electronics, which distributed the Lumitype, the first successful phototypesetting machine in the West.

New technology, new fonts

In 1965 he accepted a design position at Mergenthaler Linotype, still on Ryerson Street in Brooklyn, where he liked being close to the machines. That fascination with the process of typesetting informed his work throughout his career.

From his point of view, the changes in technology happened one after another. He spent a lot of time rebuilding traditional type libraries for new platforms. Mergenthaler the needed a library for the new phototype machines. Then Bitstream needed one for digital. All the work didn’t stop him from making new typefaces, informed by his knowledge of history and the direction of technology.

Galliard, the Granjon-inspired Mergenthaler font of 1978 (later released by ITC), has the liquid smoothness of the phototype era.

Charter, the sturdy-but-stylish 1987 text type made for Bitstream, was optimized for the low-resolution printers of the day.

Verdana, 1996, was started with pixels, to make sure that it was readable at small sizes on the CRT computer monitors of the day, which were coarse compared to today’s screens.

These typefaces, 20, 30 and 40 years old, still look fresh and relevant, and bear no evidence of any technological restraints. Carter follows the changes in the tech, but he has said that type design is about the way the shapes of letters and words convey meaning, whatever the medium. There is an essential code of the alphabet in any language, and if a typeface gets too far away from it, you can’t read it.

Shotaro, Matthew, and Sakura: The team take a tea break. (Kunihiko Okano was out of town this week.)

Several of his most successful fonts are historical revivals. Big Caslon (1994) is true to the original English classic, but it adds a long-missing display size to all the Caslon revivals out there. Miller (1997) was inspired by the great “Scotch” transitionals of the Miller & Richards foundry, but it was not a copy, but rather a synthesis of the style, one that has been enthusiastically adopted by publications like the Boston Globe to New York magazine.

Some of his typefaces are used by millions, like Georgia, a cousin of Miller which seems to be the typeface for every web site that offers serious writing in the Latin world—from Medium to The New York Times. Others are almost forgotten, like the elegant-yet-sturdy Shelley scripts, Andante, Volante, and Allegro, done for Mergenthaler in 1972. But, they all started with a good idea, a fund of knowledge, and a sure hand.

Making the superfont

In June of 2014, the team from Morisawa settled into Cambridge. “Sakura had her own apartment,” Carter says. “But the two men had an apartment with an extra room that we could use as a work room. They brought a lot of good equipment. They were fully tooled up. For software, they used Gylphs throughout, mostly because Kunihiko used it at The Hague, and, also, because it’s not too hard to convert Fontographer files to Glyphs.” Carter had thought through the design when he responded to Morisawa with a presentation “which had a bunch of characters in each style, enough to start with.”

“Of course, some of them changed quite a lot, but we did not start exactly with a blank slate.”

They divided up the work, with Carter in charge. “We kept in mind that all the styles had to work together,” he says. “I think we all got further into it and developed a sense of how it all hung together. Then when the design started to cohere, it was exciting. I was astonished how quickly they did learn.”

Not that it was easy for the designers from Japan. “What could be more different than Kanji and Latin?” Carter asks. “When I am teaching at Yale, which I’ve done for decades and decades, one of things that really hard for students to do is to is to get the weights right. We have this jungle of letters: v and a and y and x and z and so on. They’re all over the place, and it’s hard to get the angled strokes to look the same weight as an H. I find myself at crits at Yale, saying. ‘I’m sorry but your W is too heavy, or this side of the X is too light.’”

However, this was not a problem for the team. “I think the reason is that if you’re trained in Kanji, everything has to have the right weight.”

“Another key thing,” he says, “is to get the spacing even. I was surprised by how quickly the Japanese designers took this on board, since in Japanese you don’t have proportional spacing. Everything is in a square box. But you can’t just throw a glyph in the box, because the spacing is critical.”

So, they came to the project with some transferable skills. And of course, they were familiar with the Latin script, Romaji, which is used frequently in Japan, and is incorporated into Japanese fonts. “In the time I’ve been going to Japan, you see many more signs in English. You can’t live in Japan without seeing a lot of Latin,” Carter says. “Young people speak more English than the older generation.”

Carter looking over early sketches for diacritical marks.

The team quickly overcame the cultural barriers. “Of course, Shotaro was in college in this country, and he speaks very good English,” he says. “Sakura, of the three speaks the least English, but she and I communicate extremely well. She sort of draws like I do, so we didn’t have any problems understanding one another.”

Carter was delighted with their approach. “When I’d go see the team, they would have prepared questions in a very orderly way,” he says. “They always did. ‘How do I do this? . . . If do this, what do I do about that?’ Very carefully thought-out questions. Which I would have to answer! And they didn’t let me off the hook!”

The group jelled as a team, and carefully tracked each other’s progress. “They’d say, ‘Oh, if Kunihiko is going to do a number 9 like that, then I better follow the drawing.’”

“I think you can see the other designers in this,” Carter says.” When we sat down and started do this in Cambridge, I said I wanted all the designers to be given credit. Everyone was skeptical, because Morisawa typically does not credit individual designers, which makes sense when you know the fonts start with some 8,000 glyphs, and a whole crew will work on one typeface.”

“But I am happy to inform you that if you look at the colophons for the first brochures,” he says “all their names are in there. I am very grateful to the senior people at Morisawa for agreeing to that.”

The families

The families of Role took shape: Serif, Sans, Slab, Soft. The team drew Italics, of course, for everything. There are at least seven weights for each, and as many as nine. And all have optical sizes: Text, Display and Banner. There is a big character set, supporting 98 languages including Vietnamese, Romanian and Turkish. And Role is replete with Open Type features: Old-style figures, small caps, scientific inferiors, fractions, case-sensitive publications, ligatures and superscripts. No wonder it took three years to get to the design “freeze.” And then they must have all taken a very deep breath.

“Role Serif has very simple structure,” Carter says. “And I did a thing which I think I first did with Miller, which always had a text and display: I used slightly different forms for some letters. In the display and banner sizes, some of the letters have slightly fancier forms in the display and banner, which I kind of like to do.”

These touches, the tails of the lower-case b, l, u, and y, and the upper-case B have little tails that give it a little feeling of “hand,” or “craft.” This in a time of algorithmic type design.

“I just don’t like—I mean I probably shouldn’t say this,” Carter says, “but in part because of variable fonts, we’re getting deluged with big, big families which are very, very rigorously programmatic. So, everything stays the same, while being different, if you know what I mean. So, I like to make some un-programmatic differences.”

These design details, as we found while laying out the magazine, make great display typography, where the glyphs in headlines fit so nicely together, they look like logotypes.

Asked if Role could become a variable font itself, with “g-subs” to bring in these details at the larger sizes, Carter replies, “We started before the craze. But I am sure it would work as a variable.”

The Sans comes across as a fresh, contemporary humanist. “In my original version,” Carter says, “it had quite of lot of modulation—thick-thin—for the text version. But when we got to the display, we decided we didn’t really like that so much. So, we backed off the contrast in the text. It is not a monoline now—it’s not like a geometric monoline, but it is a lot more monoline than it originally was.”

“But as you go up in optical size,” he says, “the thick-thin reasserts itself. But since it is not rigorously programmed, it changes a little bit in nature according to size.”

“The slab was largely done by Kunihiko, and it’s one of my favorite things about the whole project,” Carter says. “I think it will help make the whole typeface last. . . The challenge was to make something that seems timeless. But this pushes you toward the vanilla, the generic, which would be really boring. . . The slab is anything but boring.”

“I’d never been involved in a rounded before,” says Carter. “But as you probably know, rounded sans are far more popular in Japan than in the Latin world. So, it’s quite an important part of the family.”

The marketing people called it Soft, so all the styles began with an S, he explains. “It was an eye-opener to me because these rounded sans are very legible,” he says.

“The first version we did, looked sort of kitsch,” Carter says. “You think, okay, we’ll take the sans, round the stroke ends and terminals of the stems, what’s the problem? But it looked terrible. We found that we had to round them less; it was quite an interesting exercise.

“The first one looked a little “Hello Kitty!” Carter says. It’s a very subtle business. We went through three versions of this, until we got the roundness right.”

Next steps

So, Matthew, now what? “Well, we’ve finished the Vietnamese, and now we are moving on to the Greek and Cyrillic,” he says.

It’s clear that the team has succeeded in creating a flagship for the company’s expansion in the Western world. “They are working with the winners of Morisawa Competition,” Carter says. “Winners have the option of getting their designs produced and distributed by the company. As one of the judges, this makes me happy.”

“And Cyrus Highsmith and Occupant Fonts are working with Morisawa in Providence, creating two new families a year,” he says.

“Morisawa clearly is executing a Latin strategy. This project was an important step, a carefully designed superfamily, with all the glyphs­—and languages. They were very supportive all the way through,” Carter says. “When we started, the people at Morisawa were smart enough to realize that this would take time. We’d never done this before, none of us—build a type superfamily and train a whole team. How would we know how long it would take?”

The screen is good, Carter says, but you must how the type reads on all target media.

Well, it took four years from concept to release. Supplementing the Cambridge team were more than a dozen Morisawa designers and production experts, finishing character sets, finalizing the production, doing quality control, and testing.

It seems that everyone involved is proud of the results. With the launch of Fontelier and the arrival of Role on the scene, the type world is suddenly realizing there is a major new player in Latin.

“One way or the other, there is a new mood at Morisawa,” Carter says. “The company is transforming itself. As it is nearing its hundredth anniversary it is getting ready for the next century.”

Next entry: “Good Vanilla”

Previous entry: The last phototypesetter