Minion is a project that you’ve worked on now for more than 25 years. Tell me how it began.
During the period I was working on Adobe Garamond, in the late 1980s, I had begun to study in earnest a broader spectrum of early roman types. The Aldine types seemed to me to be the purest distillation of the humanist aesthetic as type. Griffo essentially refined and resolved what Jenson and others had begun. Not only are the Aldine roman types beautiful to look at, they are very practical. When I began working on Minion, I was focused more on the humanist principles that define the Aldine types rather than using it as a direct model. I wanted to do something more contemporary with a unified calligraphic energy.
Outline drawing software was not highly developed in those days. Were you able to draw the letters on a computer the way you do now, or did you rely on pen and paper?
We were using an early version of Illustrator to digitize our drawings. In addition, the Unix-based type-editing software that we had was much more restrictive than the letter-drawing programs that are available today. In those days it was natural to produce a fairly finished set of drawings in order to minimize the amount of on-screen letter-shaping. Today the majority of my design work takes place directly on the computer. I still will use a broad-edge pen to develop a script or formal calligraphic design, as well as a simple ball point pen or a pencil to quickly flesh out a new idea.
A lot of us seized upon Minion when it was released in 1990 and began to set books in it. One reason was that it was a handsome book face, with good built-in kerning – and that was a real novelty in those days. Another reason was, it was remarkably efficient. We got more letters on a line without the text looking cramped in any way. Was the efficiency premeditated? Did you set yourself a target character count as part of the design brief?
A lot of people have said that Minion is like a modern Times Roman – not because it looks like Times but because, like Times, it’s very efficient. Certain letters, in Times Roman, are very narrow, and the face is tightly fit, yet it reads well. At the same time, some of the letters are quite wide, so there’s a nice dance of proportion. But I wasn’t looking at Times, or thinking about it, when I designed Minion. And I didn’t set any particular character count as a goal. The efficiency developed naturally in the process of figuring out the proportions. Aldine romans and chancery italics tend to have quite long ascenders and descenders, and the romans have fairly large capitals. My goal with Minion was to make a distinctly contemporary all-purpose family with more normalized proportions for general use. I also wanted to imbue the design with a subtle stylized calligraphic sensibility to help unify the roman and italic. I didn’t want to make a historical copy, and I didn’t want to make a face that was just for book production.
Historical Aldine romans were devised for spacious settings using ample leading. Today’s composition families are expected to perform well in a variety of situations, requiring each letter case and figure style to work together harmoniously. By normalizing the relationship between the capitals, the x-height, the ascenders and descenders, and the default lining figures – which involved lowering the height of the caps and shortening the ascenders and descenders – I was able to bring all of the components into balance.
As a result of these adjustments, the lowercase letters naturally gravitated toward a slightly narrower width. Narrowing the face a little made the ascenders and descenders appear slightly longer, thus adding to its classical appearance. Keeping the lowercase round bowl elements looking very round, rather than super-elliptical, helped keep the face from looking artificially condensed. If Minion were wider – with its particular parameters of x-height and caps and ascenders and descenders – it would look a bit odd. I’m particularly happy with the way it turned out. In hindsight, I feel as if I found a sweet spot of proportion which contributes to its practicality.
How did the face acquire its name? Minion is the traditional English name of a small type size, about seven point, but outside the realm of type it has other meanings.
Naming typefaces can be a tedious and frustrating process now, when every possible common name already seems to be taken. A software company the size of Adobe has to have exclusive rights to its product names, and finding a suitable one can be a chore. It was easier back in 1990.
Minion is a face with historical associations, so it made sense to look for a name that had a longstanding typographic connection. Minion just sort of popped out at me, but its attractiveness had nothing to do with any particular point size. Its other meanings could, I guess, lead to some negative connotations. To hedge against that, we described it as “a beloved servant” in the specimen book.
Not that long after Minion was first released, you added a set of ornaments, a Greek, and a Cyrillic. To those of us involved in polyglot typography and multilingual publishing, this made the face much more versatile and useful. But it isn’t the kind of thing most type designers were doing in those days. It also can’t have been planned as a corporate money maker. It can only have been a labor of love.
Yes. But you know, in those early years – 1988 when we started the Originals program, and 1989, when we began releasing original fonts – desktop publishing and PostScript fonts were just beginning to take off. One of our objectives in the early days of the program, as it remains today, was to reintroduce typographic niceties from the past in our fonts while breaking new ground in type design and type technology. We began by adding small caps and old-style figures, and so on. Then as the program progressed, it became apparent that glyph sets would need to expand further in order to accommodate larger global markets, which involved doing more non-Latin families. Greek and Cyrillic were natural early candidates – Greek for its historical relevance alongside early roman type, and Cyrillic for its wide usage and its similarities with Latin types.
As you say, it was also a labor of love. Looking at old books, I’d ask myself, “Could I make a companion Greek for this new roman and italic?” Greek types of the Renaissance often have an untamed appearance compared with roman types of the time. The prospect of updating the style was a challenge, and it was something that really interested me. We were aiming to move the best of classical type design into the modern era by expanding the typographic palette. We were also developing new font technology to move things even further. Everything seemed wide open at that time.
You make it sound easy. But in the industry at large, there was plenty of resistance to such long-range thinking: an obsession with making money in the present and a consequent lack of interest in looking very far into the future or the past. There were technical obstacles too. Except for the alphabet itself, almost everything had to be reinvented, and that took time. I should know this, but I don’t: Was the “expert set” invented at Adobe?
I believe so, yes. I think the idea of including fonts with small caps and old style figures originated during discussions between Sumner Stone and Adobe’s Type Review Board. Fred Brady and others suggested additional miscellaneous glyphs to fill out the set. And since the technology has moved on, maybe we have to explain that an “expert set” was a separate font containing the small caps, ligatures, and old-style figures. Typesetting software wasn’t very smart in those days. Type users had to select these glyphs more or less by hand, but the “expert set” is where you found them. We also introduced additional sets for swash capitals, ornaments, and historical ligatures. And Adobe was very receptive to these additions.
Was Adobe Garamond basically finished when you went to work on Minion, or did these projects overlap?
I began Minion soon after completing Adobe Garamond. I worked on Utopia and Adobe Garamond simultaneously, and it was very stressful to complete them in time for their projected release date. Just when I thought the Garamond was nearly finished, I remember throwing out most of what I’d done, only to begin a new version based on the better photographic source material acquired during our visit to the Plantin-Moretus Museum. Very late in the game I was also scrambling to include swash caps, a titling font, historical ligatures, and a few ornaments.
You’re famous in some circles for that kind of perfectionism. And for more persistent kinds of perfectionism too. Minion had only been out for a couple of years when you rebuilt it as a multiple master typeface. And in 1999 or 2000, you made the first OpenType versions of Minion, folding the Latin, Greek, and Cyrillic, along with the ornaments, the small caps, and everything else into a single font. I studied those fonts pretty closely when they were released, and I was amazed and delighted by what I saw. There was phenomenal attention to detail. For example, all the diacritics were subtly redesigned and repositioned, made a little narrower and lifted farther up above the letterforms. I’m sorry to say it, but in the English-speaking world, most type designers don’t know or care very much about such details. And not everyone takes font upkeep and editing that seriously.
It wasn’t just me. We often receive useful feedback after a font is released letting us know about technical or design problems in our fonts. These requests are filed in a bug report for later revisions, which are often easy to implement when the fonts are periodically rebuilt.
There are other times when I feel it is necessary to rework an existing design. The opportunity to do this usually comes less often, for instance, when we plan to extend the language coverage, or when a major font conversion will take place. But revisions can be problematic in some respects. It’s a pet peeve of mine that when we do revisions, there tends to be resistance to changing the font metrics. The problem, of course, is legacy documents. If the metrics are changed, the text will reflow. If I’m revisiting a font, I often want to change the letterfit. It’s very difficult to make this happen, unless the revision has a distinct name or version number: Minion Pro instead of Minion, for example. Luckily, we’re moving into an era where fonts are considered to be applications in themselves, and font producers and the type-using public are becoming more accepting of new versions of font families.
Besides the enlarged character sets which OpenType made possible, there was the hugely important resuscitation of optical sizes. This again, if I’m not mistaken, is something you pushed for and something Adobe pioneered. In the old days, everyone who knew anything about type knew that different sizes were really different: different in weight, in proportion, in letterfit. Yet when type became photographic, almost everyone forgot there was such a thing as absolute scale. The multiple master format solved the problem handsomely, but fonts were produced in that format for only a few years, 1992 to 1998. After that, you adopted a different solution: building OpenType fonts with fixed optical sizes. But Minion roman and italic, as I remember, existed in two optical sizes – text and display – from the beginning.
Yes, that’s right. I really see my interest in optical sizes beginning with Adobe Garamond when I included a titling version of the regular roman caps, but Minion was issued in 1990 with a display size that included the full character set. This set the stage for doing optical sizes when multiple master technology came into being. I was already thinking about going further with it at that time and this was the perfect vehicle. I also remember that Type Review Board member Stephen Harvard was a huge advocate for doing optical sizes. In addition, we had viewed and photographed several sizes of hand-cut type by Claude Garamond and Robert Granjon during our trip to Antwerp.
And now Minion is set to become one of the largest type families, with multiple versions of Greek, a revised Cyrillic, Pan-African Latin, Armenian, Vietnamese, and International Phonetic. You’ve talked about adding Hebrew and Arabic as well. All this in a large range of weights and optical sizes – and all harmoniously designed. Times Roman has accumulated a huge character set as well, but not under the guidance of one designer. It’s become a big grab-bag of glyph sets that happened to be available. Minion, on the other hand, has matured into a gigantic, unified work. Did you plan this way back when?
I’ve always believed that Minion was the kind of design that lent itself to additional non-Latin variants. The latest plan for extending Minion came from David Lemon and Thomas Phinney, who suggested that we update a small number of our core designs to have even greater coverage than usual, including IPA and extended Cyrillic. This was great news for me: I was very eager to revise the Greek and Cyrillic portions of the family. While there aren’t many faces that have this level of coverage, there are examples of people who have undertaken the task. John Hudson produced Brill, for example. The real challenge is to keep everything harmonious throughout and yet be true to the inherent qualities of the various scripts. A lot of different histories start converging in one place.
Armenian, like Latin, is a script with a lot of tricks up its sleeve. For instance, Armenian has m-ligatures just as Latin has f-ligatures. Depending on the style, it can call for a lot of other ligatures too. In Minion Armenian, you designed a number of alternate forms for descending letters that would otherwise collide in certain combinations – in the same way that f, g, j, and y can collide in some Latin italics. The result is very elegant – so elegant, in fact, that it’s virtually invisible. Beatrice Warde would be delighted – and so am I. When you’re learning a new script like this, how do you go about it? Did you study the language as well as the writing system? Did you immerse yourself in Armenian books and newspapers? Hang out with Armenian friends? Drink Armenian wine and eat Armenian food?
The project actually began rather serendipitously when a type designer named Edik Ghabuzyan offered us his rendition of a Minion Armenian. The prospect of including Armenian in the extension intrigued me, so Edik and I began to collaborate on the design for a time. It was really a process of learning while doing and reading whatever material I could find on Armenian design. Once I began to get a feel for the principles of the script I began to develop it further. After a bit of back and forth with Edik, we came to an impasse. He felt that a highly progressive approach was in order, and I felt the design should be more traditional. So I decided to develop a new font based on Armenian alphabets I’d sketched with a broad-edge pen. These exercises allowed me to connect more directly with the Armenian writing system, and to instill in the letters the humanist qualities that are essential to Minion Latin.
Armenian follows a different handwriting and type tradition than Latin, so I gave special attention to balancing the traditional aspects of Armenian fonts with the practical concerns of making a companion script for the Latin. The trick is to make both scripts set well together while not betraying either one’s classical heritage. Traditional Armenian text forms differ from the Latin in significant ways. First of all, they’re heavier in weight and slightly inclined. They also have a highly angular lowercase stroke pattern. Because there is a precedent in contemporary Armenian fonts for abandoning these somewhat archaic traits, I could focus on other traditional design features that I wished to retain in the design. In addition, because the middle x-height zone in Armenian consists primarily of repeated counter shapes, the script requires more prominent ascenders and descenders to help distinguish letters. So the Armenian projectors in Minion were made longer than those of the Latin to establish good legibility. The capitals are taller too, as Armenian caps require more emphasis in order to have a comparable level of presence. The x-heights remained the same, as any change here would alter the apparent size of the font. The result is an Armenian family that conforms to the Latin in terms of color and rhythm, while retaining the stroke gestures and terminals of traditional Armenian.
With most of the non-Latin faces I produce, I review the established design with native designers. They often shed light on aspects of design that a non-native designer may not be aware of. In the case of Minion Armenian, we contacted the designer and Armenian advocate Hrant Papazian, who offered valuable advice for fine-tuning some of the glyphs. We also worked together to find a solution for dealing with the awkward collisions that naturally occur between certain letter pairs. We reviewed a number of ligatures and contextual alternates that I had drawn as possible solutions, and in the end we abandoned most of the traditional ligatures in favor of a more discreet approach using contextual alternates.
Minion is a humanist face, and everyone who’s interested in typographic history soon learns how humanist ideals have affected the Latin alphabet. With Hebrew and Greek, the situation is more complex. Greek and Latin were closely intertwined throughout the Renaissance. Hebrew not so much, though it was right there in the ghetto, very close by. But in Armenian and Cyrillic, letterform history has travelled a different path that doesn’t run through Florence and Venice and Paris. Artificial scripts like International Phonetic have very little history, period. Can you infuse Renaissance spirit into a letterform just by the way you hold your pen? Or your imaginary pen, if you’re designing on a computer?
I do have an affinity with humanist lettering, so it’s quite natural for me to apply these principles to whatever script I’m working on. That’s true whether I’m working calligraphically or constructing forms directly on the computer. I have to be careful, though, not to apply this approach uniformly to everything. Some projects are essentially devoid of overt calligraphic influence, relying instead on a more architectural style of construction – Acumin for instance. A basic understanding of the writing system that informs any particular style of type is essential when approaching any new project.
When approaching a non-Latin for the first time, I like to become familiar with the earlier manuscript hands, as well as the early printing types. I’m always looking to see if the presence of the pen remains prominent as the writing system continues its evolution, or if other factors begin to affect the shapes of the letters. In a face like Minion, where the influence of the broad-edged pen is clearly evident, the non-Latin counterpart can either take on the design characteristics of the Latin, or it can draw directly from the native writing system or from any other sources that may be in play. I eventually look to calibrate the activity levels of each component alphabet in order to create a balanced effect.
The two variations of Greek in Minion 3 are a case where I designed both a traditional and a Westernized alternative. Traditional Greek, as seen in Renaissance books, has a reversed ductus. The angle of stress is reversed from that of Latin types. Early Greek types are often beautiful to look at, but they often appear somewhat untamed when set next to their more sedate Latin counterparts. In the default Greek fonts in Minion, I kept the Greek ductus but tried to achieve a very pure and principled traditional style, approaching the level of refinement present in the Latin. Conversely, the Western-style Greek in Minion adopts the ductus of Latin. I found it very educational to work in both styles in the same font. My intent was not to make them wildly different in form. I focused instead on making each style feel resolved and authentic within its respective writing discipline.
The idea of unifying scripts with divergent histories is very much a progressive idea. It’s all about balancing authenticity with the practical concerns of unifying scripts that are intended to be used together in a single document. As a Western designer, I tend to view the Latin as the cornerstone of most multi-script type families. The design parameters established in the Latin serve as my basic guide for any non-Latin addition. Scripts like Russian, Greek, and IPA already conform to the Latin parameters of x-height, cap-height, and so on. Other scripts, such as Arabic, Hebrew, and to an extent Armenian, have unique design characteristics that don’t always fit the Western mold. In these cases, it’s important to retain the design features that are familiar to readers, while finding other ways to make unrelated scripts harmonious.
During the last few decades, non-Latins have begun to coalesce into the established Latin type classifications. It is an exciting thing to witness and participate in. Cyrillic is a good case in point. When modern Cyrillic came into being during Peter the Great’s reform of the alphabet, it was reimagined in the style of the latest Latin fonts of the time. This style of Cyrillic text type has persisted until fairly recently. When Minion Cyrillic was first released in 1991, there were few if any humanist Cyrillic fonts available for setting text.
That’s right. There was Hermann Zapf’s Trajanus Cyrillic, made in the late 1950s for hot metal setting on the Linotype machine, and Vadim Lazursky’s Lazurskogo Cyrillic, made very soon afterward for the Poligraf, a Soviet clone of the Linotype. Both were wonderful designs – but almost no one ever bought them. There was a Western European version of Lazurskogo, known as Pushkin, cut for handsetting in the late 1960s, but it was a private typeface. No one but its owner, Giovanni Mardersteig, ever used it. Poligraf also made its own tweaked version of Trajanus, called Ladoga, in 1968. In the 1970s, Lazursky even published a book, in Russian, about Aldine typography, but that didn’t cause any revolution either. Those were the Brezhnev years; the tractor-factory mentality was still pretty strong. In the 1980s, Lazurskogo and Ladoga were both issued as phototype, but many Soviet publishers were perhaps a little afraid of them, even then. Typographic perestroika was a slow and fitful process – and the Poligraf faces only worked on Soviet machines.
The Belgian designer Chris Brand drew a humanist Greek and Cyrillic in the late 1960s, to go with his Albertina Latin, but the Cyrillic wasn’t produced until after Chris’s death in 1999. So until Minion Cyrillic, there was just about nothing: no humanist Cyrillic most typographers could actually get their hands on. It was practically impossible to make modern Russian writers – Akhmatova and Pasternak, Brodsky and Solzhenitsyn – look like they belonged in the larger world. We had to dress them up like Bolsheviks or Tsarists, because no other costumes were available. Typographically, you let a lot of Russian writers out of jail. And Ukrainian and other writers too.
Yes, looking back I was a little naive about the cultural politics within Russian. It isn’t always easy to insert one’s culture into another – there’s a natural resistance. Who would’ve thought that other regions actually have an allegiance to their artistic and cultural heritage?
I’ve often heard you mention the kind of intense concentration required to create a coherent suite of typefaces. Nevertheless, over the years you’ve done many different families of type across a wide design spectrum. You’ve also returned to certain faces, Minion especially, on several occasions. What’s involved in coming back to a typeface, or type family, and picking up the thread after several years of working so intensely on something else? Do you have to get reacclimatized?
I actually don’t get to revisit my typefaces as often as I’d like. If I had my way I’d like to continually tweak and expand my projects at will. It feels like an obsessive quest for perfection that is both a blessing and a curse at times – the curse part is that I tend to be very critical of my own work. Once a typeface is released, I am prevented from tinkering with it further. So when I do get an opportunity to revisit an old project, I typically want to tear into it and fix things. It doesn’t take long for me to get back into a project and pick up where I left off. Once I’ve established the design principles for any font, they remain familiar to me. With each successive project I feel as if I add to my mental Rolodex of font styles and letter shapes that I can instantly access. With every new design I refer to my mental inventory of past designs, both as a resource for ideas and as a guide for what to avoid. It’s a combination of using things from the past and not wanting to repeat too much from the past.
You’ve been with Adobe almost thirty years, and one of these days, presumably, you’ll retire. Can you contemplate leaving Minion behind for others to work on, or would you rather it stayed the way you made it?
Well, I may not have much say in the matter. I wouldn’t want to see my work corrupted, but people may certainly want to add to it. I wouldn’t want to see anyone rethink the things I’ve already resolved. In my view there’s a fine line between identifying and correcting actual errors, and changing something that shouldn’t be, needn’t be changed.
You were a gymnast in early life. And I can see these enormous and very carefully constructed type families of yours as real athletic achievements. I can see in them a certain gymnastic tension and balance of forces. Is there anything in type design that seems to you like gymnastics?
Yes. As in gymnastics, there are long-term goals that you have to feed every day in order to keep moving forward. And discipline – though I’m much more disciplined now than I was in school. Going to the gym every day to practice and refine my routines, and to see what I was capable of doing, while making it all appear aesthetically pleasing — this feels a little like what I do now. I guess this approach applies to almost anything that you build up over time. It’s a matter of discipline and constant refinement. It’s also a process of constantly analyzing and measuring: an attempt to see deeper than the day before.
As a kid, I remember spending a lot time in my room drawing. I don’t really remember now what I liked to draw, but it was something I naturally gravitated to. When I began doing gymnastics, I’d draw out my routines as sequenced human figures. I’d carefully draw out all the body positions, all the moves. I have an inclination toward working with systems, working with things that are developed incrementally to a point were nothing seems out of place to me. Developing font families opens up a similar world of possibilities. When I came to Adobe, I thought, Okay, now you can do that, you can really do that. I was all over it.
