Loading ...

Not so long ago, when I wanted to set a Greek or Russian quotation, I had to go through laborious shenanigans, synthesizing missing characters and tricking uncooperative software into relaxing its iron group. Now multilingual typography is comparatively easy. There are good Pan-European fonts, capable of setting several hundred different languages that use some variant of the basic Latin alphabet, and 50 or so that use some variant of Cyrillic, and both modern and classical Greek. Good Chinese, Japanese, Korean, Thai and Hebrew fonts are available. So is the software needed to run them. More complex scripts, such as Arabic and Tibetan, which were never really comfortable in metal, are still a serious challenge to the computer, but solutions for these are also coming along. And most of us working with type in a serious way now have digital foundries on our laptops. When we need a new glyph, we can make it with digital tools nearly as well, and almost as quickly, as our forefathers could using files, gravers, candle black and steel. The multilingual flexibility that typography possessed 500 years ago has been more or less restored in the last 10 years through software advances.

I’m grateful for these resources. I have more cause for celebration than for complaint, but something is still missing. That something is multicultural vigor. The best multilingual typography of the sixteenth century possessed a physical and intellectual vitality that is rarely achieved today.

The problem is, our solutions are too pat. The multilingual work we do—which was often unbearably crude a decade ago—is more often now too smooth, too clean, too seamless. Many Pan-European fonts are in fact so smooth, so clean, so self-consistent, that a reader barely registers the shift from one alphabet to another.

That, by and large, has been the designers’ and manufacturers’ conscious intention. They seem to want a harmonious world, where trade, information and money flows freely: an orthographic counterpart of the European Union, where most of the internal borders have vanished. Are they wrong? I’m reluctant to say so, but their decision comes at a price.

Latin and Greek types of Nicolas Jenson, Venice, 1472, with the same lines reset in Adobe Jenson Roman and New Hellenic Greek.

For better or for worse—and it was mostly for the better, in my opinion—punchcutters and printers in the fifteenth and sixteenth centuries took for granted that every written language had a visual tradition of its own. Along with its palette of sounds, represented by the letters used to write it, every language has its rhythms and intonations: its native accent. And writing tries to measure up to speech. A language that sounds different from others, as every language does, has a reason for looking different too, and once upon a time it usually did. The scribes of different languages and regions had their own distinct repertoires of scribal habits and gestures, their own ways of trimming and holding the pen, their own graphic vocabulary of ligatures and swashes and punctuation marks. There were scribal hands and type fonts specially associated with French, German, Spanish and Dutch, even though they shared a common alphabet with Latin and Italian. Where the alphabet changed—from Latin to Greek or Hebrew or Cyrillic, for instance—scribal conventions changed even more. And typographic conventions followed the lead of the scribes.

Nicolas Jenson, Simon de Colines and other great Renaissance typographers happily mixed Latin and Greek on the printed page. When they did so, they placed not only different alphabets but different scribal traditions side by side. They wanted, as any typographer would, a balance of size and a balance of color—equal degrees of blackness in different blocks of text—but not a misleading sameness of texture. They wanted the Greek to look and feel like Greek, and the Latin like Latin. The idea that Latin and Greek faces should look like parts of the same big, transnational alphabet, or that they should look like they were written with no change in the posture of the hand, would have seemed absurd to a Renaissance printer. But the alternative wasn’t disharmony; it was lively coexistence.

Colines, who was born around 1480—the year in which Jenson died—succeeded in matching the weight and color of Latin and Greek better than any typographer had ever done it before. His Greek and Latin capitals do indeed form one big alphabet, but these are inscriptional letters, evoking a shared tradition of drawing and carving. The lower case in Colines’s Greek fonts differs at every point from the lower case in his romans and italics, because the Greek fonts embody a fundamentally different approach to writing. Yet these Greek and Latin fonts function perfectly together, producing harmonious textures and proportions. The result is a kind of typographic polyphony which very few of us achieve.

Latin and Greek from a two-column page by Simon de Colines, Paris, 1530, with the same passage reset in Robert Slimbach’s Poetica and Matthew Carter’s Wilson Greek.

Over the next 300 years, a number of skilled punchcutters continued this tradition of balancing Greek and Latin without trying to homogenize them. Robert Granjon, Miklós Kis, William Caslon and his son, and the excellent eighteenth- century Scottish punchcutter Alexander Wilson—who was also a professional astronomer—all belong to this lineage. Many Latin types in current use are historical revivals or derivatives of the Latin side of this tradition: Adobe Jenson, Monotype Bembo and Poliphilus, Adobe Garamond, Linotype Janson, Adobe Caslon and many more. But there are only two digital Greeks in circulation that represent the wonderfully varied other half of the same historical enterprise. They are Victor Scholderer’s New Hellenic (now available from the Greek Font Society) and Matthew Carter’s Wilson Greek. Though these two fonts look nothing like each other, they have one important thing in common: They sustain an authentically Greek scribal and typographic tradition. They make no attempt to look and feel like Latin.

Cyrillic too was once a script with its own distinct scribal and typographic tradition. The Cyrillic that Robert Granjon cut at Rome in 1582, based on Russian and Serbian models, looks nothing like his romans or italics, but it can happily share the page with them. Cyrillic type, however, was given a sudden and powerful push in the Latin direction at the beginning of the eighteenth century. In 1702, Louis XIV published—mostly for other monarchs—a lavish book describing his royal medals, printed in his new Neoclassical French type. The czar, Peter the Great, responded by commissioning a new Neoclassical Russian type to match. Cyrillic and Latin have moved on largely parallel tracks ever since.

Greek script and type were westernized much more slowly, with no important government intervention until the script reform of 1982. It was then that polytonic Greek script, with its many diacritics—used by classical scholars for 2,000 years—was replaced by the monotonic system, which uses only the acute and the diaeresis. Classicists, of course, were unaffected by this decree, and have continued using polytonic script. But even classicists have not been immune to centuries of creeping change in Greek letter shapes.

The movement to westernize Greek, and make it look as much as possible like Latin, reached a peak around 1930. Eric Gill’s Perpetua Greek, designed in 1929, and Jan van Krimpen’s Romulus Greek, designed a year or two later, are both based on the idea that Greek type should abandon any claim to an independent scribal tradition. These were the first attempts to produce Pan-European families of type, and many of the Pan- European digital fonts produced in the past decade, using OpenType technology, follow the path marked out by Gill’s and van Krimpen’s experiments.

For years, in Russia and Greece, indigenous type design has been heavily overshadowed by Cyrillic and Greek “adaptations” of standard and popular Latin faces, from Baskerville to Helvetica and Souvenir and beyond. These adaptations are usually locally made; it is only the models that are imported. But classicists, who use written language to stay in touch with history, cannot use these faces. We still need what Renaissance readers and printers needed. We need Greek to look like Greek, Russian like Russian, and Latin-based scripts to look and taste like Latin.

The letter omicron is the Greek short o. It looks and even sounds like Latin o, and it has the same historical root. No surprise, then, that omicron and o can be identical in form, so that only context makes one Greek, the other Latin. In most of the Pan-European fonts produced in recent years, omicron and o are indeed identical.

Left: Three twentieth-century Greek foundry fonts: Eric Gill’s Perpetua (1929), Jan van Krimpen’s Romulus (circa 1935) and Hermann Zapf’s Heraklit (1954). The Heraklit text is reset in both Palatino Linotype and Palatino Nova Greek. Middle: Latin and Greek in three recent Adobe OpenType families designed by Robert Slimbach. Right: Claude Garamond’s 18-pt Greek (Paris, 1543) with the same passage reset in Robert Slimbach’s Garamond Premier Pro Greek. (The typo—eta for epsilon—at the beginning of line four of the original is corrected in the reset version.)

The Greek letter eta—the Greek long e—looks something like an n with a lengthened tail, and there are many Pan-European fonts in which eta and n are as much alike as possible. From a classicist’s point of view, this is not a good idea. Eta and n have nothing in common, and their resemblance is accidental. They should look and taste quite different from each other. In some Pan-European fonts they do. (Robert Slimbach’s Arno and Garamond Premier, Hermann Zapf’s Palatino Nova and Gary Munch’s Candara are examples.) In many other such fonts (Slimbach’s Minion, Warnock and Myriad, for instance), eta and n are as alike as they can get.

All the Pan-European type families I can think of include an italic, and here some other problems arise. In both Latin and Cyrillic, the contrast of roman and italic (or upright and cursive) is an old, established tradition. For years, cursive Cyrillic type was rare, but cursive handwriting was practiced, and that is how the tradition survived. In Greek, however, upright and cursive remained alternative and independent styles, developing on different lines and almost never paired. Most Greek type is cursive to begin with, though it often has no slope. Italicizing it makes no scriptorial sense, and is usually accomplished just by forcing it to tilt. The result is more often a parody than a variant. Noncursive Greek letterforms (New Hellenic is an example) have no cursive counterpart, typographic or handwritten. They too can be forced to tilt—but again, the result is usually parody.

One of the great successes in recent Latin type design is Robert Slimbach’s Adobe Garamond. It was first produced in the late 1980s, based on roman letterforms cut in the early sixteenth century by Claude Garamond and on italics cut some decades later by Robert Granjon. When the family was first updated to OpenType technology, in 2000, East European Latin glyphs were added, but no Greek or Cyrillic. In 2005, the design was issued again, revised and enlarged and renamed Garamond Premier. In that form it does include Cyrillic and Greek.

Garamond Premier Cyrillic was what the market said it wanted: something as close as possible to the Latin face, and nothing like Granjon’s Cyrillic. The Greek was more experimental. It owed something to the study of Greek fonts cut by Garamond and Granjon, yet it looked nothing like those models. The Greek fonts cut by Garamond and Granjon involve hundreds of alternates and ligatures. OpenType could handle this complexity, but Garamond Premier Greek has turned the other way. It has only three or four alternates and no ligatures at all. Its merit is that it differs fundamentally from its Latin companion. Even omicron and o are quite different. I have used it with pleasure to set Greek texts, though I cannot use it to give them a sixteenth-century flavor.

When I try to bring multilingual typography up to sixteenth-century standards, I find that none of the Pan-European fonts is what I need. I am far better off mixing and matching, picking the Latin faces I need from one source and picking the Greeks from another. ca

Robert Bringhurst is a poet, typographer, linguist and cultural historian. He has published more than fifteen books of poetry, including Bergschrund (1975), The Beauty of the Weapons (1982), The Calling: Selected Poems 1970–1995 (1995) and New World Suite No. 3, a poem in four movements for three voices. Many of his lectures are collected in The Tree of Meaning: Thirteen Talks (published in Canada in 2006, scheduled for U.S. release in the spring of 2008). Other works include The Solid Form of Language: An Essay on Writing and Meaning (2004). Bringhurst wrote the Typography column.


With a free Commarts account, you can enjoy 50% more free content
Create an Account
Get a subscription and have unlimited access
Already a subscriber or have a Commarts account?
Sign In

Get a subscription and have unlimited access
Already a subscriber?
Sign In