Digital Web Magazine

The web professional's online magazine of choice.

The Destination Matters More Than the Journey

Got something to say?

Share your comments on this topic with other web professionals

In: Articles

By Dean Allen

Published on July 10, 2001

I've often been asked (alarmingly, by senior designers and at least one creative director) to recommend a good introductory book on typography. Many excellent guides exist—Robert Bringhurst's The Elements of Typographic Style is likely the most complete; it's also a deeply satisfying read—however I usually offer that those with a nascent interest in typography would do better to start with the book Bringhurst used as a structural model, the timeless (if shabbily typeset) manual on writing with clear, vital language, The Elements of Style.

Strunk and White's small masterpiece has a good deal more to offer than rules of verb tense and sentence structure, even to mere designers like you and me. In its haughty pronouncements and admonitions there's a solid underlying logic, one that might be oversimplified thus: build a sound structure, then give it life.

It may be said that typographic structure, like that of writing, is built from parts both familiar and new. Letters, numerals, dots, squiggles and lines have inalterable meaning in the eye of the reader; each new combination and arrangement of these parts carries new visual and linguistic meaning. Good designers know that this meaning can thrive or wither in the crucible of publication, a fate decided by two monumental factors: a reader's desire to be informed, and a reader's willingness to be enticed.

It makes all kinds of sense for designers to give first priority to being informative, so much so it hardly seems worth mentioning. Until, that is, one takes a spin through the visual landscape—in which information frequently suffers at the hands of designers who see content as the raw material with which to satisfy a need for attention and praise, or, worse yet, as an amorphous quantity of stuff to be stylishly crammed into the space available.

It can't be said that this disrespect is new development, nor is it due to the tools and media in use. Hack designers were around in 1501 Venice, in the Paris of the Didots, in Stanley Morison's London. What was true then, is true now: designers more interested in process than outcome do shoddy work.

Design software tools available now—regardless of intended medium—are ludicrously powerful, offering options and efficiency and the means to produce great work with speed. Most software publishers, however, have paid scant attention to how their products can be used to produce lucid, engaging text.

Carefully-won standards of practice hardwired into the DNA of any capable typographer are utterly flouted in expensive, industry-standard design programs, requiring constant override and compensation, forever demanding compromise. To the question of why bodies of text look so artless and clumsy in print and on the Web, one could shrug and whisper that it's just too difficult to make it look good with the tools at hand. And, so long as it's legible, what does it matter? It's just text.

"It's just text!" So goes the college cheer of the hack designer.

Readers want to read. Get out of the way. Justify their interest.

Just as it's reasonable to expect text to be in one's own language, with the words in the right order, it's reasonable to expect to read unhindered. A welcoming block of text is one of even colour—that is to say, consistent in overall tone: dense settings give off darker colour, loose settings are light. The ideal colour is defined by personal taste, but evenness and consistency go straight to the goal of unhindered reading. There's a great deal going on as the reader navigates lines of text, gathering images made of letters, absorbing meaning. The eye demands a clear path, the building blocks for which good type designers provide, while most typesetting environments (such as the aforementioned software tools) don't.

Well-made typefaces are designed with consistent spacing in mind: between letters, words, sentences, and lines. Each adjacency of these elements contributes to the whole; each instance of a shape carving into that which surrounds. Design software and CSS code provide means for global adjustment to these pairings, and individual instances of tweaking can always be done on the fly. But all of this is meaningless if, upon final presentation, software arbitrarily inserts and removes space in the interest of questionable goals such as justification, or if instructions are ignored entirely.

This is perhaps the primary hindrance to well-made text at present. Spacing instructions are ignored in applications such as Microsoft Word, in presentation software such as PowerPoint, and most all Web browsers (with Netscape 4 being a particularly egregious offender).

Graphics programs like Photoshop and Flash are quite good at respecting spacing, though no one would confuse these programs with having anything to do with reading.

For the designer's part, the best way to ruin spacing is to justify text—forcing the software to insert and remove space between words and letters—especially in narrow columns, and inevitably in Web settings: the damage that a browser inflicts on crucial wordspacing to make both sides of a column smooth—though browsers get that wrong too—is unspeakable. The only solution, then, is to left-justify text, leaving the right column ragged to pick up the slack of leftover space.

Some will argue that fully justified text is required in certain situations, that editorial convention demands it, that justification implies a certain formality in contrast to the casualness of ragged setting... and this is bunk. It's true that certain reading situations benefit from justification (usually it's the white space around that benefits rather than the text) but, when looking at columns of ruptured, ruined words, one must ask which element deserves the best treatment, the text or the design.

While letter spacing—rendering space between letters elastic—is ruinous to body text, especially to lowercase letters, it is essential whenever capital letters and small capitals stand adjacent. The historic beginnings of what are now capital letters were of monumentality, of being carved in stone, and as such these beefy letters have a visual force which comes in handy when indicating the beginning of a sentence, but is compounded when more than two gather at once. So with groups of capitals in titling and subheads, or with small capitals in the flow of text, it's important to separate these letters with extra space. Usually only a little is required: it's easy to arrive at an adequate amount of letterspacing when evaluating how the strings of letters affect the colour of the surrounding text. The letter-spacing attribute is part of CSS1 as defiined in the Technical Recommendation.

The space between lines of text—still called leading by some, though line increment might be a better term—deserves careful attention, and is never adequately arrived at by default settings (it's worrying to the soul to know that millions work for hours a day with text in solid leading: 12-pt text on 12-pt lines, ascenders and descenders duking it out in no man's land). Line increment in web documents is easily adjusted with stylesheets by employing the line-height attribute.

Seeing contrast on its own merits

Typography components of a communication design education often dwell on catch-all graphic conceits like contrast and harmony, proportion and scale. As valuable as these ideas are in establishing wayfinding systems and hierarchies of information—shouting above the din, if you will—these visual principles aren't the best starting point when designing text.

Again, a main goal of the designer should be to help in the navigation of information. Design instinct, however, sometimes veers toward overstating the obvious. Think of a 72-point headline in a bold sans-serif face, a web banner with bing-bong sounds, flames and... oh, let's throw in the dancing minions of a corporate antichrist for good measure.

These are extreme examples of a frequent error in design judgement. The problem with shouting is that it can't be ignored. Reading a nicely crafted body of text while strutting and preening, if ostensibly helpful, aids to navigation are constantly calling the eye away is a hallmark of design that's about the designer, not the subject at hand.

Another aspect of typographic contrast many designers fail to acknowledge or utilize is meaning. Words have meaning registered by those who read, rather than merely look at, designed objects. By way of example, a small bit of text that begins with the word 'chapter' already has at least one parameter of distinction from that which surrounds it.

Some designers and writers tend toward theatrics in their work, wanting to add punch and emphasis to a point. Many choose boldface letters, or add drop-shadows or slanting or underlines or neon outlines or animation. None of these additions will (apart from irritating the reader and negating above-stated goals) add liveliness if there's nothing lively to begin with. In other words, good patter is being used to hide bad matter.

Making (short) demands on the reader

Uneven settings and high contrast require the reader to overcome hurdles, sometimes deliberately imposed. The practicalities of available space to contain text serve up other potential problems.

Reading is rhythmic and intuitive. Start at the top, leave at the bottom, begin again at the top; enter at left, exit at right, return and begin again one line down. The rhythmic movement through typographic space implies limits to that space: lines of text that are too long require extra effort to find the beginnings of subsequent lines. Text on a web page in a maximized window (already suffering from myriad indignities) that spans from left edge to right is simply murder to navigate. True, you can always print, but why make the demand?

Numbers have been batted around for centuries over ideal character count on a line. Somewhere between 60-70 (with spaces) seems to be an ideal in ink-on-paper. On the screen, however, already a quaking minefield of distraction, the number is closer to 40 characters per line.

Remembering certain goals

Typography is idealized writing. It is also recorded speech. A good designer will inevitably be fascinated by visual and verbal language both, and respect the typographic goal of making visual the patterns and logic of language. Keeping in mind the goal of minimizing disruption on the page, analphabetic marks frequently require extra attention. The North American habit of using double quotation marks to mark dialogue, for example, will perhaps one day be abandoned in favour of the much less fussy British method of using single quotes: this small change greatly reduces the noise on a page heavy in dialogue. And the time is coming to abandon the Victorian-era eccentricity—used extensively in this article—of using the em dash to set off phrases. An en dash—with normal word spaces on either side—is much more poised, and less ruinous to the quality of text.

A concern to keep in mind when evaluating punctuation marks and other modifiers in digital type is unwelcome collisions and unresolved alignment between letters and brackets, braces, parentheses, quotation marks, and ellipses.

Selecting type

If there is guiding principle in the intelligent selection of type, it begins not with stylistic appropriateness but with practical value: can the typeface you choose do the job you require? Basic requirements for a working type family might be a full complement of roman, italic and small caps faces. If you require mathematical symbols or fractions, you need a family that furnishes these. If you need characters from foreign languages with accents and other modifiers, or possibly need to invoke Cyrillic or Greek text, then you'll need a family with these parts.

Letterforms, when arranged into sentences and paragraphs and instructions on medicine bottles, have a voice of their own which can either shout or whisper. They can also harmonize or clash with the voice they convey. The design of type betrays much about the politics, philosophies and stylistic ways of its point in history, characteristics that can help or hinder: a book on classical French cooking will hardly benefit from being set in boiled-mutton Victorian English letters. Choosing an ideal typeface or group of faces capable of the job at hand, out of the tens of thousands available, may seem ungainly at first, but infinite Web resources and many books will steer the way.

It's crucial to learn the distinction between type designed for print use and type intended for tube (monitor or television) display. Out of the box, many operating system and software installations will include core sets of fonts. In recent years some have been quite good screen fonts: Georgia and Verdana, both by Matthew Carter, are quite dependable when used as intended, which is to look reasonably good on a coarse computer monitor. These faces don't print well, indeed most everything that comes with an OS or browser install currently won't print at all well, and most are junk on screen as well. Note that I'm not naming names here.

On digital vs. analog

Raging debates among music enthusiasts over digital vs. analog reproduction are paralleled in typography: where audiophiles long for vinyl in the age of the compact disc, students of type prefer the letterpress page to one made entirely from digital means. The process of raised, inked metal being pressed into paper is three-dimensional, sculptural, unpredictable and organic: in letterpress pages there is ambience and vitality. Alongside smoothly calendared paper decorated with ink plopped off rubber sheets at increments specified with cold, digital precision, such pages betray an almost lost ideal, one of text that has force, has life.

Go grab one of your Tufte books and compare the body text—handset in metal Monotype Bembo—to a page set in the pale, sickly Bembo available digitally.

Some new digital type, notably from Enschedé and the Dutch Type Library, has been coming close to that vitality on the page, but so much of what is available to designers now suffers a similar fate to many photographs after being run egregiously through Photoshop: one looking just as bad as the next.

On 'classicism' and the bleeding edge

Early in every designer's career there exists an instinctive desire to experiment, to innovate, to try the bounds of established style; my own early work offers textbook examples of what can go wrong before experience and sobriety kicks in.

Innovation—finding something new in existing arenas—is thrilling. Misfired innovation, unaware of its surroundings, simply is not. Innovation for its own sake has its applications, if only as a means of self-help.

Fashion and the cult of the new are forever nipping at the heels of typography, at least in its applications for adding market value to products. What's tangibly innovative among those with a claim to typography, however—what fills me with joy every time I come across it—is the ceaseless curiosity required to learn from historical example rather than treat it as a convenient archive of lifestyle-specific clip-art or as a blanket definition of 'classical' design; the humility to step back from the work long enough to see if the train is making it to the station; and, above all, the tenacity required to build a sound structure, and then give it life.

Got something to say?

Share your comments  with other professionals (0 comments)

Related Topics:


Dean Allen is a writer and designer who publishes the websites Textism and Cardigan Industries, to which there is not much point. He sometimes orders pancakes, just for the pleasure of cutting them up into uppercase letterforms.

Media Temple

via Ad Packs