dot-font: Can a Book Teach You to Set Type Perfectly?
The purpose of James Felici's new book, "The Complete Manual of Typography" (Peachpit Press, 2003), might best be expressed in its subtitle: "A Guide to Setting Perfect Type." Felici insists that he's showing how to set type well, not why to make particular design decisions. There may be no such thing as "perfect" typesetting -- experts disagree about which practices are best -- but a thorough explanation of the factors involved in mastering this craft may give users the tools they need to make their own decisions and to end up with good results.
Reading this book is like sitting down with a long-time typesetter and going over the details of a complex job, pausing to ruminate on underlying principles and everyday rules of thumb, with occasional digressions to examine a fine point to the participants' mutual satisfaction. Most people will use it as a reference -- which it is -- but reading any section straight through is rewarding. The writing is clear and straightforward, and Felici has obviously thought long and hard about everything he deals with here.
Well, almost everything. I noticed a few areas that got short shrift, such as the proliferation of humanist sans-serif text faces or the complexities of setting type in languages other than English; and a few things that I'd have liked to see considered, such as the effect of the letter design itself on the best way to space a spiky, high-contrast text face like Didot or Bodoni. But on the whole, this is an admirably thorough book.
It's a how-to book, divided into three sections: about a hundred pages of "Typographic Basics," a much longer main section on the details of "How to Set Type," and a final section comprising a detailed glossary and index and a very truncated list of suggestions for further reading.
Rules of Thumb
Felici gives his readers a lot of useful advice about specific problems, and he even includes a couple of handy suggestions that I hadn't encountered before. One of them is a rule of thumb for calculating how much extra leading to use with a typeface, depending on the length of the line: "simply divide the measure (in picas) by the size of the type (in points)" and round off the result to the nearest half point. Add that number to the point size, and you've got a reasonable leading value for a block of text of that width. If you make the measure wider, then the leading gets deeper, too -- which is as it should be, so the text will be easy to read. Obviously, this formula is just a rough-and-ready guide, since different typefaces look best with different amounts of space between the lines -- but it's a start.
I don't always agree with Felici. For instance, he asserts that sans-serif text faces are inherently less legible than serif faces. "The difference isn't vast," he says, "otherwise sans-serif faces would have dropped off the map -- but it is considerable." This is like saying that men are taller than women. As a statistical average, it's quite true -- but the most interesting variations take place in the huge area of overlap. There are simply more important factors in determining whether a typeface will work well in text than whether it has serifs or not.
That said, I'm interested in Felici's suggestion that sans-serif text faces usually work best when set to a narrower measure than serif text faces. This is another of those generalizations that will vary from typeface to typeface, but it's a useful way of approaching the problem.
And I'm firmly in the camp he pooh-poohs, of typographers who believe that justified type looks best when set with no variation at all in the spaces between letters. But Felici always comes down firmly on the side of the reader; when he allows slight variations in spacing, he does so only "as long as the fundamentals of legibility and readability are honored, and this means that such variations should be kept to a minimum." The nature of typesetting text is that you are constantly confronted with intractable problems, where no solution works perfectly; the art is in finding the best out of a bunch of imperfect answers.
For that reason, even hard-and-fast "rules" such as "no more than two hyphens in a row" or "never hyphenate the last word in a paragraph" are subject to review. Sometimes it's less disruptive to break one of these rules than to hew to it slavishly and make a mess someplace else. Felici is very good at showing the complex consequences of our decisions in setting type.
Chapter 10, "Controlling Hyphenation and Justification," may be the most thorough explanation of how a page-layout program actually sets type that you can find outside of an engineering spec. The example on pp. 134-136, "How H&J Works," where he imagines the decisions that a text composition program makes as a sort of internal dialog, is very clear and to the point. It doesn't, of course, delve into all the possible branches of the decision-making process, but it makes the nature of that process plain.
The one thing Felici doesn't do, in most cases, is name specific page-layout programs and explain how they differ. I suppose he chose to do this in order to avoid getting bogged down in the ever-changing details of new software versions, but it would have been helpful to see, for instance, how QuarkXPress and PageMaker and InDesign differ in their approach to text composition.
The Book Itself
The book's design doesn't live up to its content. It has the big, square format of a computer manual, which allows lots of room in the margins for illustrations but makes the book itself awkward to read (see figure 1).
And in the glossary and the index, where there are no illustrations, the huge outer margins actually get in the way: they're so deep that it's hard to see where you are as you flip through those pages (see figure 2).
I'm certainly surprised at the choice of Perpetua as the text typeface for this book, and at the small size of the text. (Perpetua has such a small x-height that it appears even smaller than it really is. But the perception is what counts.) In its original form, as a hot-metal typeface for Monotype typesetting machines, Perpetua looked quirky but had the strength to hold a page of text when it was printed letterpress; as a digital adaptation, however, Perpetua looks light and spindly and blotchy on the page. It gives this book a look of preciousness that belies its basic practicality. The text composition is good (not surprisingly!), but the design choices seem odd.
Although most of the book is about the specifics of typesetting, "The Complete Manual of Typography" does go into broader topics like the design and layout of indexes and tables of contents (in chapter 14), which overlaps with the sort of thing covered in the "Chicago Manual of Style." Felici's manual is more of a one-man show, but the two books might usefully sit next to each other on your reference shelf.
The danger with a book this authoritative is simply that readers will take the author's word as gospel, and follow his guidelines blindly -- whereas what Felici would obviously prefer is that they use his knowledge and his recommendations as a jumping-off point for their own explorations of the art and craft of typesetting.
Read more by John D. Berry.