ON THE CHOICE OF TYPEFACE
author: Beatrice Warde

The legibility of a typeface has an exact parallel in the audibility of a human voice. A lecturer must make every word audible and distinct; yet within the limits of audibility lie the whole range of speaking tones from a metallic monotonous drawl to the infinitely flexible and persuasive tones of the good speaker.
Type, the voice of the printed page, can be legible and dull, or legible and fascinating, according to its design and treatment. In other words, what the book-lover calls readability is not a synonym for what the optician calls legibility.
In choosing a type design for book printing the problem of ocular legibility has in most cases been solved in advance; that is, it is very unlikely that a typefounder or composing-machine manufacturer would produce and offer to good printers a face of which any two characters had confusing similarity, or in which any one letter ignored the "code" which governs its design in roman or italic. The size must be chosen in view of whether the work is one of reference, that is, to be read in short sections by people who are concentrating, or a novel to be read uninterruptedly by people who are enjoying themselves, or an educational book for young and reluctant eyes. Here again the makers are not likely to cut a small size so small as to be "illegible"; though any size may be called "unreadable" when it is too small or even too large for a given purpose—a reader's, not an oculist's purpose.
The moment the question shifts to readability, however, these elementary precautions give way to endless and delightfully varied experiments no less effective in each minute difference than is a change of timbre in a speaking voice. Set a page in Fournier against another in Caslon and another in Plantin, and it is as if you heard three different people delivering the same discourse—each with impeccable pronunciation and clarity, yet each through the medium of a different personality. Perhaps the layman would not be able to tell one old-style setting from two others of the same group; yet he could not read the three pages in turn without at least a subconscious discrimination. The smallest variation in serif-construction is enormous compared to the extent to which a disc of metal, in a telephone receiver, vibrates to electric shocks produced by one voice and another; yet we find it easy to deduce from one such set of vibrations that an old friend is asking us to "guess who this is!"

Physical Considerations

The beginner in book typography is prone to import aesthetic sentimentality into what is first of all a matter of convenience. Baskerville and Fournier were both designed during the eighteenth century, and some people think that they represent in miniature, and in terms of their respective national cultures, the clarity and good manners of that age. But should you label an old or modern author "dix-huitieme" and start matching his words to what you consider a type of the era, it would be better first to remember that Baskerville, being relatively generous in set-width, will "drive out" the book; whilst Fournier, a neatly condensed face, will be more frugal of space. Thus Pride and Prejudice by Jane Austen, produced by Peter Davies Ltd., had a large amount of text to begin with, and not too many pages were to separate one illustration from the next: Fournier, in a beautifully legible small size, solved the problem. Baskerville, conversely, printed on bulky paper, has saved many a fine book from seeming to offer less than the money's worth. The typographer, whether he be connected with the printing or publishing office, should be able at a few minutes' notice to calculate ("cast off") how many pages the copy will come to in a given face, taking into account the point size, set-width, number of lines, and leading between the lines.
The word "set," which appears in our type specimen books, means that an actual type of the widest letter in the font (such as cap. M) will be as many points wide as the num­ber given, and that the narrower letters will be proportionate to that width, if the type is cast on that indicated "set." Thus a font like Baskerville, of which the 12-pt. is "12 set," is going to occupy more space, word for word, than Bodoni 135, 12-pt., which is 11 3/4 set; Centaur and Garamond 12-pt., which are 11 1/4 set, will take less space; and Fournier, which in 12 pt. is 10 1/4 set, will vary from the width of Baskerville by the proportion of 10 1/4 set to 12 set.
Some typefaces are more successful in the sizes above 11-pt. than in those below it. The fine cut of Bodoni demands injustice exquisite printing for 6- and 8-pt.; Caslon and Garamond seem to many to improve as the sizes increase. Aldine Bembo 270, used on these pages, is one of the new Old Faces that preserves all its freshness and charm in the smallest sizes. Fournier and Monotype Plantin, for different reasons, are highly successful in the smallest settings as well as in and over the normal sizes. Centaur is a fine type in any size, but certain subtleties of cutting cannot be appreciated below 24-pt., and these details go to make it as successful an upper and lower case for poster work as has ever been designed.
If the quality of paper is known in advance—as it must be in most cases, and especially where illustrations are used—this will influence the choice of a typeface. Old Face was not designed for calendered paper, which did not exist until Baskerville's experiments; the difficulty arises in the fact that a smooth-finished surface of paper takes the inked copy with such ease that little or no impression into the fabric of the paper is necessary, and, therefore, the only ink which comes off the type is that on the actual printing surface. In general, calendered or shiny surface ("art") paper needs such a face as Monotype Plantin, which is not noticeably thinned down by such treatment.
In the old days a printer had no reason such as these for stocking different typefaces. He worked on one kind of paper: handmade pure rag, with the corrugated surface left by the wires of the paper-mould—a surface now known as antique laid. He had only one process by which pictures and type could be printed simultaneously. Nowadays he also has to be the master of a process as different from the old type-printing as the "kiss" impression of thousands of shallow dots of metal on smooth paper is different from the pressure of a deep-cut type and wood-blocks into damped paper. The modern printer is versatile, as his ancestors never dreamed of being; he prints from a rotary as well as a flat surface, and often from rubber or copper cylinders. He has long recognized the necessity of using a special kind of paper for each process. Nowadays, if only to prevent set-off, he has learned to stock special inks for special papers. But some survival of craft tradition prevents many printers from realizing that a face, like an ink or a paper, can be suitable or unsuitable for a given process. There is still a widespread feeling amongst them that the typographers ought to settle on one perfect type, and thus eliminate the expense of stocking, not that one font of 12-pt. which the old printer would call simply "our pica," but at least three or four different sets of 12-pt. composition matrices—chosen, be it noted, not for aesthetic reasons, as all can be "good" designs, but for as practical reasons as hold good in the paper-stores. Quite apart from the survival of the "One face" tradition, there is the fact that a composition series costs money. It is therefore necessary for laymen, buyers of printing, to discipline their enthusiasm for new faces.
If a Monotype user has four body composition faces, and each is well-designed and adapted to a particular printing process, and if the four designs are sufficiently different to convey four different "tones of voice" it would be inordinate to expect that man to increase his type repertory without very good reason. A customer can confer a great benefit upon a hitherto undistinguished printing office by clamoring for one fine composition face where there was none before; but on the other hand to wave aside Bembo and insist on Centaur or vice versa, is an ungrateful act. Besides, if there is a really defensible necessity for Bembo in that particular job, why not reward the master printer who, independently and of his own judgment, invested in that type without being prompted? In short, the man who wants a choice of good typefaces must go where they are or else accept what he is offered—unless he is willing, in token of his sincerity, to go shares with his printer and help purchase that font.
Let us leave this matter with the admonition that most old faces look anemic on coated paper, that a few types like Plantin Light 113, Ehrhardt, Bembo, Imprint, and Bell, are adaptable to varying processes, and that no printer ought to put in a composition face except that a number of customers over a number of years may be advantageously served by it.
One other mechanical point in the choice of typefaces has to do with combinations of different alphabets. Nowadays italic is thought of as a part of the whole font loosely called "roman," but the appearance in a page, or even a long sentence, in italic would show why this form of letter, at least until the middle of the sixteenth century, was considered as an entirely separate alphabet. When italic was thought of as separate cursive a certain latitude and individuality was allowed to it. Garamond italic, for all its whimsical and charming irregularity of slope which lends piquancy to certain italicized words, does not invite the effort of reading in entire poems or paragraphs as well as the disciplined Baskerville or Bell italics. The kind of cursive called Chancery, to which family Blado, Bembo, Arrighi, and Lutetia italics belong, has such beauty in its own right as to justify its use in long passages or even whole books; and as far as combination is concerned, there seems to be a closer correlation between the Chancery letter and the essential form of Roman Old Face than can otherwise be found before the eighteenth century. Another question in regard to combinations; Is an exotic font to be used anywhere in the text? If so, neither it nor the body roman must be too discrepant in weight, serif treatment, and general appearance. Perpetua is one of the few types which may be said to have "a greek" in the sense that most romans have "an italic"; in general one must be one's best to see that a warm Renaissance letter like Poliphilus is mated either with New Hellenic (for color) or the Aldine Greek, Series No. 283, rather than with a Greek cursive of the brilliance of Didot's. Even the extent to which capitals are used has some bearing on the choice of typefaces. The almost superstitious regard for Caslon Old Face has been such that only a typographer of our own time has dared to point out that its capitals, especially the capital M, are so heavy in contrast to the lowercase that very frequent use of them on a page creates a spotty effect.
And still we have not reached the really interesting part of choosing a typeface. All this preliminary matter has consisted of a recognition of certain physical facts—which, if the craft is to maintain its touch with the real world, must always be considered first and foremost. But beyond all the questions of relative width, color, suitability for certain processes, and optical legibility, lies the whole fascinating field in which the skilled typographer is at home. We must perforce leave him at this point. Looking at a number of books, he will improvise his own dogma as to the very delicate matter of suitability—a matter in which practically every canon of good taste and every detail of a cultural background and literary training are involved. We can offer only two generalizations to accompany him on his journey.
The first is that before any question of physical or literary suitability, must come the question of whether the face itself is tolerable or intolerable as a version of the roman alphabet. If a single letter is warped, emphasized above its fellows, made grotesque (as in this ugly g) or snub (as in any non-kerning f); if the letters, however pretty in themselves, do not combine automatically into words; if the fourth consecutive page begins to dazzle and irk the eye, and in general if the pages cannot be read with subconscious but very genuine pleasure, that type is intolerable and that is all there is about it. It must be wiped out of the discussion. There are bad types and good types, and the whole science and art of typography begins after the first category has been set aside.
The second generalization is, briefly, that the thing is worth doing. It does genuinely matter that a designer should take trouble and take delight in his choice of typefaces. The trouble and delight are taken not merely "for art's sake" but for the sake of something so subtly and intimately connected with all that is human that it can be described by no other phrase than "the humanities." If "the tone of voice" of a typeface does not count, then nothing counts that distinguishes man from the other animals. The twinkle that softens a rebuke; the scorn that can lurk under civility; the martyr's super-logic and the child's intuition; the fact that a fragment of moss can pull back into the memory a whole forest; these are proofs that there is reality in the imponderable, and that not only notation but connotations is part of the proper study of mankind. The best part of typographic wisdom lies in this study of connotation, the suitability of form to content. People who love ideas must have a love of words, and that means, given a chance, they will take a vivid interest in the clothes which words wear. The more they like to think, the more they will be shocked by any discrepancy between a lucid idea and a murky typesetting. They will become ritualists and dialecticians. They will use such technically indefensible words as "romantic," "chill," "jaunty" to describe different typefaces. If they are wise, they will always admit that they are dealing with processes of the subconscious mind, mere deft servants of the goddess Literature. But just as the poet prefers that the wireless announcer at the reciting of his verse over the wireless should choose neither a harsh nor a maudlin tone, but a sympathetic one, so will any author cock an anxious ear before the printing type that carries his words, and ask in his pride neither for officious flattery nor harsh mistreatment, but for justice tempered with mercy.

Taken from the book The Crystal Goblet: Sixteen Essays on Typography, edited by Henry Jacob and published by Cleveland and New York: The World Publishing Company, 1956