Introduction

History

Accents have enhanced the basic 26 letters of Latin alphabet since the Middle Ages. They delimit the way a word is pronounced and they help distinguish between words which would otherwise look the same. This is necessary, for all European languages have more phonemes than the 26 basic letters. Another option of increasing the number of characters is more old-fashioned, but still in use: it is the diphthongs or two following letters, which are read as one phoneme (for example ch in Czech). In some languages one phoneme may be represented by more than two characters (for example by three, in German: sch). In certain languages there exists more than one character for the same phoneme: in Czech these are i-y or ú-ů. There is also the possibility of creating a completely new character, such as Old English þ and ð adapted by Icelandic. However, the easiest and also the most practical away of extending the Latin alphabet are the accents. That is why they are the most commonly used. Although the basic shapes of the accents have been standardised during the 20th century, the accents do develop — permanently and almost invisibly. For example, in Czech, there is a tendency to abandon a decorative style in favour of something neutral and less eye-catching: the previously common caron with sharp contrast is being abandoned nowadays, in favour of the version symmetrical along the vertical axis. In some languages, the readers are very sensitive to the proper shapes of the accents; in others, the shapes are more varied. In Czech and Romanian, it is not acceptable to interchange a caron with a breve; in Turkish, the two shapes are freely interchanged. Accented characters are often perceived as individual letters. This is reflected in the alphabet order — sometimes the accents are ignored, but in other cases, accented characters are placed following to the non-accented ones, or they are placed at the very end of the alphabet.

Digitalisation

When digitalising an existing face, the typographer has to make an important decision: whether it is better to use the original shape of the accents, although it might seem out of date, or whether to design new accents. Both ways are possible. For the digitalised version of Preissig’s antiqua, František Štorm has prepared two versions: one with the original accents (very unusual from the contemporary point of view), and one with modern ones. Well known is the example of Futura, whose accents were designed by Paul Renner. Some foundries have adopted them, some have replaced them by more conventional ones. I believe that with the most common faces, the reader gets accustomed to a certain shape of the accents. If these are replaced by new, unconventional ones, it might cause difficulties with the fluent reading of the text, even if the new version is typographically better.

One of the most common errors is using accents which do not harmonise with the character of the face and which are constructed differently than the face, such as the accents of a sanserif face used with an antiqua, or, even worse, with decorative italics. Accents must be constructed in harmony with the way the face was built. The designer’s indifference to the shape of the accents is typical for a number of type foundries. Appalling ignorance is exhibited for example by the North American Emigre foundry. In such faces one of the common errors is interchanging a caron with a breve or including a useless ogonek. Sad example of faces with incorrect accents is unfortunately provided by the ones most commonly used: the Times New Roman and the Times typefaces supplied with MS Windows and Mac OS operating systems. I fear that here one might wait for the rectification in vain…

The distance between the glyph and the accent constitutes another problem for the typographer. The reader perceives the accented character as a whole and a wrongly placed accent may lead to a wrong interpretation. If the accent is placed too close to the character, it could merge with it, especially in small lettersizes or during cursory reading. This makes reading more difficult and it can even lead to mistaking a character for another one. An accent placed too far above the character makes smooth reading nearly impossible. The same problem is encountered in cases where the diacritics is placed on the side of a character (ď); here one also has the added problem of proper kerning. The distance between the baseline and an accent should be, within one face, constant. It is necessary to take into account that some glyphs (such as the ones with round strokes, like c or s) might be taller or shorter than others. In exceptional circumstances, it is unavoidable to alter the shape of the character, if an ornament gets in the way of an accent.

Most Common Errors

Even if the face contains correctly shaped accents placed in an ideal distance of the characters, all effort might be in vain if the accents are not well placed horizontally. Such accents “fall off” the characters and they destroy the overall harmony. Symmetrical accents are, almost as a rule, placed at the optical center of the character, especially if the characters themselves are also symmetrical. Asymmetrical accents offer more options where to place them. Again, it is necessary to opt for the visual balance, so that the accents do not “fall off” the characters. Maximum care has to be taken while adding accents to the italics, especially to a strongly calligraphic one; sometimes, one has to give in to a compromise. Some foundries fault by assuming that for one typographic family, one set of accents is sufficient. Correct diacritics have to be designed for each weight separately, or at least it has to be correctly interpolated. Because the accents must respect the character of the face, it should be pointed out that italics, especially antiqua italics, has strong calligraphic leaning and the accents should therefore be calligraphic, too. With other faces (such as sans serifs) it is usually sufficient to slope the accents.

The Obstacles of Diacritics

A different approach to accents is applicable when designing a poster, a book cover or a logotype. Then it is possible to adapt the original accents, for example to fit the text into the available space. Some artists work creatively with the accents — they use them to enhance the aesthetic impact of the work. In such cases, the caron might be substituted with a horizontal stroke, a breve or a triangle. Dieresis might form a part of the character or it might be stretched to its full width. An accent might also grow out of the character. This option is not advisable in cases where such revolutionary approach may lead to misinterpretation. One example is the face used for the names of Prague Metro stations, in which a caron is replaced by a macron. (Recently issued digital version of the face does not contain that error.)

There is no single answer to whether accents may vary in width. One of the typical examples is the macron above i and above the æ ligature. If the width is set to be universal, then it comes out as too wide above the i and therefore requiring adapting of the kerning in words such as līl; on the contrary, it is too narrow for the æ ligature. In my opinion, it is better to use a variable shape of the accent, both in the case of the macron and with the ogoneks. These need to be adapted so as to harmonise with the character. In times of the printing press, the accents above the uppercase letters constituted a number of problems. That is why the diacritics above the uppercase letters usually comes out flatter than the diacritics above lowercase; this also allows the uppercase lines to be set closer together. It is not clear whether it is better to use similar accents for lowercase and for small caps (this is desirable where the small caps are used on one line with the lowercase), or whether it is preferable to use smaller accents for both small caps and uppercase. Ideal solution, of course, would be having both options available… Diacritics are a hot topic, both among type designers and the general public. Millions of readers in Europe struggle with the incorrect accents in foreign fonts, and converting between the different encodings is a nightmarish experience for virtually every computer user in Central and Eastern Europe. Apart from the extreme opinion that it would be easiest to get rid of all accents whatsoever, there are some interesting ideas resulting from the discussions. As a proof one may look at the Slovak project Diacritics. The students of the Graphics design Department of the VŠVU (College of Fine Arts in Bratislava) have attempted to design entirely new symbols; some have tried to simplify the accents, some have parodied them. The brochure is undoubtedly inspiring. However, globalisation and standardisation make the possibility of a complex change yet more difficult. Introducing new characters would prove far more complicated than the recent adding of the Euro symbol, and still, there are so many typefaces which have not had it added yet. The following overview does not attempt to be comprehensive. The list of languages does not contain some minor and rarely used languages, as well as any non-European languages. It is also necessary to consider the fact that some languages are influenced by the neighbouring ones. Thus, they adopt words with their original orthography, for which the original accents are needed. This is common in English; some words adopted from French, such as café or façade, are written as in original. Also, the overview does not contain some special symbols needed for transcription from foreign alphabets into the Latin one, or the symbols used nationally for denoting accent in textbooks and dictionaries.

Filip Blažek

First published in Typo Magazine 10 / 2004

Home
Introduction
Further Reading
Languages
Galleries

Latin Diacritical Marks

Greek Diacritical Marks

Sanskrit Diacritical Marks

Kerning

Help
About the Project