LettError

Widely considered as craftsmen rather than artists, typeface designers traditionally inhabited the most cloistered of environments. Since the mid-80s, however, they have had to come to terms with the outside world as new technology has equipped a generation of relatively unskilled type users, facilitating a proliferation of new font designs. While many designers and typographers are still reeling from the shock of the new, Erik van Blokland and Just van Rossum, a pair of young designers collectively known as LettError, have set out to take technology to its limits. Rather than simply adapting to change, they are waiting impatiently for computers to catch up with their ideas.

Beowolf, the first LettError typeface to be commercially released, is still their best known. Originally called Random Font, its starting point was the designers' understanding that PostScript fonts are sets of mathematical instructions, rather than physical forms. When letters are stored as coded outline information, they need not necessarily take the same form each time they are printed: if a random element is introduced, the same set of instructions can produce a variety of different letterforms. Beowolf is available in different degrees of randomness: Beowolf 23, for instance, is a great deal more irregular and jagged than Beowolf 21. While no two instances of a character ever come out the same, the letters of the Beowolf typeface are instantly recognisable as part of the same family.

While Beowolf and their subsequent random fonts have broken with current typographic convention, LettError view the standardisation of letterforms that resulted from mechanical typesetting not as typographic perfection, but merely as a phase in a much longer history of written communication. Erik van Blokland explains: 'For a short while, maybe 300 years, there was a system that meant letters had to be the same. A mechanical system of producing type meant that there was one master form and you made copies of that; it was all very logical. That is why all the 'A's are the same and all of the 'B's are the same. We have grown up expecting that to happen, but it is the result of a mechanical process, not for any reason of understanding or legibility.'

Van Blokland recalls that Beowolf found its name almost by chance. Suggested by a friend, the title seemed suitable because of its gothic feel, but it also referred to more than simply the style of the face. The manuscript held in the British Museum of the Anglo-Saxon poem Beowolf was probably made by scribes in the year 1000. It is thought the poem reached its first literary form late in the eighth century, but existed in any number of oral forms for centuries before that. The surviving version of the text, which is considered as the origin of English literature, was simply one of many. Just as the poem Beowolf resists traditional modes of literary criticism because it cannot be treated as a unique, fixed document, so the typeface escapes conventional typographic judgements. Discussions about the finer points of spacing and ligature are irrelevant when each letter is of uncertain form. Broader questions of legibility do still apply; Beowolf remains a recognisable alphabet when printed as part of a continuous text.

In this respect, LettError's enthusiasm for the new is tempered by an understanding of tradition. Both studied at the Royal Academy of Fine Arts in The Hague under Gerard Noordzij, whose teaching is based on the idea that handwriting is the origin of all typographic communication. LettError see their designs as continuing the traditions of written communication that pre-date printing, recognising that established type designs have taken their forms from handwritten script. LettError's 'handwritten' fonts display these ideas most clearly. Van Blokland and van Rossum each wrote out the alphabet then scanned and digitised the letters on a computer. Marketed as Justlefthand and Erikrighthand, these fonts became the first of many which allowed people to indulge in the irony of typing a letter in a handwritten script. Since then, having your handwriting digitised has become a service widely offered to computer users. While the value of this exercise might be debated - just what kind of layer of meaning are you adding to your typed communications by using this kind of font? - LettError's pioneering work in this area has opened up whole new areas in type design.

Subsequently LettError have pursued explorations along these lines with a range called Instant Types: rather than returning to the handwritten form, they have created fonts by scanning and digitising a variety of existing letterforms. Trixie, the original Instant Type, was taken directly from a typewriter that belonged to a woman whose name was... Trixie. Trixie and other fonts based on found letterforms, such as Confidential and Dynamoe, have become widely used, almost invariably as a very literal way to convey urgency or spontaneity. While accepting that some fonts tend to dictate use, van Blokland and van Rossum do not subscribe to the strict divisions that have been made between text and display faces. They argue that every typeface has both an image and a text factor, and could be used for either purpose, illustrating this argument with a graph. Image factor, the extent to which the viewer is drawn to the type, is plotted against text factor, the font's legibility. While this has the appearance of a scientific process, van Blokland admits that the way they plot a typeface is actually subjective and arbitrary. LettError are not seeking absolutes through their type designs.

Although they subscribe to the current typographic orthodoxy - the easiest typeface to read is the one you read most often, the catchphrase of the West Coast type magazine Emigre - van Blokland shows a healthy respect for typographic nicety. Convention or not, he feels Trixie makes a poor text face because it 'is dirty, has some disadvantages in the spacing and some of the shapes are a bit unclear.' While acknowledging that technology has made experimentation with typographic abstraction viable, he feels that this kind of experimentation is not part of LettError's project. 'If you are making typefaces now, then you are making them for people who need type to communicate ideas to other people.'

Van Blokland speaks approvingly of the new GX font technology. GX makes fonts about as smart as it is currently possible for them to be, taking care of the finer details such as ligatures and hyphenation. Rather than being wary that such developments might lead to an erosion of craft skills, van Blokland feels that computers are essential repositories of this kind of typographic knowledge and are all that will prevent it from being completely forgotten. In early interviews, LettError described their own plans to make intelligent fonts - letters that would respond to the weather or fall apart after a use-by date. After years of experimentation, they have discovered that - at the moment - it is technically very difficult to make a typeface that knows anything about the way it is being used; how big it is or where it is on the page, for example. Van Blokland harbours none of the current reservations about smart technology, brushing off fears that technology will take all control out of the hands of the individual. He also resists the blandly prophetic statements issued by many in his field.

Like many of their more forward-looking colleagues, LettError have designed fonts for the type magazine Fuse, which is supplied on five posters and a disk. Some involved with the magazine claim to be engaged in a project of revolutionising visual language, although what is meant by this is often not entirely clear. Jon Wozencroft, the editor of Fuse, has drawn an analogy between the recent developments in type design - and typesetting technology - and the invention of photography. According to Wozencroft, just as photography made the reproduction of reality easy and obliged painters to do something more interesting instead, desktop publishing has enabled even mere clerks to set a competent piece of type, allowing the professionals the freedom to roam new creative territory. This comparison is ripe for criticism - it adopts a discredited view of the relationship between painting and photography; it assumes that the task of the typesetter before the advent of desktop publishing was absurdly limited; it requires a belief that many issues in type design and typography have been conclusively resolved, and so on. Most irritatingly, this argument has been used as justification by some designers to adopt the title of artist, as if by doing so they were scaling the dizzy heights of a supposed cultural hierarchy.

The fact that many of LettError's inventions are not immediately marketable, are motivated by curiosity rather than commerce and would not appear to have a lot to do with the majority of day-to-day printed communication, might prompt some to suggest that they have similar aspirations. But LettError make no such grand claims for their work. As usual, they take a disarmingly common-sense approach, suggesting that artists and designers have very different motives. They do not predict any revolutions, arguing that the development of type is closely linked to the slow development of language. At a recent conference organised by Fuse, LettError went against the grain by boycotting the terms 'multimedia', 'cyber' and 'Internet', but still applaud the magazine for providing a forum for the systematic exploration of type.

LettError's technological skill and enthusiasm mean that their inventions are often several leaps away from the marketplace. Their new BitPull fonts are bitmapped typefaces which allow control of the individual pixels - BitPull does to typefaces what 'pitbulls do with little children'. Although the BitPull idea is three years old, the fonts are not yet on sale. Called a 'typetoy' by LettError, it is more of a system than a typeface. The faces work with an application written by Just van Rossum, but it is in a far from distributable form: it has no interface and you can only click on certain buttons otherwise it crashes. The latest product to see release is Kosmik, which includes a random element but abides by certain rules: each letter is selected from three alternative forms, ensuring that no three characters in a row are ever exactly alike. LettError call Kosmik a 'Flipperfont,' and believe such of fonts are especially appropriate for use on screen. By using a different letterform on each consecutive frame, the letters can be made to jiggle - instantly reminding those of us brought up on British TV of the wavering forms of Rhubarb and Custard.

LettError's experimentation with random fonts was prompted by a desire to take full advantage of PostScript technology. Historically, designers aiming to exploit technological innovation have done so under the banner of function: when typographers earlier this century proposed abandoning the uppercase, they argued it would lead to more effective use of contemporary typesetting machinery. LettError would not claim that their typefaces are efficient in that sense - an article set in Beowolf for Emigre magazine took an hour and a half to print. LettError's objective is 'to use the technology for what it is worth'. They point out that a laser printer is a very large and expensive machine capable of responding to complex sets of instructions, so why, when you ask for an 'A', should it print the same form each time? You might have to wait three quarters of an hour for your page of Beowolf to emerge, but at least the printer is doing something instead of just sitting around.