Jan Middendorp
[Originally published in the LettError Book. Charles Nypels Award in 2000]

toolspace

In the mid-Eighties there was hardly a computer to be seen at the Royal Academy. Working by hand was the norm. That was pleasant because you learnt to deal with letters and words in a very direct and physical way. Typography under Gerrit Noordzij was in the first place writing and seeing how forms arise on paper. There were computers, but their place was at home. JvR and EvB, later LettError, were already messing around with computers at secondary school, just as a decade earlier whizz kids with a technical bent soldered their own amplifiers and radios. When typography entered their lives and they got to know one another, the question was not whether a computer might be a legitimate tool for that profession, but what the best way to use it would be. Why? came later, and it was asked by others. After all, typography is a discipline full of traditions and conventions, and few typographers are in the front ranks when it comes to welcoming new technology. From the moment that LettError went public with ideas that were mainly intended to be playful but were often interpreted by colleagues as a provocation, there was a call for justification: a method behind the madness.

There is no longer any need to explain why computers can be useful for typography; hardly anyone does it without them. But the next step, that the typographers themselves influence the operation of the computer, is too much for most of them. Still, it is a natural choice for designers who do not want to be slaves to their digital tools.

A hammer can do what a hand can’t

In itself a computer is not a tool but a tool-kit, a box you put functions in as you choose. All kinds of instruments that used to litter up workshops and offices now have a place on the hard disk: marker and drawing board, scissors, paste, trash can, ledger, typewriter, calculator, fax and telephone. Some of those instruments are already in the computer when you buy it in the form of system or standard software; others are added by buying extra programs.

Can you compare a program with a tool? Not completely. In the case of a physical tool, like a paint-brush or a hammer, it is pretty clear what kind of tasks it is more or less suitable for. Like computers, those objects (to borrow a term from Marshall McLuhan) are an extension or expansion of the human body, but their scope is usually limited to a single function. For example, they are highly suitable for spreading chemicals on a surface or hammering a piece of metal without hurting your hands. The case is more complicated when it comes to a computer program. A program has more potential than a hammer, it is more complex to use, and it comes with hefty instruction manuals to guide the user through all the functions. After weeks or months of intensive use, a program can still surprise you with undiscovered functions. The learning period is a long voyage of discovery through what you can call the software world of tools.

But no matter how overwhelming a program may be when you are first introduced to it, its capacity is finite. Software is a commercial mass product, so it is not made for individuals but for a target group with hypothetical wishes and demands. If in practice that target group turns out to have unfulfilled desires extra functions are added with an upgrade, but only if there is enough interest. You can do everything with a program as long as there are enough people who want to do the same thing. None of this raises any problems for the average user, who usually does not demand more of the computer than it is capable of doing, and in fact finds reassurance in its limited capacity. That should not be the case for users in a creative profession. It is precisely the task of designers (graphic, industrial or whatever) to discover new possibilities. In their case the use of a computer is perhaps more of a handicap than an advantage.

If you walk through the real world with a pencil behind your ear, you will find the inspiration for new ideas in that world. If you sit in front of the monitor all the time, you will draw your ideas more and more from the computer. In the past you were successful if you could make what was in your mind’s eye with your own hands. Now there is a danger that designers will only function properly if their creativity does not exceed the potential of their computer: if the world of ideas fits inside the world of tools. Of course, in the long run this leads to a monotonous computer-driven uniformity.

Once an alert designer has become familiar with the software, it is to be hoped that questions will arise which the software is incapable of solving. This can be frustrating. You think of an image or a solution that requires a specific combination of functions, and then it turns out not to exist. Or you want to repeat an action a large number of times, while the program does not offer any way of doing it automatically. The toolhorizon comes into view. Should you begin to have doubts about yourself as a designer? On the contrary. It simply means that the people who devised the program did not take your idea into account, so it is a relatively new idea. And it is no bad thing for a designer to have new ideas. All the same, good advice is a rare commodity when you run up against the limits of the tool-kit in the middle of the thinking process. Should designers slow down and adjust their ideas to what the computer can handle? As we know, to design is to make images within given limitations. But not all limitations are the same. Limitations and demands imposed by a client are easier to accept than the arbitrary limitations of your digital tools.

Honing

Artisans of the past — the predecessors of our designers’ guild — were rarely satisfied with tools as they found them in the shop. They always had the tendency to personalize their tools, to appropriate them by honing them, converting them or expanding them. The more specialized the work, the greater the demand for customized or individually made instruments. For instance, letter-cutters in the past thought up methods of working faster and more meticulously, and to that end they designed not only new fonts, but also counterpunches and other new tools. It must be possible to do something similar with software. After all, programming graphic programs is much too important to leave to programmers.

For most designers it is not an obvious option to start meddling with the original code themselves. Programming is regarded as the work of specialists. So e-mails are sent to the publisher with lists of demands. With a bit of luck the engineers reply (‘it is a marketing question’), and perhaps the marketing department too (‘it is a technical problem’), and if you are really lucky a new version which contains a few improvements appears on the market between three and eighteen months later. A designer does not usually have that much time. The deadline is generally next week, or yesterday. Right from the start, LettError found it logical and fun to solve design problems with new bits of software. This was connected with their early, extraordinary interest in the inside of a computer. Now even designers without a whizz kid past can do their own programming, provided they are given the right tools to do the job. Take Python, for instance, a programming language that was developed by Guido van Rossum (Just’s brother). Python is internationally recognized as a flexible, low threshold language for addressing the computer, and it is exceptionally suitable for making additions that you have written yourself to programs without requiring a lot of prior knowledge.

Parametric design

What are the practical possibilities for a designer who hones his own tools? An interesting facet is the opportunity of automating a part of the creative process. A design usually starts with a series of sketches, a vague visualization of possibilities that gets the selection process — the choice between good and mediocre solutions — going. As long as you’re sketching on a napkin, the marks can mean whatever you like. If you start straight away with the mouse in your hand, which is more common nowadays, you are immediately confronted with tenths of millimeters and dead straight lines. The design already takes on finality when it is time for flexibility.

There is an alternative: programming-assisted design. In this case the alternatives are neither sketched by hand nor directly drawn on the computer. The designer indicates the ingredients and the margins, and asks the computer to think up a series of variables, taking certain rules into account. In other words, to vary certain parameters at random. The computer itself does not start designing, it cannot be accused of interfering with the creative process. It does what a computer is best at doing: very quickly working out the consequences of possible decisions. Fully informed, the designer selects the most attractive, the handiest or the best solution from the range of alternatives. Nothin yet? press the button again and see what the program makes with slightly altered parameters.

The designers themselves determine how complex they want this process to be, how long they want to postpone the definitive decisions. They can decide to add certain values or to fix certain variables (a color, a font, the thickness of a line). The program becomes a specialized graphic synthesizer, that like its musical counterpart, is just as limited or versatile as the person operating it.

Best and better

The random principle is good for something else as well. In the early days of digital typography, software engineers were fascinated by the possibility of creating the perfect form. Had people not been searching for centuries for the roundest circle, the most flexible curve, the clearest line? With the computer this was as easy as falling off a log. Add razor-sharp definition by using offset on smooth modern paper, and the result is the best print ever. Engineers get the warm fuzzies when confronted with such perfection. The question is whether the ordinary reader wants such an overdose of perfection. After all, a completely balanced page of text can also be completely boring. Perhaps our attention wanders if there are no rough edges. Is the best there is really better for us?

Here too parametric design can offer a solution. By allowing the computer to mess around within the — strictly defined — margins of a virtual font or layout, a new sort of liveliness can arise. In that case the computer is not used as the big leveler that it only too often is, but as a way of introducing carefully considered imperfection.

The critical outsider will note that this method also has its disadvantages. After all, sometimes designing proceeds faster and more securely if nothing is left to chance, if work starts straight away as on the computer with a precision of a hundredth of a millimeter (‘exactly one cm’ is also possible). Is it really handy to generate the layout of a calendar with a program that can shift parameters endlessly? You have to write a program like that first, and that takes a lot of time. Of course not, will be the answer, the first time naturally takes more time and trouble, but that is what makes it so much fun. Design is hardly a challenge any more, but programming is. The paradox of designing like Just and Erik is that a lot of individually written tools are only efficient (in the sense of saving time) if the same sort of design is repeated a large number of times; but that is a very rare occurrence. Explorers, and that is what they are, do not want to do the same thing twice. They prefer to leave that up to ordinary designers, and studios. Which brings us to the second paradox: such designers may never get around to programming. They hope that the scripts and programs of LettError will simply be available on the internet one day. Ready to use.