The origin of “adhesion”

In 2001 I came up with this word for a workshop I was doing. I wanted a word that I could get people to think about designing a typeface very quickly. This is in the context of people who don’t have much experience in designing typefaces, and can’t do what Gerard [Unger] suggested, use pre-existing work as a starting point. [They] start with a clean sheet of paper. It also needed to help people get over the problem of the absence of a brief, which we find is one of the most difficult problems with new students. Even if they are professionals with ten or fifteen years of experience, they’re conditioned to respond to a brief: read a brief, restate a brief, find out the limitations inherent in a brief. But if you start with a completely blank sheet of paper you have to imagine the thing yourself, and that’s a very different kind of challenge. So, what we want to do is help this period of experimentation and exploration to happen fairly quickly.

I chose the word “adhesion” because at the time people were having discussions [about] the h and the o – which I don’t think are very helpful – or “hamburgefonstiv”, “hamburgefons” or whatever. People forget that “hamburgefons” and its variants were not a design tool, they were a testing tool. There’s photographs of people at Linotype etc. looking at these strings of characters to space and to see how the letters would fit together, but once they had already been designed, under a process (which Gerard mentioned, and our archives here show, and indeed Walter Tracy’s Times Europa s up there shows) – [which] was extremely protracted in the early stages: there was a lot of effort to get the basic shapes right. There was a lot of know-how being passed down from the company – because there was an explicit company doing that stuff – and this is completely absent now, so we need to accelerate this process.

So the point with “adhesion” was to identify a set of shapes that allow people to – very quickly – get a feel for the style of the typeface; and also the differentiating elements in the typeface, but without having the risk of every small change needing to propagate through the rest of the typeface and taking forever, and leading them down blind alleys.

So, there’s an o and an n. The o is there only to make a word – I don’t think the o needs to be there, the o is an aberration in that it is the only completely symmetrical letter; as a round counter it is not very helpful. In terms of round counters of letters, the d or b are be much more helpful, because they help you decide how does a round counter stick onto a vertical stroke, which is an integral part of the Latin typographic script, and is not at all answered by the o – or indeed the n to a large degree. So, one of the b, d, p, q letters needs to be there.

The h and the n are very helpful to have because you can begin to build in people the skill of how do bits that stick out influence the perception of shapes. So if you design the h the same as the n, students will fairly quickly through just these two shapes realise that the fact that the stem of the h ascends, will make the curve of the h look different in relation to the n’s. We can begin to get some idea of how things interact.

The most important letter for identity is the a. Because the h and the n, and the o, and to a large degree the d, are useful to give the underlying pattern and uniformity in a typeface – but the distinguishing features will come much more from letters like the a which has the key decision that we make between the balance of the top and the bottom halves within the x-height. How dominant is the top in relation to the bottom, or vice versa, which we can see propagating though to the e and the s, and so on. But also the treatment of the open stroke: is it something that is heavy at its tip, is it something that is light? Is it something that curves in quite a lot, or leaves a big gap between itself, its tip, and the closed bowl? And also the treatment of the underside strokes again in a, d, and we can look at how these things propagate. The e and the a build this set of relationships, [they are] the two main letters that interrupt the zone of the x-height, which is a key design feature that can very quickly give a lighter or heavier feel to the typeface.

And the s is the really tricky letter in the lot, it’s the one really difficult letter. Because it has the problem of making a concave and convex curve look part of a single stroke. It also helps people learn quite a lot about conventional structures, where thicks and thins might start, and also gives a very quick idea of how fast, or how slow the typeface might be on the page. A wider s will make the typeface look much more slow because there’ll be a stronger horizontal emphasis in the centre; a narrower s will have a diagonal stroke which will make the typeface look “faster”.

There’s no descenders, because the depth of descenders can change quite a lot in a typeface, and depending on the style, and indeed the brief, the descenders might have different characteristics. We have very good examples, like Lexicon, where typefaces have different ascender and descender lengths with no detriment to their quality.

But this set of letters allows people to very quickly try out their ideas without the problems of all the diagonals (v, w, x, y) which are a set of problems in themselves; without letters that are traps, like the g, which are extremely individual – but exactly because of the individuality you need to build them into the context of the rest of the typeface, so that they both support and emphasise its individuality. [“adhesion”] allows [students] a good enough combination of vowels and consonants so that they can get decent [texts]. I’ll plug Miguel Sousa’s adhesiontext website because it came out of this problem: from having a small number of characters, how do you get valid test text strings. He built a website that you can enter any set of characters and will return a string of words sources from online documents like dictionaries, using the characters that you have selected. (It’s now a standard design tool for all of our students.)

But it means that very quickly this process of experimentation that Gerard alluded to can happen for people who are not used to building consistency and variance in typeface design. And very quickly we can begin to look at things like that. [Showing early test document by a previous student.]