Kanat: The beginning (2014/15)
In October 2014, I was quietly working at my desk – at a FabLab called ZB45 – when Wiek and Corine of the [Typo]Grafische Werkplaats Amsterdam walked in with an idea. They wanted to task the design of letterpress typefaces in “modern” 3D outputting techniques. So the owners of the maker space said: How convenient, we have a type designer right over there!
Of course, I was only interested in designing Arabic fonts and didn’t think an Arabic letterpress typeface would be widely used in Amsterdam. But Wiek and Corine believed there would be an audience for it. So I trusted their judgment and began to sketch.
Before this project, I had a deep love for printmaking as well as prior training in Arabic letterpress technique – thanks to a lovely Lebanese printer called Fadel. He was the owner and printer of the now inexistent Ras Beirut Tipo Press. Without these experiences, I would have been quite overwhelmed starting this project.
I was motivated; at that moment, my work was centered around defying the idea of treating the Arabic script like Latin (which, to me, meant designing Arabic characters in separate “boxes”). So I thought it would be fun to challenge my ways and ask: How would I create an Arabic typeface for moveable type technology, to do the opposite of what I envision for Arabic, and design letters in boxes?
Now where does one begin to address this type of project: with the technique or concept?
A few aspects were clear from the start: In its shaping, I decided to create a detached Arabic typeface to “naturally” fit the vertical separation of characters in letterpress technology. And I imagined the letters would be pretty bold, with short ascenders and descenders, to generate a compact effect. In technique, I would need to explore the different 3D technologies extensively since I had little knowledge of these methods. Experimentation in technique, therefore, became my starting point.
Technique: wood, rubber, or plastic?
My initial tests explored texture types and line thicknesses. How thick can a line be? How thin can a negative space be? Do they break or disappear in specific 3D techniques? How curved can a curve be? Does anything happen to the curve architecture between the digital and the physical output?
I was searching for the graphic possibilities and limitations of the three techniques I had at hand in the FabLab: 3D printing, laser-cutting, and CNC milling. You can imagine that the letters must be quite strong to withstand frequent inking, cleaning, and mostly the weight of a giant metal drum rolling over them for years and even decades or centuries at the printing press. So I designed pilot shapes containing different stroke thicknesses, negative space thicknesses, and curve speeds to test which parts would remain intact and which would fall apart because of the material or the heavy use of a press.
First tests (top two photos by Claudia Willmitzer)
Form: detached yet connected to history (2016)
In search of a graphic direction, I searched for wood typefaces in the history of Arabic script. But I couldn’t find much except for the following two examples: One was a missing block for the Roman Scherer A.G. Holzletternmanufaktur Arabic letters produced by Dafi Kühne. Unfortunately, I could not reach Dafi in time to ask about its story. And the second example was sent to me by Onur Yazicigil. It was a title block in Thuluth script from the Bulaq Press in Cairo. And that was it! I did keep in mind how the dots of the letters were built as changeable inserts in the Holzletternmanufaktur Arabic.
Left: Dafi's Holzletternmanufaktur Arabic. Right: Onur's Bulaq Press woodtype
In terms of form, there have been several past attempts at making detached Arabic typefaces. In Lebanese script history, architect and designer Nasri Khattar is the most notable one. You can read more about his experiments in Yara Khoury's book: Nasri Khattar, a Modernist Typotech, published by Khatt Books, 2014.
Sketches of Unified Arabic typeface by Nasri Khattar, AUB Library Special Collections, personal photo
In my design process, I decided to choose a script source to base my decisions on. It seemed logical to begin with Primary Kufic since it’s one of the more geometric script styles and a style that I simply love. It also has an intrinsic boldness that I imagined for this letterpress typeface (at large type size) and is challenging because it is already pretty abstract. My design problem was, therefore: How to abstract a script that’s already innately highly geometric? How to make it detached? And how to make it more readable for modern readers (because it contains letter shapes that could now be ambiguous and misread by people)?
Primary Kufic, Single folio, 9th century ad, 20.8 × 27.8 cm. In: François Déroche. The Abbasid Tradition: Qur'ans of the 8th to the 10th centuries, The Nasser D. Khalili Collection of Islamic Art, vol.I. London: Nour Foundation, 1992.
Development of the Kanat letterpress typeface (2016/17)
I followed different models of the Primary Kufic (instead of relying on only one). I also consciously broke many of the rules of letter connectivity in the Kanat to reduce the number of wood blocks in the typeface as much as possible: Given the scarcity of resources in the world, it was important to me to limit the number of plastics or wood used for the production of the typeface.
In terms of characteristics: The letters are bolder than in the Primary Kufic. The general contrast follows the contrast placement of newer Arabic script styles, meaning the vertical strokes are thinner than the horizontal ones. (It’s usually (but not always) the other way around in the Primary Kufic.) The counter shapes are also open on one or two sides for a higher geometric impact in the design.
Characteristics as originally sketched
It was time to return to the production process, with the design system now being mainly established. I’ve explored many ways of building the letters: Should they be modules with loose dots or full dotted characters (image of modules and flexible dots)? I’ve reconsidered the detachment level of the typeface: Would the typeface be fully attached, semi-detached, or fully detached (image of the sketch)? What would the typeface’s final size be? And how thin could the counters go?
Sketch of the different stencil levels
Evolution of letter Waw from the first sketch to the final digital drawing (in red)
To answer these questions, I tried one sentence in different outputting processes. It was time to decide which technique to adopt finally. I had narrowed my options to 3D printing using flexible plastic and CNC-milling (in wood). And I compared the results with wood laser-cutting because the laser cutter rendered the most accurate shapes.
Test sentence in CNC, 3D printing and Laser-cutting and the prints
Even though 3D printing (using one of the earlier Ultimaker machines) gave an attractive, industrial texture, I didn’t want to use so much plastic. Also, each letter took 9 hours to print, and I would still need to mount the 3D print on a base to reach Dutch type height. There was also the worry of letters becoming more and more fragile over time and use.
Together with Rudi Oostermeijer of the GWA, we observed that the wood yielded a beautiful, warm texture. And the production would be more practical because letters could be carved and cut into one piece. Of course, there are higher-level 3D printing machines where this would also be possible, but I didn’t have the budget for this kind of production.
Challenging characters evolution
Characteristics of the Kanat in 2017 before becoming the wood typeface.
So the Kanat wood typeface came to life at a type size of 189pt, in Dutch type height, and produced in pear wood using CNC-milling (with two different blades, thanks to the craftsmanship and dedication of Rutger Paets). One of the last production questions was: How many sorts do we need for each character? Since the local migrant population in Amsterdam predominantly writes in Arabic, I decided to cover a glyph set suitable for the Arabic language. So I counted the frequency of letters in different texts and made an average based on the type size and maximum printing width for the presses at the GWA.
We ended up with a total of 167 blocks in February/March 2017.
Kanat wood type
Images of the milling, and first use of the Kanat Wood (last two photos by Claudia Willmitzer)
I’m still slowly looking for another craftsperson to produce the punctuation, Persian, and Kurdish characters and to repeat some characters that changed in the improved digital partner of the Kanat Wood typeface. And I will probably venture into another design for an Arabic letterpress typeface sometime in the future.
This project was made possible with the following people, in chronological order: Jaap & Liesbeth of Stichting MOVING CONCEPTS, Corine Elemans and the GWA team, Kristien van den Oever and AGA Lab, Rutger Paets for his generosity and CNC-milling the letters.