Climbing Everest

“Designing Zapfino Arabic takes everything I ever learned about type design and then some… Have never, ever, worked on anything this challenging.” — Facebook status: Oct. 24, 2013

Designing an Arabic companion to the Zapfino typeface was a running joke with my Monotype colleagues, one that I would react to half in amusement and half in terror. It was exactly a decade ago that I was introduced to Professor Hermann Zapf and we would start work on the Palatino Arabic design. Many years later, and after finishing Palatino Sans Arabic, the idea of Zapfino Arabic started floating around but I did not yet have the right skill-set to take it on. I am no calligrapher.

The turning point came with the Afandem typeface I designed for my PhD studies where one of its variants is based on manuscript Naskh. It was my first attempt at a calligraphic design, and I enjoyed the design so much that I wanted to explore this elegance of movement once more. Once the PhD work was more or less done, I finally cooked up enough courage to take on Zapfino Arabic. I knew from the start that it would be difficult and that the design process would be an interesting one to share – something I did regularly on Facebook and Twitter.

A question of slanted writing

My first challenge was how to determine the direction of the slant. Should I follow the logic of Zapfino and slant forward (towards the right) or follow the actual slant direction and lean backwards? The latter turned out to be the better choice as opposing directions would have been very disruptive for the flow of bilingual designs, and such a steep angle would close up the counters of left-facing Arabic letters.


The next dilemma was that of structure, which turned out to be a never-ending headache. The problem was none of the known calligraphic styles would fit with Zapfino. The angle, rhythm, and stacking order were all unsuitable. The closest option would be Nastaaliq, which is a back-slanted style that could potentially provide a good starting point. However, Nastaaliq is also characterized by a strongly slanted baseline (the words climb upwards) and that would definitely create a very different rhythm. So, the design would need to use the character shapes of a Nastaaliq but combine the structure of a more horizontal stacking style like Naskh.

Designing the isolated shapes was easy and fun: I would look at Nastaaliq references, and draw letters that have a similar yet more exaggerated structure (Fig. 1). However, it wasn’t long before the design hit a wall. Designing the connected forms was a disaster. The combining logic of Naskh did not work on Nastaaliq shapes, and the end result was too stiff and unyielding. This bit of ugliness I did not share on social media. How could I? The stacking structures, if left to follow Nastaaliq logic, created holes in the text and at times it was difficult to imagine that the design would ever work. A new solution was needed.

“Decided to push Zapfino Arabic a bit more in the Naskh direction so am redrawing almost everything and adding a few hundred glyphs. Brilliant idea.” — Facebook status: March 18, 2014


Most of the Nastaaliq influence had to go and the typeface moved into that of a Naskh structure that leans backwards (Fig 2). The isolated and final shapes of some characters like Beh, Ayn and Jim retained some of that Nastaaliq flavor, but the stacking logic of Naskh demanded Naskh structures, and at the end of the day, that was the right way to go (Fig 3). The word shapes felt right, and the flow of text had an authentic look to it. The shape of each letter was tied to the logic with which it connected to its neighbors. Though there was room for playing, the process of how letters become words had its own set of expectations.


“I could draw as many pretty shapes as I want, but they won’t become letters until they conform to the conventions of structure & proportion.” — Facebook status: March 19, 2014

A typeface, even one as calligraphic as Zapfino Arabic, is meant to be read. It will not be real until it meets readers’ expectations of what a letterform should look like. Because it was an invented hybrid solution, it needed to feel like a natural extension of the letters we know and love, rather than an entirely new typeface. It can look fresh, special, unique, and as contemporary as you want it to be, but it can’t be new. How can one recognize a new symbol, if it is not a natural evolution of the symbols that people are already familiar with?


As a type designer, my job is to design words (Fig. 4). There was a point in time, when I had designed enough character alternates that could fit the Naskh stacking order and I could finally look at word shapes. In a typical design, I would churn out the basic character set and then start testing. In this case, I had to do the basic set and another 200 characters just to be able to see what the words would actually look like (Fig. 5). This meant that I was designing blind for a significant part of the design process, as I was only able to judge a small set of words or character strings. Once everything was put together, I discovered that the spacing needed to be adjusted. Here we go again!


“Re-spacing Zapfino Arabic… Or in other words, redrawing every connecting stroke. ‪#‎typedesignOCD‬” — Facebook status: Aug. 13, 2014‬‬‬‬‬‬‬‬‬‬‬


Once the word shapes were looking half decent, I started to notice that the rhythm of lines and paragraphs were off. Some words were very flat, others stacked up high. Thanks to valuable feedback from non-Latin design and technology expert John Hudson, who also advised on several aspects of the design, I introduced a slight slant to the baseline (Fig. 6). The incoming stroke is a bit higher than the outgoing one. That meant even the flat connections bring in a bit of a natural slope to the writing. This improved the flow, and after many rounds, the design was finally getting closer to being finished.


Real democracy

It’s impossible to talk about Arabic type design and not mention the role that typesetting technology has played in shaping its evolution from handwritten forms to those cast in metal and later on in pixels. For the first time in the history of Arabic type design, it’s possible to create high-quality Arabic typefaces in standard software, and bring in the elegant calligraphic forms that twist and turn in total abandon.

“Confession time: I make funny sound effects (swish-like) when I explain to colleagues what I’m trying to draw in Zapfino Arabic.” — Tweeted on April 3, 2014

This is thanks to three factors: the Unicode Standard, which has standardized text definitions; the OpenType font format, which allows the kind of complex shaping that Arabic sometimes requires, and the Glyphs font editor, which has many features that are quite handy for designing in Arabic. Among those features:

• The ability to design glyphs in context allows the designer to see what the whole word looks like and go in and edit each shape in real time.

• OpenType features can be activated, allowing designers to edit the specific glyph required in each sequence.

• It previews cursive attachment so designers can visualize how the letters cascade.

• The creation of the OpenType features is very straightforward and simple so a designer with a passing understanding of OpenType would be able to, with a reasonable amount of effort, create very complex and sophisticated features.


Glyphs allows designers to overcome the barrier between design and its technical implementation (Fig. 7). If the design requires another set of 50 alternates, I just go draw them, create the required classes and with a few lines of code the job is done. I even managed to do contextual kerning: adjusting the space between two glyphs taking into consideration the next five characters in that sequence. I didn’t even know that this was possible.

“Writing my contextual behavior feature for Zapfino Arabic straight in @glyphsapp. No VOLT. You should see the smile on my face!! :D” — Tweeted on Jan. 17, 2014

Is this not liberation? Is it not true democracy when non-Latin scripts get as much technical freedom and support as Latin? This was my reaction after switching to Glyphs:

“Did not want to leave the office! For the first time in months the design of Zapfino Arabic is not a struggle but addictive joy. Have a nasty cold, but was glued to my laptop, sniffing nonstop and drawing like there is no tomorrow. All for the love of design.” — Facebook status: Dec. 2, 2013

Letting go

A typeface is never really finished. At some point in time, you stop redesigning every glyph, you make an assessment that this is probably as good as it will get at this point in time, and you just give it up. Sometimes you stop too early and regret it afterwards, and sometimes you get to a point that you can’t believe that this is your own work.

With Zapfino Arabic, difficult as the design was, letting go was scary. Is it good? Did I get it right? Is it comfy to read? I kept posting requests for feedback online, as I really needed to make sure that my hybrid solution was easy to read and that it felt natural to readers. The feedback was positive, and I will be eternally grateful for the suggestions, the messages of support and the luxury of time that I was able to spend on this project.

“Thank you all for the feedback and the support. This is one big design studio :) #ZapfinoArabic” — Tweeted on Aug. 23, 2014

This was an experience like no other, and one that I had wanted to capture in all of its raw emotions. Too often we talk about typefaces long after we are done with their design. Though to a large extent that is still ok, the après-design discussion cannot adequately portray the euphoria and yes, the terror, of finally finishing a difficult typeface.

“Zapfino Arabic is a couple of hours away from being done. How does it feel? Like I climbed Mount Everest, only to discover a cliff on the other side and now I must jump.” — Oct. 9, 2014


Zapfino Arabic is available from Linotype.

By Nadine Chahine.

previous post: Notes on the first Books Printed in Italy

next post: This Month in Typography