How to adapt a Western font for Arabic or Hebrew
Why simply "redrawing" Latin doesn't work
A Latin font is a set of isolated signs of fixed form. Arabic is a contextual system: the same letter changes its shape depending on its position in the word (isolated, initial, medial, final), and these forms are connected by ligatures. If a designer mechanically transfers the proportions of the Latin design to Arabic glyphs without understanding the calligraphic logic, the result is the effect industry calls "Latinized Arabic" — the letters are visually recognizable, but the native speaker immediately sees that the contours are not connected as they should be, angles are not where a calligrapher's hand would place them. It works like incorrect accent in a foreign language: the meaning is understood, but it sounds foreign and causes distrust in the brand.
The problem with Hebrew is different: the alphabet has no lowercase and uppercase letters, and all signs are visually closer to Latin capitals. Hence, a typical error is either Hebrew text looks compressed against the backdrop of Latin, or, on the contrary, "presses" it with its bulk.
How to choose a calligraphic style for Arabic
There is no universal correct answer here — it is a question of the function of the text.
- Naskh — the workhorse for long texts: rounded, readable, the standard for UI and body copy;
- Kufi — geometric, angular style, closer to display fonts, works well with grotesque Latin in logos, but is poorly readable in large blocks of text;
- Nastaliq — historically used for the Persian language, visually cascading and expressive, but modern digital platforms often replace it with Naskh simply because quality Nastaliq fonts with full OpenType support are still rare.
A hybrid approach is not a compromise of 'both this and that,' but a conscious decision: the display version of the logo can lean towards Kufi for geometry, while the text version for the interface is strictly Naskh. Mixing them at the same hierarchy level is a mistake.
Case: how Adobe adapted Myriad
A telling example of an engineering approach is the Arabic version of Myriad. When directly transferring Latin proportions, Arabic text in interfaces appeared smaller and lost readability due to the specifics of the interline spacing in Arabic script, where diacritics and connecting strokes occupy vertical space differently than in Latin. The solution was not aesthetic but metric: Adobe scaled the Arabic version to 1.4x relative to the Latin one, compensating for the difference in optical perception of size. This is a specific figure that can be used as a reference in your own tests, but not as a ready-made recipe — the coefficient depends on the specific font and its x-height.
Hebrew: balance of letter heights
The main technical task with Hebrew is to fit letters that do not have case distinctions into the visual hierarchy of Latin with its x-height and cap-height. The rule is simple, but easy to violate: the height of Hebrew letters should be positioned between the x-height and cap-height of the Latin font.
| Error | Result |
|---|---|
| Hebrew height ≈ x-height | The text looks unreadable, 'sinks' into the line |
| Hebrew height ≈ cap-height or higher | Hebrew visually dominates, Latin seems secondary |
| Height between x-height and cap-height | Both alphabets are read as a single system |
Here we should dispel a common misconception: stylistic associations are not automatically transferred between scripts. A high-contrast Didone reads as a language of fashion brands in Latin — but the same contrast in Arabic or Hebrew can simply break readability or evoke associations unrelated to fashion at all. You must test not "does it look similar visually," but how the language speaker reacts.
Technical checklist before font release
Adaptation is not considered complete until an engineering, not just visual, check is passed. Mandatory checks:
- Contextual forms — all four positions of the letter (isolated, initial, medial, final) are rendered correctly and connect without gaps;
- Kashida — stretching of connecting lines does not distort the proportions of the word when aligning text to width;
- Nukta (Hebrew diacritics) — up to three signs on one letter, placement is configured via OpenType features GPOS MarkToBase, with the diacritic position calibrated relative to the visual, not the mathematical center of the glyph;
- Test on real carriers — at least 2-3 people for whom the language is native, read the text aloud and separately evaluate "does it look familiar";
- Check in conservative and secular contexts — a form acceptable in Lebanon may cause rejection in more traditional regions of the Persian Gulf.
Who to hire and at what stage
If the brand team does not speak the target market language, hiring a consultant is not an option, but a mandatory project condition. A telling example of the correct methodology — Intel Clear Arabic, developed with consultants Christian Sarkis and Thomas Milo: the team consciously chose traditional calligraphic forms instead of modernized ones to avoid provoking a negative reaction on conservative markets. This decision was made at the concept stage, not in the final proofreading.
The practical sequence of hiring is as follows: a typographer-language carrier is engaged at the style selection stage (Naskh/Kufi/traditional Hebrew forms) — before a single glyph is even drawn; a linguistic consultant checks contextual forms and diacritics at the technical implementation stage; a focus group of native speakers tests the final result before release. Hiring a consultant only for the final proofreading is a typical mistake that saves budget but does not solve the problem: by that time, structural errors are already embedded in the font, which are cheaper to avoid than to redo.
The problem with Hebrew is different: the alphabet has no lowercase and uppercase letters, and all signs are visually closer to Latin capitals. Hence, a typical error is either Hebrew text looks compressed against the backdrop of Latin, or, on the contrary, "presses" it with its bulk.
How to choose a calligraphic style for Arabic
There is no universal correct answer here — it is a question of the function of the text.
- Naskh — the workhorse for long texts: rounded, readable, the standard for UI and body copy;
- Kufi — geometric, angular style, closer to display fonts, works well with grotesque Latin in logos, but is poorly readable in large blocks of text;
- Nastaliq — historically used for the Persian language, visually cascading and expressive, but modern digital platforms often replace it with Naskh simply because quality Nastaliq fonts with full OpenType support are still rare.
A hybrid approach is not a compromise of 'both this and that,' but a conscious decision: the display version of the logo can lean towards Kufi for geometry, while the text version for the interface is strictly Naskh. Mixing them at the same hierarchy level is a mistake.
Case: how Adobe adapted Myriad
A telling example of an engineering approach is the Arabic version of Myriad. When directly transferring Latin proportions, Arabic text in interfaces appeared smaller and lost readability due to the specifics of the interline spacing in Arabic script, where diacritics and connecting strokes occupy vertical space differently than in Latin. The solution was not aesthetic but metric: Adobe scaled the Arabic version to 1.4x relative to the Latin one, compensating for the difference in optical perception of size. This is a specific figure that can be used as a reference in your own tests, but not as a ready-made recipe — the coefficient depends on the specific font and its x-height.
Hebrew: balance of letter heights
The main technical task with Hebrew is to fit letters that do not have case distinctions into the visual hierarchy of Latin with its x-height and cap-height. The rule is simple, but easy to violate: the height of Hebrew letters should be positioned between the x-height and cap-height of the Latin font.
| Error | Result |
|---|---|
| Hebrew height ≈ x-height | The text looks unreadable, 'sinks' into the line |
| Hebrew height ≈ cap-height or higher | Hebrew visually dominates, Latin seems secondary |
| Height between x-height and cap-height | Both alphabets are read as a single system |
Here we should dispel a common misconception: stylistic associations are not automatically transferred between scripts. A high-contrast Didone reads as a language of fashion brands in Latin — but the same contrast in Arabic or Hebrew can simply break readability or evoke associations unrelated to fashion at all. You must test not "does it look similar visually," but how the language speaker reacts.
Technical checklist before font release
Adaptation is not considered complete until an engineering, not just visual, check is passed. Mandatory checks:
- Contextual forms — all four positions of the letter (isolated, initial, medial, final) are rendered correctly and connect without gaps;
- Kashida — stretching of connecting lines does not distort the proportions of the word when aligning text to width;
- Nukta (Hebrew diacritics) — up to three signs on one letter, placement is configured via OpenType features GPOS MarkToBase, with the diacritic position calibrated relative to the visual, not the mathematical center of the glyph;
- Test on real carriers — at least 2-3 people for whom the language is native, read the text aloud and separately evaluate "does it look familiar";
- Check in conservative and secular contexts — a form acceptable in Lebanon may cause rejection in more traditional regions of the Persian Gulf.
Who to hire and at what stage
If the brand team does not speak the target market language, hiring a consultant is not an option, but a mandatory project condition. A telling example of the correct methodology — Intel Clear Arabic, developed with consultants Christian Sarkis and Thomas Milo: the team consciously chose traditional calligraphic forms instead of modernized ones to avoid provoking a negative reaction on conservative markets. This decision was made at the concept stage, not in the final proofreading.
The practical sequence of hiring is as follows: a typographer-language carrier is engaged at the style selection stage (Naskh/Kufi/traditional Hebrew forms) — before a single glyph is even drawn; a linguistic consultant checks contextual forms and diacritics at the technical implementation stage; a focus group of native speakers tests the final result before release. Hiring a consultant only for the final proofreading is a typical mistake that saves budget but does not solve the problem: by that time, structural errors are already embedded in the font, which are cheaper to avoid than to redo.
Recent articles
MegaFon, BMW, VTB and more 240+ companies. See how it looks in the showcase reel:
Our branding agency is approached to order:
Branding agency services
