Serbian Cyrillic from monastery walls to screens and Unicode

Most people who ask where Cyrillic comes from already have the answer wrong before they finish the question. They say: Saints Cyril and Methodius invented it. Partially true, and also backwards.

Cyril invented Glagolitic. The round, ornate, deliberately alien-looking script he and his brother Methodius created around 862–863 AD was designed to write Old Church Slavonic for the Slavic peoples of Moravia — and it was designed to look like no existing script, because that was politically safer. If it resembled Greek, the Greek clergy would claim authority over it. If it resembled Latin, Rome would. So Cyril made something new.

Cyrillic — the script we actually use — came later, probably in the 890s, developed by Clement of Ohrid, one of Cyril and Methodius's disciples, at the Preslav literary school in what is now Bulgaria. He named it after his teacher. It borrowed heavily from the Greek uncial alphabet, which made it easier for Greek-educated clergy to adopt. Glagolitic was more original; Cyrillic was more practical. Practical won.

The script arrived in Serbian lands through the medieval church. The Nemanjić dynasty — the ruling house that built Serbia into a regional power between the 12th and 14th centuries — was deeply tied to the Orthodox church, and the church wrote in Cyrillic. The oldest surviving monument of Serbian Cyrillic is the Miroslav Gospel, written around 1180–1190 for Prince Miroslav, the brother of Stefan Nemanja. It's sitting in the National Museum in Belgrade now, behind glass, 362 parchment leaves covered in the kind of calligraphy that makes you feel genuinely embarrassed about your own handwriting. The Getty Museum has some of its illuminated pages digitized if you want to look without booking a flight.

The great scriptoria of Serbian medieval culture were mostly in monasteries. Hilandar, founded around 1198 on Mount Athos by Stefan Nemanja and his son Rastko (who would become Saint Sava), became the center of Serbian manuscript culture for centuries. Monks there copied texts, translated from Greek, developed the conventions of Serbian Church Slavonic. If you want to understand why Serbian Cyrillic looked the way it looked for 600 years, Hilandar is the answer.

Printing arrived earlier than most people expect. In 1493–94, Đurađ Crnojević established a printing press in Cetinje, in what is now Montenegro. The Oktoih — a liturgical book — came off the press in 1494, making it one of the earliest printed books in the Balkans. Gutenberg's Bible was 1455. These are not ancient people making primitive marks; this is early modern technology, adopted with real speed.

But early printing changed almost nothing about the script itself. Church printing was conservative by design. The letterforms stayed close to the manuscript tradition. The language stayed close to Church Slavonic, which by the 18th century was increasingly distant from how anyone actually talked.

By the late 1700s, educated Serbs were writing in a confused mixture of Church Slavonic, Russian, and vernacular Serbian — sometimes within a single document. The alphabet had letters no living dialect actually needed. Different writers used different conventions. There was no standard.

This is where Vuk Stefanović Karadžić enters the story, and his entry is loud.

Vuk was born in 1787 in a village in western Serbia, largely self-taught, and possessed of the kind of obstinacy that either ruins a person or makes them historically important. In his case, the latter. His program was simple and completely radical: the written language should match the spoken language. "Write as you speak, read as it is written." Every sound gets one letter, every letter represents one sound.

He published his first grammar in 1814 and his Srpski rječnik — Serbian dictionary — in 1818, both supported by the Slovenian linguist Jernej Kopitar in Vienna. The 1818 dictionary was printed in a reformed alphabet he designed himself: he cut letters that Serbian phonology didn't need (the ones that existed for Church Slavonic or Russian reasons), added new ones the language did need — Љ, Њ, Џ, Ј — and organized everything around actual spoken Serbian.

The reaction from the Serbian establishment was approximately what you'd expect when someone tells an institution that everything it has been doing for centuries is wrong. The Serbian Orthodox Church was furious. The government in Serbia was furious. His books were banned in the Principality of Serbia for decades. He kept writing anyway, from Vienna, Bratislava, and wherever else would have him.

He won, eventually. The reformed alphabet was officially adopted in Serbia in 1868, four years after he died. The man did not live to see his victory, which seems deeply unfair.

What Vuk created is essentially what we have today: 30 letters, perfectly phonemic, one-to-one correspondence between sound and symbol. In practical terms, Serbian Cyrillic is one of the most logically organized writing systems in the world. You genuinely cannot misspell a word if you can say it — there are no silent letters, no irregular spellings, no historical holdovers that the pronunciation left behind centuries ago. English speakers should probably not read this paragraph if they want to maintain any sense of satisfaction about their own orthography.

The typographic history of Serbian Cyrillic in the 20th century has a wrinkle that anyone drawing Cyrillic type needs to understand: the italic conventions.

Russian Cyrillic italic, which became dominant in the early Soviet period, uses cursive-derived letterforms — so the italic б looks like a cursive б, the italic г looks like a cursive г, and so on. These forms became so standardized in Russian typography that when Cyrillic type got exported globally, people assumed they were universal.

They're not. Serbian Cyrillic traditionally used different conventions, particularly for letters like б, г, д, п, and т in italic. The Serbian italic forms often look closer to their upright counterparts, or take different historical cursive shapes entirely. Most commercial typefaces, designed primarily for Russian, get this wrong. A font marketed as "supporting Cyrillic" often means "supporting Russian Cyrillic" — which is fine for Russian, and wrong for Serbian. This matters for anyone actually setting text in Serbian. And it matters to us, because we're a foundry from Serbia, drawing typefaces, and we know how bad it looks when the letterforms are subtly wrong in your own language.

Today, Serbian Cyrillic is constitutionally the official script of Serbia — it's in the 2006 constitution, explicitly. In practice, Latin script is used freely and increasingly, especially online, in advertising, in anything aimed at a younger audience. This is partly a technological legacy: early digital systems handled Latin better than Cyrillic, so a generation of Serbs grew up typing in Latin because it was easier. Unicode fixed the underlying problem, but habits stick.

Whether Cyrillic will lose ground in Serbia the way it has already largely lost ground in Croatia and Bosnia is an open question. Some people will tell you it's inevitable. Others will tell you the script is doing fine and the panic is overblown. I don't know the answer. I do know that when we release a typeface that handles Serbian Cyrillic correctly — with proper italic forms, proper spacing, proper diacritics — it feels like it matters. Or we can make desirable Cyrillic localization.

Maybe that's just the foundry bias talking.

Bibliography and further reading

• Vuk Stefanović Karadžić, Srpski rječnik (1818; revised ed. 1852). The primary source. Both editions are digitized and available through the Digital National Library of Serbia at digitalna.nb.rs.
• Cubberley, Paul. "The Slavic Alphabets." In The World's Writing Systems, ed. Peter T. Daniels and William Bright. Oxford University Press, 1996. Dry but thorough.
• Lunt, Horace G. Old Church Slavonic Grammar. 7th ed. Mouton de Gruyter, 2001. For the pre-Vuk background.
• Birnbaum, Henrik. "On Some Evidence for the Glagolitic Origin of Cyrillic." International Journal of Slavic Linguistics and Poetics, 1–2 (1958). If you want the Glagolitic question in depth>
• Hilandar Research Library, Ohio State University — hilandar.library.osu.edu. Extensive digitized manuscripts.
• Miroslav Gospel facsimile and commentary, National Museum Belgrade — nms.rs.
• Sava Mrkalj, Salo debeloga jera libo azbukoprotres (1810). An earlier, failed attempt at Cyrillic reform — worth reading as the thing Vuk improved on.
• Unicode Standard, Cyrillic block (U+0400–U+04FF) — unicode.org. Essential for anyone working with the script digitally.
• Sofija Stefanović, Srpska ćirilica i njene sudbine — harder to find outside Serbia, but worth tracking down for the typographic history specifically.