The Armenian Alphabet That Accidentally Preserved How Ancient Greek Sounded
Here's a fun linguistic puzzle: What do you call a document written entirely in Armenian letters that contains zero Armenian words? A headache for librarians, for one thing. But also one of the most fascinating phonetic time capsules in the history of language.
Sometime around the 5th to 7th century AD, someone in Egypt sat down with a sheet of papyrus and transcribed Greek words using the Armenian alphabet. Every. Single. Word. Greek in content, Armenian in script. The document ended up at the Bibliothèque Nationale in Paris (catalogued as BnF Arm 332), where a French scholar named Auguste Carrière bought it from a dealer in the late 19th century, then it basically vanished into the archive fog until historian Dickran Kouymjian rediscovered it in 1993.
For decades, scholars had worked from a photograph of only one side of the papyrus. One side. Of a document that proved to have two. That's either deeply funny or deeply frustrating depending on how much coffee you've had.
Why This Matters to Anyone Who Loves Words
When people wrote languages phonetically using foreign scripts, they captured something priceless: actual pronunciation. Not the "official" pronunciation. Not the literary standard. The real sounds coming out of real mouths at a specific moment in time.
This papyrus is the only known early example of the Armenian alphabet written on papyrus material. That alone makes it remarkable. But what it tells us about Greek pronunciation in late antique Egypt? That's the good stuff.
Three Sound Changes, Frozen Mid-Shift
Languages don't change overnight. There's always a messy middle period where some speakers use the old sound and some use the new one and everyone argues at dinner parties. This papyrus caught Egyptian Greek right in the middle of several of those shifts.
The B-to-V problem. The Armenian alphabet has a letter Բ for the /b/ sound and a letter Ւ that indicated something closer to /v/. Our papyrus author was spelling Greek words that start with B using Բ, not Ւ. So they were still hearing and saying a real /b/. But here's the tell: one word in the document spells out the Greek σάβανον (meaning "linen cloth") as ՍԱՒԱՆ, using that Ւ in the middle. The /b/ in the middle of a word was already drifting toward /v/. The shift was happening, but not evenly. Edge cases first, always.
The kh-to-x problem. The Greek letter Χ started life representing an aspirated /kʰ/ sound, like the K in "kite" with extra breath. It eventually shifted to the fricative /x/ heard in the Scottish "loch." The papyrus author was spelling Χ-words with the Armenian letter Ք, which represented the older /kʰ/. That means as of the 5th to 7th century in this corner of Egypt, the sound had not yet completed its journey. Your pronunciation of "chaos" is further down the road than this scribe ever traveled.
The H problem. The word ՀԷՄԱ in the papyrus transcribes ancient Greek αἷμα, meaning "blood." You know this root: haemoglobin, haemorrhage, haematology. The Armenian letter Հ represents the /h/ sound, so the papyrus author was writing and presumably saying something like "hema." Modern Greek speakers pronounce it "ema," no H at all. H-dropping in Egyptian Greek was already happening at the time, but it was common without being consistent, which suggests it wasn't yet considered standard. The H was on its way out. Just not quite gone yet.
Conservative Tastes in a Changing World
Put it all together and the Egyptian Greek dialect reflected in this document has what scholars call a "quasi-classical" character. Conservative. Holding onto older sounds a bit longer than other dialects. The /b/ was still mostly /b/. The /kʰ/ was still aspirated. The /h/ was still hanging around.
There's also evidence in the document for what linguists call iotacism (the merging of various vowel sounds into a single "ee" sound) and the -ίον diminutive endings that are scattered all through Greek. So it's not frozen in amber. It's moving. Just at its own pace.
The Accidental Gift
Language change is invisible from the inside. You don't notice your vowels shifting. You don't feel a consonant softening over decades of casual speech. That's what makes documents like this so valuable: somebody was just trying to learn Greek. Maybe a student. Maybe a translator. They weren't trying to document phonology. They were trying to get the sounds right.
And in doing so, they handed us a photograph of a language in the middle of becoming something else. The linen cloth word with its half-shifted /b/. The blood word that still remembered its H. The Χ that was still breathing hard.
Next time you type "haemo-" into a search bar, you're using a root that this anonymous scribe in Egypt wrote down as ՀԷՄԱ, in Armenian letters, on a papyrus that sat unexamined for the better part of a century. Language is long. And sometimes it leaves notes.
Source: Languagehat