Read the roll of fallen angels in 1 Enoch chapter 19 and you will find Turiel named three times and Azazel named twice, in a single list of twenty-one. Nothing separates the repeats, and nothing in the text explains them. Readers write to us about it, which is reasonable, because in almost any other book that would be a printing error.
It is not. The repetition is in the Ge'ez, and it is the visible end of a chain of copying that ran through three languages over roughly two thousand years. Once you know what happened to the book, the doubled names stop being a defect and become the clearest evidence in it.
1 Enoch was written in Aramaic, translated into Greek, translated again into Ge'ez, and only then into English. A proper name is the part of a text least protected by context, so it is the first thing to drift. The angel lists are where that drift is visible to an ordinary reader.
Start at the front of the book. Two hundred angels swear an oath on Mount Hermon, and the text then names their leaders and calls them "captains, leaders over groups of ten." Two hundred angels in groups of ten wants twenty captains. The list gives nineteen.
That short count is old, it is in the Ethiopic manuscripts, and it is not something an English edition can quietly repair. Somewhere between the Aramaic and the Ge'ez a name fell out of a list, and no surviving manuscript tells us which.
Then turn to chapter 19, where a second roll of angels appears. Compare the two and the second one is not an independent tradition at all. It is the first list again, coming back in damaged form:
| First list (chapter 2) | Second list (chapter 19) |
|---|---|
| Semyaza | Semyaza |
| Kokabiel | Kokabel |
| Baraqiel | Baraqel |
| Batarel | Bataryal |
| Ananel | Hananel |
| Rameel | Rumael |
| Turiel | Turiel, Turiel, Turiel |
Every pair in that table is one angel wearing two spellings. And in the second list the spellings have multiplied far enough that a copyist meeting three forms of Turiel wrote all three down, as though he had three angels in front of him rather than one name that had been through three alphabets.
The book's route to English is unusually well documented, because pieces of it survive at every stage.
Aramaic. The original. Fragments of eleven manuscripts turned up in Cave 4 at Qumran, the oldest copied in the second century BCE, published by J.T. Milik in 1976. They are fragments, not a book. No complete Aramaic Enoch exists anywhere.
Greek. Someone translated it into Greek, and three witnesses survive. A codex dug up at Akhmim in Egypt in the winter of 1886 carries chapters 1 to 32, bound alongside the Gospel of Peter and the Apocalypse of Peter. The Byzantine chronicler George Syncellus, writing around the year 800, quotes chapters 6 to 10 and 15 to 16 in his chronicle, which happens to include the first angel list. A papyrus split between the Chester Beatty and Michigan collections carries chapters 97 to 107.
Ge'ez. The Ethiopic translation was made from the Greek, somewhere between the fourth and sixth centuries, along with the rest of the Ge'ez Old Testament. This is the only version of 1 Enoch that survives entire, anywhere, in any language. Everything else is pieces. The Ethiopian Orthodox Tewahedo Church has kept it in the canon and read it continuously ever since, which is the only reason the book exists to be argued about.
English. James Bruce came back from Ethiopia in 1773 with Ge'ez manuscripts, one of which reached the Bodleian. Richard Laurence published the first English translation from it in 1821. August Dillmann edited the Ethiopic text in 1851. R.H. Charles published his translation in 1893 and revised it in 1912, and that 1912 text is the one nearly every modern English Enoch descends from, ours included.
Now put a name into that chain. Aramaic writes it with no vowels. Greek supplies vowels, and has no letter for some of the consonants it is trying to carry. Ge'ez re-vowels it again from the Greek. Each stage has a scribe who has never met the angel and has no way to check the spelling against anything, because the name occurs once. Meaning protects ordinary words in a sentence; a proper name in a list has no such protection. So the lists drift, and they drift faster than the narrative around them.
There is a second, smaller confusion sitting on top of the first, and it is worth clearing up because it makes the same angel look like two more.
Charles was writing for scholars in 1912, using the transliteration habits of his time. The circumflexes in Semjâzâ and Jômjâêl mark Ge'ez vowel values. More importantly, his letter j is doing the job of an English y, which is the German convention he had inherited. So:
Editions that reprint Charles without touching this leave the reader looking at Semjaza in one book and Semyaza in another and reasonably concluding they are different beings. They are the same angel, spelled by a man who expected his readers to know what his j meant.
Our edition prints the standard modern forms throughout and drops the diacritics. That is a spelling decision, not a translation decision, and it does not move a single word of the text.
We left them.
Deduplicating the list would have produced a tidier page and a false one. Turiel is in that list three times because the manuscripts say so, and an English edition that silently reduces it to once has edited the evidence out of the book while telling the reader nothing. The same argument applies to the nineteen captains who ought to be twenty.
So both lists carry a short editorial note saying plainly what the reader is looking at: that the repetition is in the Ge'ez as Charles rendered it, that it is not an error in this edition, and that the doubling is itself a record of how the book travelled. The note is set apart from the text so nobody mistakes our voice for Enoch's.
That is the general rule we work to. Where a text is strange, the strangeness is usually the text. Numbers that do not add up, lists that miscount themselves, an angel who arrives three times under three spellings: assume it is the book until a manuscript proves otherwise, and say so in a note rather than fixing it in silence.
Because the list was copied through three languages before it reached English. 1 Enoch was written in Aramaic, translated into Greek, then into Ge'ez, and a proper name in a list has no surrounding meaning to protect its spelling. By the time the second roll of angels was copied, one angel's name existed in several forms, and a scribe who met three of them wrote all three down. Turiel appears three times in that list, and Azazel twice.
No. It is in the Ethiopic manuscripts, and it is in R.H. Charles's 1912 translation, which nearly every English edition of Enoch descends from. An edition that prints the name once has removed something the manuscripts contain.
Aramaic, with parts possibly in Hebrew. No complete Aramaic copy survives. Fragments of eleven manuscripts were found at Qumran in Cave 4, the oldest copied in the second century BCE and published by J.T. Milik in 1976.
Because it is the only complete version that survives in any language. The Greek exists only in pieces and the Aramaic only in fragments. The Ge'ez translation, made from the Greek between roughly the fourth and sixth centuries, preserved the whole book, and the Ethiopian Orthodox Tewahedo Church has kept it in its canon ever since.
They are the same name. Charles used the scholarly transliteration of his day, in which the letter j carries the sound of an English y. Semjaza, Jomjael, Baraqijal, Rumjal and Danjal are Semyaza, Yomiel, Baraqiel, Rumyal and Danyal. Editions that reprint Charles unaltered keep his spelling; ours uses the modern forms.
The arithmetic in the text says so. Two hundred angels descend, and the list is described as their captains over groups of ten, which wants twenty captains. Nineteen are named. The shortfall is in the Ethiopic manuscripts, and no surviving witness shows which name was lost.
Related reading: