A study from the corpus
How many angels are actually named?
Angelology looks like a filled-in system: seven archangels with portfolios, nine choirs in three spheres, a name for every office. Most of it is real, in the sense that people held it and wrote it down. Almost none of it is scriptural, in the sense of being in the books.
This is a count of where each name first appears. It is a bibliographic question rather than a theological one — which lists of books received which text — and the answer sorts cleanly, with two names in the protocanonical material and everything else arriving later or elsewhere.
Counted across 47 angelic entries at build time. Corpus last revised 29 August 2026.
Published · Updated
The count
2
Named in every canon
Gabriel and Michael
36
Personal names in all
34 entries, two of which name a pair
13
Not names at all
Ranks, kinds and offices, listed below
47
Angelic entries
Everything this corpus files as angelic
- In every canon2
- Deuterocanonical1
- Outside the canons23
- Qur'an and hadith4
- Later tradition4
Where each name enters
Every named angel, filed by the standing of the earliest text in this corpus that carries the name. The groups are about which canons received a book, not about whether a book is worth reading.
In every canon
2 namesBooks received as scripture by Jews, Catholics, the Orthodox churches and Protestants alike. If a name is here, no tradition has to argue for it.
Deuterocanonical
1 nameReceived by Rome and the East, moved to an Apocrypha by the Reformers. A name here is scriptural for most Christians in the world and not for all of them.
Outside the canons
23 namesBooks that no major Western canon received — 2 Esdras in an appendix, 1 Enoch canonical only in Ethiopia. A name arriving here is popular rather than authorised.
Qur'an and hadith
4 namesA separate scriptural tradition rather than a later layer of the same one, and it names its own. Counted apart for that reason.
Later tradition
4 namesRabbinic codification, patristic and medieval devotion, Renaissance correspondence tables, and the reference works that repeat them. This is where most angel names come from.
The ones that are not names
Counting these as named angels is what produces the inflated numbers you meet elsewhere. Each is a plural, a rank or an office, and each line says which.
The ninth choir: the word itself, used as a rank.
A rank. The word occurs twice in the New Testament and names no one.
The living creatures of Ezekiel, made the first rank in Maimonides' scheme.
Likewise a plural; the singular is a kind of guardian figure known across the ancient Near East.
A rank, from the same list of abstractions in Colossians.
A rank, and one of the plurals Paul uses for authorities of unspecified kind.
A rank, from the same passage.
A Hebrew plural for a kind of being — the burning ones — not anybody's name.
The group. Its membership is named in other books, and this entry is the body itself.
A body rather than a being — the assembly this corpus treats in its own study.
A rank in the Dionysian scheme, taken from a plural noun in Colossians.
A rank; the name is a Latin translation of a Greek abstraction.
The honourable scribes: two angels at each person's shoulders, named by function and never individually.
What the count shows
Two names are in books every canon receives, and both arrive late in the Hebrew Bible: Michael and Gabriel appear in Daniel, and Gabriel again in Luke. Before Daniel, angels in scripture are anonymous. The messenger who wrestles Jacob refuses to give a name; the captain of the LORD’s host is asked and answers with an allegiance instead. The convention is not that angels lack names — it is that they do not hand them over.
Raphael comes from Tobit, which most Christians in the world receive as scripture and Protestants print as Apocrypha. Uriel comes from 2 Esdras, which no major Western canon received, and enters popular reckoning as the fourth archangel anyway. After that the names arrive in books nobody was ever asked to accept as canonical: 1 Enoch supplies most of them, Islamic tradition names its own, medieval devotion and Renaissance correspondence tables supply the rest — and a synod at Rome in 745 cut the licit list back to three, which tells you how many were circulating by then.
None of this makes the later names fraudulent. Traditions elaborate, and the elaboration is a real object with a date, which is what this corpus is for. What the count does is put the weight back where the evidence is: a reader who has met a seven-archangel chart with portfolios and colours should know that two of the seven are in Daniel, one is in Tobit, one is in an appendix, and the rest are somebody’s considered opinion.
Method, and what would change it
Every angelic entry is sorted into a name, a class or an office by an explicit list in the study’s data file, and each name is dated by the earliest source in the corpus that carries it, with that source’s canonical standing attached. Writing this study moved two citations: Raphael’s Tobit passage and Uriel’s 2 Esdras passage were filed under the corpus’s combined scripture row, which is fine everywhere except in a study about where a name first appears.
- A protocanonical book naming an angel this study does not count — the test is a personal name, so Daniel's princes of Persia and Greece do not qualify and neither does the captain of the LORD's host, who is asked his name and does not give it.
- A demonstration that one of the thirteen entries in the not-a-name list is used as a personal name in its own language rather than as a class or an office.
- An earlier attestation for a name this corpus dates late. Camael and Jophiel are the likeliest to move: both are here on early-modern and medieval devotional material, and either could turn out to have a source nobody in this corpus has read.