Sibyls and Sibylline oracles. 1. The Sibyls

The Sistine Chapel ceiling at the Vatican has paintings of seven ancient Hebrew prophets, but also of five Sibyls — non-Christian prophetic figures. How and why did pagan prophets came to have such a place of honour in the centre of western Christianity?

Central part of the Sistine Chapel ceiling (Michelangelo, 1508–1512)

The full answer is surprisingly long. We have to cover genres of ancient Greek poetry; Roman myth and ritual; ancient Jewish and Christian poems about the end of the world; and also music by a 16th century Dutch composer.

The short answer

An answer that focuses just on the Sistine Chapel looks like this. A 4th century Christian writer, Lactantius, wrote about a canon of ten Sibyls: ancient seers lost in the mists of time. They were pagan, Lactantius says, but they accurately predicted the coming of Christ, so you could say they were ‘virtuous pagans’.

Lactantius’ virtuous Sibyls became influential in 15th century Italy, the so-called Quattrocento, around the time of the introduction of the printing press. They appeared in many artworks and books: they’re a major feature of the floor decorations in Siena Cathedral, for example (1480s). They were a well established motif by the time Michelangelo came to his own fresco painting in the first decade of the 1500s. This is why he included Sibyls alongside Hebrew prophets: they were all considered to be pre-Christian prophets of Christ.

And if that’s all you came here for, great! You’re done now. But there’s a lot more to it ...

The start of a longer answer

We have to distinguish four distinct meanings of ‘Sibyls’. A difficulty with Sibylline things is that when people talk about any of these four meanings, they tend to mix them up. One is just a list of names; one refers to a group of texts that is now lost; two of them refer to real texts that still exist today.

1. The Sibyls themselves: originally there was one mythical woman named Sibylla, but more Sibyls appeared as time went on. In the 1st century BCE Varro drew up a canon of ten Sibyls: the Erythraian Sibyl, the Persian Sibyl, the Cumaean Sibyl, and so on.
2. The Sibylline books. These were real sacred texts in pagan Rome, now lost except for two long quotations. They were used for rituals by an important body of priests, the decemviri, or later the quindecimviri.
3. The Sibylline oracles. These texts still exist today, but they’re not pagan. They’re two collections of Jewish and Christian poems, ranging from the 1st century BCE to the 7th century CE, on a range of religious themes, in ancient Greek hexameter verse.
4. The prophetiae Sibyllarum (‘prophecies of the Sibyls’). This is a set of twelve neo-Latin epigrams written in the Sibyls’ wave of popularity in the Quattrocento, possibly for a theatrical production. They’d be forgotten today if not for a stunning musical setting by Roland de Lassus, for an unaccompanied four-part choir (ca. 1558).

There’s some overlap, so for clarity I’ll consistently use the labels in bold.

Lactantius mixed up the first three meanings. He had a copy of nine of the Sibylline oracles (the Jewish and Christian poems); he thought they must be the same thing as the Sibylline books (the Roman ritual texts); and he attributed each poem to one of the Sibyls in Varro’s canon. This is how he came to the conclusion that the ‘real’ Sibyls genuinely foretold the coming of Christ.

Modern scholars sometimes make similar mistakes. But in terms of origins, these four meanings are quite distinct.

The Erythraian Sibyl as she appears on the pavimento of Siena Cathedral (ca. 1480)

1. The ten Sibyls (or one, or three, or eleven)

Once upon a time, in sixth century BCE Greece, there was a genre of oracular poems that dealt with things like directions on performing rituals, enigmatic animal imagery, and foretelling the future. These are the so-called ‘oracle collectors’, or chresmologoi (χρησμολόγοι).

The poems were real, but the oracle collectors were fictional personas. Real poets would adopt a persona, ‘becoming’ Abaris or Orpheus or Epimenides, and incorporate the character and their fictional biography into real poems. Think of them as stage acts, a bit like Ziggy Stardust or Slim Shady. Except a persona wasn’t the personal property of a specific real poet: different poets in different periods could ‘be’ Orpheus or Linos. We hear of one poet, Onomakritos, getting into trouble for adopting the persona of Mousaios. The personas were both men and women — men like Bakis, Epimenides, and Euklos; women like Phemonoe, Boio, and Sibylla.

Herodotos 7.6. Herodotos doesn’t understand that Onomakritos is taking on the role of Mousaios: he represents him as adding an interpolation into the text of Mousaios.

That is, ‘Sibylla’ started out as a specific person. A fictional person, but still, a single individual. References to Sibylla — as a name, not ‘the’ Sibyl — appear in Herakleitos (500s BCE), Aristophanes (400s), and Plato (300s).

But a few decades after Plato, Philitas of Kos wrote that there were three Sibyls: one from Delphi (central Greece), one from Erythrai (East Ionia, in Anatolia), and one from Sardeis (Lydia).

There were three Sibyls. One of them was Apollo’s sister, as she says in her own poem. The second was the Erythraian. The third was from Sardeis. In the same way there were three Bakises: one from Eleon in Boiotia, one Athenian, and one Arkadian.

From this point on, Sibylla was no longer a specific person. Instead, Sibyls were a class of prophetic women.

Herakleitos, 22 B 92 Diels-Kranz (= Plutarch, Moralia 397a); Aristophanes, Knights 61, Peace 1095, 1115–1116; Plato, Phaidros 244b; (pseudo-?)Plato, Theages 124d; Philitas of Kos reported in schol. Aristophanes, Birds 962 (cited as ‘Philetas of Ephesos’).

The next we hear of Sibyls is in the 1st century BCE. At some point between 64 and 49 BCE, the great Roman scholar Varro wrote a book called Ancient matters concerning the gods (Antiquitates rerum divinarum). The text is lost, but a few hundred fragments survive.

Edition: Cardauns 1976. For the date: Varro dedicated the book to Julius Caesar as pontifex maximus: so it is later than Caesar’s appointment to that role in 64 BCE, and most likely before his appointment as dictator and the outbreak of civil war in 49 BCE.

Fragments 56a–60 deal with a canon of ten Sibyls, distinguished by where they were based. Varro’s interest wasn’t purely antiquarian: he was writing not long after the loss of the so-called ‘Sibylline books’ in a fire in 83 BCE — we’ll come back to them in Part 2 — and the Romans farmed out the job of replacing the burnt books to a number of oracular institutions all over the place.

In that context, it makes sense that a canon of Sibyls would be quite important to real religious practice. It also explains why Varro’s canon is organised by where each Sibyl was based. Varro cited his sources too — though the sources are all lost, just like Varro’s own book, and our source doesn’t repeat all of Varro’s citations.

The fact that he cited a different source for each Sibyl indicates that they each had a distinct backstory — they started out as independent characters. They weren’t all named ‘Sibylla’: Varro reports personal names for several of them, and we hear of other names elsewhere. So Varro says the Cumaean Sibyl was named either Amaltheia, Herophile, or Demophile; another list of the ten calls her Taraxandra, Vergil calls her Deiphobe, pseudo-Aristotle calls her Melankraira. Take your pick.

Several later sources report Varro’s canon of ten Sibyls. The following is a synopsis.

1. Persian Sibyl (or Chaldaian, or in two lists, Hebrew); recorded by Nikanor (= FGrHist 146 F 1b). Most reports (not Lactantius) give her name as Sambethe.
2. Libyan Sibyl; mentioned in the prologue of Euripides’ play Lamia (= play description w/o no. ed. Kannicht). In this usage, ‘Libya’ refers to the whole of the Maghreb from Morocco to modern Libya.
3. Delphic Sibyl; recorded by the Stoic philosopher Chrysippos in his book On Divination (= Stoicorum veterum fragmenta ii.1216).
4. Cimmerian Sibyl (in Italy); recorded in Naevius’ Punic War (= fr. 17 ed. Warmington) and Calpurnius Piso’s Annals (= FRomHist 9 F 2). ‘Cimmerian’ in this context means a priestly group attached to the nekyomanteion of Lake Avernus in Campania, near Cumae. We don’t know whether Varro imagined this to imply a connection to the Cumaean Sibyl. The Cimmerian Sibyl is stated to be the mother of Evander in two lists.
5. Erythraian Sibyl (in Asia Minor); recorded by Apollodoros of Erythrai (= FGrHist 422 F 1).
6. Samian Sibyl; named Phoito (spelled Phyto in most sources, but Phoito is the older spelling: see Tibullus 2.5.68. Note that these spellings were homophonous by the 1st century BCE); recorded by Eratosthenes (= FGrHist 241 F 26a).
7. Cumaean Sibyl; named either Amaltheia (or Amalthia), Herophile (or Hierophile), Demophile (only in Lactantius), or Taraxandra; alternatively Deiphobe (Vergil) or Melankraira (ps.-Aristotle). This is reputedly the Sibyl that sold the Sibylline books to the king of Rome (see Part 2).
8. Hellespontian Sibyl (or Gergithian); recorded by Herakleides of Pontos (= fr. 131a ed. Wehrli). She allegedly lived at the time of Solon and Cyrus.
9. Phrygian Sibyl.
10. Tiburtine Sibyl (at Tibur, now Tivoli); named Albunea or Abounaia.

The lists: Lactantius, Institutiones divinae 1.6.7–12 (= Varro, Antiquitates rerum divinarum fr. 56a ed. Cardauns); Souda σ.361; Sibylline oracles prolegomena 32–50 ed. Geffcken; schol. Plato Phaidros 244b; Theosophia 3.A.1.3 ed. Beatrice (≈ Tubingensis epitome 75–77 ed. Erbse); John Lydus, De mensibus 4.47.3 (Lydus adds an eleventh Judaean Sibyl, named Chaldis). Various Souda articles deal with individual Sibyls or alternative lists: β.354 (Delphic and Chaldaian); γ.355 (Erythraian, Sicilian, Leukanian, Sardinian, Gergithian, Rhodian, Libyan, and Samian); ε.357 (Kolophonian, apparently identified with Herophile in Pausanias 10.12.5); ζ.358 (Thessalian, named Manto, another famous chresmologic persona); η.359 (Phrygian, named Sarysis, Kassandra, or Taraxandra); and θ.360 (Cumaean and Thesprotian). Cf. also Clement of Alexandria, Stromateis 1.21.108.1–3, referring to Phrygian/Delphic, Erythraian, Egyptian, and Italian Sibyls (the first two = Herakleides of Pontos fr. 130 ed. Wehrli); and 1.21.132.3, with a list of nine: Samian Sibyl, Kolophonian Sibyl, Cumaean Sibyl, Erythraian Sibyl, Phyto, Taraxandra, Macedonian Sibyl, Thessalian Sibyl, Thesprotian Sibyl. Pseudo-Aristotle, Mirabilia 838a,5–10, states that Sibylla (or ‘the’ Sibyl?) was a chresmologos at Cumae, and that she was also called Melankraira; pseudo-Aristotle, Problemata 954a,34–38 refers to ‘Bakises’ and ‘Sibyls’ collectively.

If you read all of the note in the textbox above, you’ll see that there’s a mass of material: it isn’t very consistent, and over time there were major shifts in how people described the Sibyls.

One important thing to emphasise is that there are no stories or myths about the Sibyls collectively. A couple of them have stories about them. The Cumaean Sibyl is important in the origin story of the Sibylline books (see Part 2); some like Phoito have an independent existence to an extent.

The confusion of names probably comes from attempts to harmonise multiple disparate traditions. The separate listing of the Cimmerian and Cumaean Sibyls is always going to be a bit puzzling, given that they’re both based at Cumae.

One source, Pausanias (10.12.1–6), tries to resolve some of the multiplicity of Sibyls by identifying Herophile as the Sibyl, and saying that she moved around — from Samos to Kolophon, Delos, Delphi, and the Troad — and that Erythrai also laid claim to her. As a result, in Pausanias, Herophile accounts for four of Varro’s Sibyls (Samian, Delphic, Hellespontian or Phrygian, and Erythraian).

The only ones that seem to have an irresolvably separate existence are the Persian, Libyan, and Tiburtine Sibyls — and there are probably sources out there that I haven’t found who identify them with someone else. The Tiburtine Sibyl seems to be the safest: her name looks suspiciously Latin — Abunaea or Albunea or something of the sort — which tends to shield her from the harmonising we see in sources written in Greek.

In Part 2 we’ll turn to the Sibylline books, and what’s known about them. In the meantime, for further reading on chresmologoi and the Sibyls, Parke’s (1988) book on Sibyls is excellent (if a bit overwhelming).

References

• Beatrice, P. F. 2001. Anonymi monophysitae Theosophia. An attempt at reconstruction. Leiden.
• Cardauns, B. 1976. M. Terentius Varro. Antiquitates rerum divinarum. Teil I: die Fragmente. Wiesbaden.
• Diels, H.; Kranz, W. 1960 [1903]. Die Fragmente der Vorsokratiker (6th ed.), 3 vols. Berlin. [Internet Archive]
• Geffcken, J. 1902. Die oracula Sibyllina. Leipzig. [Internet Archive]
• Parke, H. W. 1988. Sibyls and sibylline prophecy in classical antiquity. London/New York.
• Warmington, E. H. 1936. Remains of old Latin, vol. 2. Cambridge (MA). [Internet Archive]
• Wehrli, F. 1953. Die Schule von Aristoteles. Texte und Kommentar. Herakleides Pontikos. Basel.

And the word was God (John 1.1)

Ἐν ἀρχῇ ἦν ὁ λόγος, καὶ ὁ λόγος ἦν πρὸς τὸν θεόν, καὶ θεὸς ἦν ὁ λόγος.

In the beginning was the word, and the word was an attribute of God, and the word was God.

John 1.1 (my translation)

This isn’t a topic I’d normally write about, because it’s so basic: it seems like low-hanging fruit. But I was a little staggered today to be directed to a Youtube video where a biblical scholar, with a PhD, who ostensibly knows Greek, spends 45 minutes discussing theological reasons why the phrase θεὸς ἦν ὁ λόγος (‘the word was God’) doesn’t have a definite article before the word ‘God’.

The opening of John in the codex Sinaiticus, fol. 247r. (Image source: Codex Sinaiticus website)

The video is sensationalised, to be sure — the title has an emoji! But the reason ‘God’ doesn’t have an article is a very simple grammar rule. When you’re learning ancient Greek, it’s covered in week 1. ‘God’ can’t have an article. Because it’s a complement.

Articles in ancient Greek

Ancient Greek had a definite article, ὁ ἡ τό (and other inflected forms), corresponding to English ‘the’. It started out as a demonstrative pronoun, and acquired its role as an article in the Archaic period, ca. 600 BCE. It has hung around to the present day. If an ancient Greek-speaker wanted to refer to a specific individual thing, or person, they normally put an article before it: ὁ ἄνθρωπος ‘the person’, ἡ ἀγορά ‘the agora’, τὸ δένδρον ‘the tree’, and of course ὁ θεός ‘the god’.

The article is common (but not used invariably) in some situations where English doesn’t use an article, particularly with names. It’s perfectly normal to refer to Sokrates as ὁ Σωκράτης. In the Bible it’s perfectly normal to refer to God as ὁ θεός. So you have to rely on context to tell whether ὁ θεός should be translated as ‘the god’ or ‘God’.

Strictly speaking it might be best to refer to the article not as ‘definite’, but as ‘definitive’, since it often plays a deictic role: Sokrates, that guy right there. Modern German has a loose parallel, in the use of the definite article before personal names: the difference between saying die Ulrike and just Ulrike is subtle, but the first one carries a reminder of which Ulrike we’re talking about.

For details of when names do and don’t have articles in Greek, see Smyth 1916: 211–213 (§768–779).

Ancient Greek didn’t have an indefinite article, corresponding to English ‘a, an, some’. (Modern Demotic Greek does, but it’s a much later arrival.) So when a noun appears in ancient Greek without an article — δένδρον instead of τὸ δένδρον — it will generally be because it’s indefinite: ‘a tree’. There are cases where the article can be optional, as with personal names.

There are also situations where an article is invariably used, and situations where an article is invariably omitted. And one of those situations is a sentence of the form ‘A is B’.

The confusion

C. J. Cornthwaite starts with the question of whether θεός is substantival (‘the word was God’) or adjectival (‘the word was divine’). That is technically a legitimate question — though there’s no reason to prefer reading a noun as adjectival when it works perfectly well as a noun. Anyway, in thinking about that, he spends a great deal of time questioning why it’s θεός instead of ὁ θεός, and what the significance of that is.

That tells me there’s a serious problem with how Greek is being taught. But just as importantly, because he’s a very popular para-academic, he’s going to be leading a lot of people astray. So it seems like it’d be wise to address the matter.

Here’s the bit where he states the options as he sees them:

So, the way I see it, there are three different ways to read this text based on Greek grammatical rules.

So. The first is just reading this as saying ‘the Word was God’. Basically like, we're talking about God, you know, ... the one true God. ...

The second way to read this would be to say that θεός here is actually functioning as a way to signify that the Word was ‘divine’ or ‘godlike in quality’. ...

And the third way that you could translate this is the way that the Jehovah's Witness Bible translated it, as ‘the Word was a god.’ So, this kind of sounds like you're saying it was, like, one of many, because it's indefinite.

In terms of understanding this problem, those are the three options. And all of them come from the difficulty of knowing why John left the article off here, and what that is supposed to mean — why it says, basically, ‘god’ instead of ‘the God’. That’s the question.

Cornthwaite 2025, 6:03–7:30

and then he spends several minutes talking about how the Septuagint, an ancient Greek form of the Jewish scriptures, sometimes refers to God without an article, and why that might be. His discussion involves logos theology, construct chains in Hebrew syntax, and why translators ‘chose’ to omit an article in some situations.

What the discussion doesn’t involve is Greek syntax. The answer is so much more basic. It has nothing to do with theology, and it isn’t a choice.

The ancient Greek complement: an introduction

A sentence of the form ‘A is B’ has three elements: A, also called the ‘subject’; a copulative verb, like ‘is’ or ‘was’; and B, also called a ‘complement’.

It’ll always be obvious which bit is the verb. But A and B are the same part of speech: they’ll both be nouns, or names, or some other kind of substantive. Yet they have distinct syntactical roles. So how do you tell which is which?

In English, it’s done by word order. The sequence of the three elements is fixed: A, verb, B. (Unless you’re Yoda, in which case it’s A, B, verb.) You always know that the thing that appears first is the subject, and the thing that appears later is the complement.

But in ancient Greek, the three elements can be placed in any order. Word order doesn’t tell you which is A and which is B. So the way Greek tells you which is the subject and which is the complement is by using ... (drumroll) ... an article.

The rule is very simple and universal:

• The subject, A, is indicated by the presence of an article.
• The complement, B, is indicated by the absence of an article.

Actually the rule is more general than this: it applies to all kinds of predicate substantives, not just in ‘is’ sentences, e.g. θεὸς Ἰσραὴλ πάσῃ τῇ γῇ κληθήσεται ‘he will be called the God of Israel over the whole earth’ (LXX Isaiah 54.5).

As a result, the following all mean the same thing, just with slight shifts in emphasis:

ὁ Ὅμηρος ἦν ῥαψῳδός.
ῥαψῳδός ἦν ὁ Ὅμηρος.
ὁ Ὅμηρος ῥαψῳδός ἦν.
ἦν ῥαψῳδός ὁ Ὅμηρος.
ἦν ὁ Ὅμηρος ῥαψῳδός.
ῥαψῳδός ὁ Ὅμηρος ἦν.

Homer was the rhapsode or Homer was a rhapsode.

But they swap around if you just move the article:

Ὅμηρος ἦν ὁ ῥαψῳδός.
ἦν ὁ ῥαψῳδός Ὅμηρος.

The rhapsode was Homer.

And this is all that’s going on in John 1.1 θεὸς ἦν ὁ λόγος. λόγος has an article to show that it’s the subject; θεός has no article, to show that it’s the complement.

There’s no deep theology going on there. Well, nothing deep about the article at least — this still doesn’t tell you whether to translate θεός as ‘God’ or ‘divine’ — but that question has nothing at all to do with articles. θεός will still be a complement, so either way it won’t have an article.

Don’t believe me? Fine: here’s a standard textbook, Reading Greek, 2nd edition, published by the Joint Association of Classical Teachers. Here’s a snippet from the introductory chapter, Section I.

§47 in Reading Greek, JACT 2001, page 37.

Athenaze covers this too, but only by example: it doesn’t state it as an explicit general rule. It’s also explained explicitly in Smyth’s Greek grammar (Smyth 1916: 214, §784–786) and the Cambridge grammar of classical Greek (Van Emde Boas et al. 2019: 330, §28.9), but because they list some exceptions where both A and B have articles, that may give a false impression that the rule is somehow optional. (There are no exceptions about the absence of an article, which is the situation in John 1.1.) Reading Greek is refreshing because it focuses on telling the beginner the thing that they need to know.

It boggles my mind that actual scholars could have any uncertainty over this. It makes me wonder what Greek textbook is being used in Religious Studies departments.

By the way, my translation at the start isn’t ideal: a first century reader would certainly understand the metaphysical language in John 1.1 in Stoic terms, as I explained here a couple of years ago. It’s just that that would be a bit too unfamiliar to hit modern readers with right away. Someone living in the philosophical framework of the time would most naturally interpret John 1.1 as

Ἐν ἀρχῇ ἦν ὁ λόγος, καὶ ὁ λόγος ἦν πρὸς τὸν θεόν, καὶ θεὸς ἦν ὁ λόγος.

In the beginning was the principle of intelligibility, and the principle of intelligibility was an attribute of God, and the principle of intelligibility was God.

John 1.1 (my preferred translation)

References

• Cornthwaite, C. J. 2025. ‘The Origins of John 1:1 will BLOW YOUR MIND! 🤯 - WARNING! Greek & Hebrew DEEP DIVE.’ ‘C. J. Cornthwaite’ channel, 3 April 2025. [Youtube]
• van Emde Boas, E.; Rijksbaron, A.; Huitink, L.; de Bakker, M. 2019. Cambridge grammar of classical Greek. Cambridge.
• JACT 2001. Reading Greek: grammar and exercises, 2nd edition. Cambridge.
• Smyth, H. W. 1916. A Greek grammar for schools and colleges. New York. [Internet Archive]