The Ogdoad of Hermopolis

Egypt had more than one answer to the question of what came first, and they were maintained side by side for thousands of years without anybody appearing to mind.

The famous one is Heliopolitan: Atum alone, who becomes three, then nine — the Ennead. The other is older, according to the man who wrote the standard English account of both, and it belongs to a Middle Egyptian city the Greeks called Hermopolis and the Egyptians called Khemennu: the city of the Eight Gods.

Not one god. Eight. Four males and four females, in pairs, and none of them is a person in any ordinary sense. They are water, boundlessness, darkness and rest.

The eight

Budge gives the roster as "Nu and Nut, Hehu and Hehut, Kek and Keket, and Kerh and Kerhet (or Nau and Nait), i.e., four pairs of deities, each pair consisting of a male and a female deity" [budge-1904]. At their head, in the form the city practised, stands Thoth.

Two rosters, and they disagree

Before going further: the list above is Budge's, and it is not the list you will find in most modern references. This corpus's own census gives the Ogdoad as Amun–Amunet, Nu–Naunet, Heh–Hauhet, Kek–Kauket.

Set them side by side:

Three pairs are shared. The difference is at the ends: the modern roster opens with Amun and Amunet and has no fourth pair of night-gods; Budge opens with the water and closes with Kerh and Kerhet.

And Budge explains the divergence without knowing he is doing it. In his account of the fourth pair he notes that "the texts in some places give quite different names where we should expect to find theirs" — and among the substitutes for Kerh he lists Amen [budge-1904].

So the standard modern Ogdoad is the roster that took the Amun substitution as canonical, promoted him to first position, and dropped the pair he had been substituting for. That is not an error in either list. It is a real historical process — Thebes writing itself into the oldest company in Egypt — visible in the gap between a 1904 reference and a modern one.

This corpus keeps Budge's eight, because they are the eight the archived source documents and because Amun already has his own entry as the great god of Thebes. But the standard roster is the one a reader will meet elsewhere, and the difference is the interesting part rather than a footnote.

What this cosmogony actually claims

Read the four pairs together and a coherent position emerges, and it is not the position most creation stories take.

There is water, and it is going nowhere. There is darkness on it. There is duration so long it cannot be counted. And there is rest — the state of not yet having begun.

Nothing here is made. The Ogdoad is not a set of creator gods; it is a description of the conditions before creation, itemised. What the eight constitute is a negative — everything that was true when nothing was happening. Creation then consists of making distinctions inside that: raising a mound out of the water, separating light from the dark on its surface, dividing the sky from the ground.

Which is why Nu never goes away. You cannot destroy the thing you drew distinctions in. The flood came back over the fields every year, and the theology and the agriculture agreed about what that was.

The frogs and the snakes

The Ogdoad has the most consistent visual signature of any group in the corpus, and it is not a flattering one.

The male gods are given the heads of frogs or serpents. The female gods are given the heads of frogs or cats [budge-1904]. Frogs and snakes are what you find in mud when the water goes down: the creatures of the silt, produced by wet ground apparently out of nothing. For a set of gods who are the primeval swamp, it is exactly the right choice, and it is a long way from the falcons and lions of the state gods.

The arrangement is not rigid. In the scene identifying Kekui and Kekuit with Ka and Kait, the heads swap — the god takes the frog, the goddess the serpent [budge-1904].

The argument: were they ever real?

This is the interesting part, and Egyptology has not fully settled it. Budge lays out two flatly opposed positions from the two dominant scholars of his generation.

Brugsch: the eight are primordial. They are "the oldest example of the ogdoad," genuine primeval deities, and their conception belongs to a far older stratum than anything at Heliopolis [budge-1904].

Maspero: the eight are grammar. The high priest of Hermopolis carried a title meaning that he served "him that is chief of five" — so the city's gods were originally five: Thoth plus the four gods of the cardinal points. Female counterparts were then assigned to those four, "hence the Eight gods." On this reading the four goddesses "are only gods formed according to the laws of grammar, four being masculine, and four feminine" [budge-1904].

Maspero went further and mapped the Hermopolitan eight onto the Heliopolitan nine, pair for pair: Nu and Naunet = Shu and Tefnut, Hehu and Hehut = Geb and Nut, Kekui and Kekuit = Osiris and Isis, Kerh and Kerhet = Set and Nephthys — the whole company a copy of Heliopolis with Thoth swapped in for Atum [budge-1904].

Budge sided with Brugsch, and gave two reasons. First, the correspondence "is not supported by the evidence of the texts." Second, Thoth is inseparable from the dog-headed ape, and ape-worship "takes us back to a very remote period" — so long before Heliopolis that "the two conceptions belong not only to different stages of civilization, but probably to two distinct races of men" [budge-1904].

That last clause is 1904 racial anthropology and should be discounted as such; the chronological instinct behind it may still be sound.

Budge also thought the leadership was borrowed. "There is no proof whatsoever that Thoth was the original leader of this company of gods; on the contrary, there is reason for thinking that if the Eight ever had a leader in the beginning of their existence he must have been a form of the Sun-god" [budge-1904]. The priests of Hermopolis, he argues, did what the priests of Heliopolis did — adopt an existing company and install their own god at the top.

Where this corpus stands

The evidence in these eight entries does not support a single verdict, and the useful thing is to say why.

Against Maspero: Kekuit has an independent function that no male member has. She is "the raiser up of the night," responsible for the stretch of darkness immediately after dusk, while Kekui is "the raiser up of the light," responsible for the stretch immediately before dawn [budge-1904]. A goddess invented purely to give a god a feminine ending does not get her own half of the night.

For Maspero: Kerh and Kerhet between them accept at least eight substitute names in Budge's account — Ni, Nenu, Nut and Amen for him; Ennit, Nenuit, Nut and Nit for her [budge-1904]. That is not how a deity with a cult behaves. It is how a structural slot behaves, filled by whoever a given tradition had to hand.

So the eight are probably not uniform in kind. Some look like observed phenomena that were given names; some look like positions in a scheme that needed filling. Treating them as a single class is what makes the argument unresolvable.

Budge, having set the whole thing out, concedes the point himself: "It is not easy to reconcile the various views which Egyptologists have held about the above four pairs of deities, and it certainly appears as if the ancient Egyptians themselves had no very clear ideas" [budge-1904].

One test, and the Ogdoad partly fails it

If Brugsch and Budge are right that the eight are older than Heliopolis, they should be visible in the oldest material Egypt has left — the Pyramid Texts.

Budge himself supplies the list to check against: an enumeration of "the great gods who were certainly worshipped in the early archaic period" as made known by the Pyramid Texts at Sakkara [budge-1904]. Checked against the eight, the result is clean and one-sided:

• Nu is there, at Unas 199, among Ptah, Horus, Neith,

Ra, Hathor, Kheper, Khnemu and Sebek.

• The other seven are not there at all. Naunet, Hehu, Hehut, Kekui, Kekuit,

Kerh and Kerhet are absent from that roster.

So the primeval water is demonstrably Old Kingdom. The eightfold company is not — at least not on this evidence.

That does not vindicate Maspero, whose specific argument was about grammatical counterparts and the "chief of five" title. But it does undercut the strongest version of Brugsch's and Budge's position. What the record supports is that a god of the primeval water is very old, and that organising him into a set of four pairs is a later act of systematising — which is exactly what the fourth pair's eight interchangeable names would lead you to expect.

One caveat, stated plainly: this is an argument from silence in a single nineteenth-century compilation, and Budge's roster is a selection rather than a concordance. It should be re-run against Sethe before it is treated as settled.

A naming hazard

Three different deities in Budge's two volumes are transliterated Nut:

1. Nut, the sky goddess. 2. Naunet, the female counterpart of Nu. 3. A substitute name for Kerh — a male god.

Any automatic extraction from Budge's text will conflate these, and this corpus does automatic extraction. The distinction is recorded in 05-TRANSLITERATION.md; it is flagged here because the Ogdoad is where two of the three collide.

Sources

• [budge-1904] — Budge, The Gods of the Egyptians (1904), vol. I, the section on Thoth and his paut, and the treatment of the four pairs. Public domain, archived in sources-texts/. On this material Budge is unusually good: he presents the opposing scholarly positions fairly, states his own with reasons, and concedes where the evidence runs out.
• [brugsch-religion] — Brugsch, Religion und Mythologie der alten Aegypter, pp. 129–142, cited by Budge. Public domain; not archived here.
• [maspero-myth] — Maspero, La Mythologie Égyptienne, p. 257, cited by Budge. Public domain; not archived here.
• [lanzone] — Lanzone, Dizionario di Mitologia Egizia, pll. 168 ff. and p. 685, cited by Budge. Public domain; not archived here.

Open threads

• Get the Ptolemaic Ogdoad material. The eight had a real late revival, and the temples at Dendera, Edfu and Hibis carry Ogdoad inscriptions. Those sources would test almost every open question on these eight entries, particularly whether the fourth pair ever acquired stable names.
• Find the Hermopolitan high priest's title Maspero built his argument on, and check the "chief of five" reading independently.
• Establish whether any of the eight is attested in the Pyramid Texts. Budge's Pyramid Text god-roster does not appear to include them, which — if it holds — is a serious problem for Brugsch's antiquity claim and should be stated plainly.
• Locate the Lanzone plates for the Ogdoad registers and the Ka / Kait scene.
• Check the reciprocal claim in Neith: she is said to absorb Naunet's attributes and to substitute for Kerhet, i.e. to stand at both the second and eighth positions.