Here's a word: si.
In Akan it means to stand. It also means to sit. Which is already odd, because those are opposites.
It means to set something down and to set something in motion.
It means to descend — to come down onto something.
It means to establish, to institute, to found.
It means to purify, to cleanse.
It means the lower part of the back.
And it means what happens when a spirit takes possession of a person: it comes down and becomes seated in them.1
Nine senses. One syllable.
That looks like a mess. It turns out to be the most tightly organised thing in this corpus.
Look again and the contradiction dissolves.
Standing and sitting aren't opposites here. Both are being fixed in a position. You stand fixed; you sit fixed. The English word seat does the same work — a seat is where you sit, and to seat someone is to install them.
Setting down and setting in motion are the same too. You set a date. You set a clock going. You set a stone in a wall. In every case you're putting something into a determined state.
Descending is how a thing arrives at that state. It comes down and lands.
Establishing is descending made permanent.
The lower back is what lets you be fixed upright at all — the pelvis, the sacrum, the thing you sit on.2
And possession is a spirit descending and becoming seated in a body.3
So it isn't nine meanings. It's one motion — coming down into a fixed position — applied to a body, a stone, a king, a spirit, a civilisation.
Most of the comparative material here matches one word to one word. This is different in kind.
A single root generates:
That's not a coincidence being pointed at. It's a claim that one semantic core organises a deity, an organ, a piece of furniture, a political concept and a rite.
Whether or not the Akan-Kemetic link holds, the internal coherence is real and it's checkable in the Akan dictionary alone.
Here's the part that's genuinely elegant, and the teacher walks it deliberately.
Take Asar's myth and read it through the senses of si.
He is appointed to rule — si, to institute, to be seated as king.
He establishes civilisation — si, to found, to set in operation.
He is killed — and awu is death, giving awu-si, and the compound names sudden death.6
He descends into the ancestral realm — si, to go down.
He is purified by funerary rite — si, to cleanse.7
He becomes seated as sovereign of the spirit world — si again.
He is raised up — si, to rise.
"We see the entire story play out through the definition stated above."8
That's the claim, and it's a strong one: the myth is the word conjugated.
Two readings, and they lead somewhere different.
Reading one — the myth was built from the word. The narrative is an elaboration of a root's semantic range. Storytellers took a word meaning establish-descend-purify-rise and wrote a life that performs all of it.
Reading two — the word was shaped by the myth. The root accumulated its senses because the god who bore the name did all those things, and the language absorbed it.
Both are plausible. Neither is testable from a transcript.
But there's a third possibility the corpus doesn't consider, and it's the deflating one: that any sufficiently polysemous root can be mapped onto any sufficiently eventful narrative, because both have enough parts.
English set has over four hundred distinct senses. Give me a myth and I'll walk it through set — the king is set on the throne, the sun sets, the trap is set, the bone is set, the jelly sets. It would feel meaningful and mean nothing.
The corpus never tests against that. It should.
That said, the si case isn't quite as free as the set comparison suggests, and it's worth being precise about why.
The senses aren't scattered. They cluster tightly around one physical image — coming down into a fixed position. Even the anatomical sense, the lower back, is the same image made flesh.
English set is scattered. Its senses have drifted apart over centuries and no single image holds them.
So the difference is coherence, not count. A root with nine tightly-related senses is a different object from a root with four hundred loose ones, and the myth-fits-the-word observation is worth more against the first than the second.
That's a real distinction and the corpus doesn't make it. It should — it's the corpus's own best defence.
You're watching him read the dictionary entry and he doesn't stop at the useful senses.
Si, to stand. Si, to sit. Si, to place. Si, to rise. Si, to descend. Si, to cleanse. Si, the lower part of the back, from one thigh bone to the other.
He reads the anatomical one in the same tone as the others. He doesn't flag it as the payoff, though it plainly is.
You try to hold all nine senses at once and find you can't — until you stop treating them as a list and picture the motion instead.
Something comes down. It lands. It stays.
Now all nine are available at once, and you're not memorising anything.
Later you're reading about a completely different tradition and you catch yourself asking a new question.
Not what does this word mean but what does this word do — what single motion would generate all the senses it carries.
Sometimes there is one. Often there isn't, and the senses really are unrelated. Learning to tell the difference is the transferable skill.
Take one branch of the cluster all the way down, because it's where the abstract and the physical meet.
The reduplicated form. Si doubled gives sisi — glossed as arrangement, institution, and by the teacher as to lead the way in instituting.9
The furniture. The Akan royal seat is the agua or adwa, and the teacher's point is that a throne is not a chair with decoration. It's the si — the fixed seat from which order is established.10
The anatomy. Your own si is the lower back and pelvis: what fixes you upright, what you sit on, the base the spine descends into.11
And the body's own throne. The sacrum — the fused vertebrae at the base of the spine — is the seat of the trunk.12
Now the move. If the throne is the seat of the nation, and the sacrum is the seat of the body, then they are the same structure at two scales, and the word says so.
That's the corpus's characteristic operation in miniature: a political object, an anatomical object and a divine name collapsing into one root.
What holds up. The Akan senses are dictionary-verifiable. The sacrum genuinely is the structural base of the trunk. The throne genuinely is called by a word in this family.
What doesn't follow. That the ancients intended the anatomical parallel. The word covers both because both are seats — which is what the root means. Finding that a root meaning seat applies to two seats is not a discovery about anatomy. It's the root working normally.
That's the trap in the whole method, visible clearly here: a polysemous root will always connect its own referents, and connecting them can feel like uncovering a hidden design when it's just the word doing its job.
Strongest support. The internal coherence. Nine senses organised around one physical image is unusual and checkable in the Akan dictionary without any Kemetic claim at all.
The central tension. The myth-fits-the-word observation is never tested against a control. A polysemous root can be mapped onto almost any narrative, and the corpus offers no case where the mapping failed.
A second. The corpus's own best defence — that these senses are tightly clustered rather than scattered — is never articulated, so the reading looks looser than it is.
Open. Run the control. Take a comparably polysemous Akan root with no divine name attached and try to map a myth onto it. If it works as well, the si reading is an artefact of method. If it doesn't, the reading gains real support. Cheap to run, and nobody has.
Against Awusi / Asi — Establishment and Institution Etymology, this page is the fuller version of a reading that page holds in part.
That page carries the establish/institute strand. This one adds the rest of the cluster — descend, purify, possess, the lower back — and argues they're one motion rather than a list.
The difference matters for what the etymology can support. A root meaning establish explains a founder-king. A root meaning come down and be fixed explains a founder-king and a possession rite and a sacrum, which is what the corpus actually needs it to do.
Against Four Words Offered as Proof, there's a useful contrast in evidential type. That set matches Akan words to Kemetic words — cross-linguistic, contestable, family-boundary problems. This page's core observation is entirely internal to Akan and needs no cross-linguistic claim to stand.
Which makes it, quietly, some of the most defensible material in the corpus. It just isn't presented that way.
Motif as Recurring Symbol Anchor
A motif and a polysemous root do the same structural work at different levels of a text.
A motif returns — water, a door, a bird — and each return carries the accumulated weight of the previous ones. The reader builds a private meaning nobody stated. Its power is that it operates below the level of argument.
The si cluster does that inside a single syllable. Every time the word is used, in whatever sense, it drags the others behind it. Say si meaning purify and descend and establish are faintly present, because they live in the same sound.
The insight the pairing produces is about what happens when the motif is a word rather than an image.
A motif can be avoided. A writer can decline to use water in a scene. A root cannot be avoided if you're speaking the language — every mention of the lower back, of sitting down, of founding a village, is a mention of the deity's name-root.
Which means the theology isn't transmitted by teaching. It's transmitted by speaking normally. A child learns the connection between establishing, descending and being seated before anyone tells them a god's name, because they learn a word.
That's a genuinely different transmission mechanism from doctrine, and it explains something the corpus asserts but doesn't account for: how a cosmology survives the loss of its institutions. It survives in the vocabulary of people who no longer know they're carrying it.
Sharpest implication. If a divinity's name-root is an ordinary word people use daily, then the cosmology is embedded in the language rather than in the religion — and killing the religion doesn't kill it. That's the strongest available argument for the corpus's whole revival project, and the corpus never makes it in these terms.
Generative questions.