There's a goddess on a Middle Empire relief with an object balanced on her head. It's lumpy and bilobed and has two horns curving off it.
Egyptologists call it a cow's uterus. They're confident about the identification and vague about the meaning: it links her to royal births, probably, in some emblematic way.1
Here's the other reading. It's a cow's uterus because that is where she lives. Not a badge announcing her portfolio. A location marker. She is in that region of the body, and the picture is telling you which region.
Once you've heard that, you can't un-hear it, and every headdress in every temple starts looking like a pin dropped on a map.
Names and images are two separate channels, and it's worth keeping them apart.
The name rule says: the senses of a divinity's name describe a structure. That's linguistic, and it runs through dictionaries.
The image rule says: what a divinity wears, holds, stands on, or is headed by, tells you where in the body she is seated. That's visual, and it runs through iconography.
They're independent, which is what makes them useful together. When both point at the same organ, you've got two channels agreeing. When only one does, you've got a weaker claim — and the corpus almost never says so.
The four-headed ram. Banebdjedet appears with one head and also with four.2 Behind the thyroid sit four parathyroid glands.3 Number matches number, in a structure obscure enough that a guess would have to be lucky in a specific way. This is the strongest instance in the corpus and the page on that deity says so.
The clasping arms. Anuket's name carries a determinative of two arms in the embracing position.4 The placenta wraps the fetus. And the same glyph doubles as a sieve — which is the other half of what a placenta does, passing nutrients one way and waste the other.5 Two functions, one glyph, and the organ genuinely has both.
The fish on the head. Hatmehit wears a fish.6 The pharynx, in cross-section, is fish-shaped — nasopharynx, oropharynx, laryngopharynx tapering back.7 Weaker: fish-shaped is doing a lot of work, and plenty of things are fish-shaped. But it's shored up by the second channel, because mehit independently means fish, and by a function claim about phonation that has nothing to do with shape.
The unguent jar. Bast's name is written with a bas jar; the jar is read as the pericardium, the fluid-filled sac around the heart.8 This is the weak one, and the reason is simple: Bast has a well-attested association with perfumes and oils. The jar is explained without any anatomy at all. Nothing in the method tells you when to stop.
The interesting thing isn't any individual reading. It's the blanket refusal.
Whenever Egyptology reaches the word symbolic, this teacher treats that as an admission of not knowing. His clearest statement of it comes at the crocodile stelae — Heru as a child standing on two crocodiles, gripping serpents and scorpions. Europeans, he says, called it symbolic. Then somebody went to Nubia and found children standing on actual crocodiles, and families keeping crocodiles in the front room until they get too big.9
"It's not an arbitrary or symbolic representation. It's the real thing."10
Whatever you make of the specific case, the underlying move is a genuine methodological commitment: treat depiction as report until you're forced to treat it as metaphor. Set the default to literal and make the metaphorical reading earn its place, rather than the reverse.
That inverts the usual scholarly setting, and it's not obviously the wrong one. A great deal of ancient depiction is report — of clothing, of tools, of animals, of procedures. The habit of reaching for "symbolic" first has a poor track record in the history of reading other cultures' pictures.
But a default is not an argument, and this one breaks in a predictable place.
Some things in Kemetic iconography really are emblems. Crowns denote regions. Sceptres denote office. The false beard is not an anatomical claim about chins. The tradition plainly had a symbolic register, and used it constantly.
So the literal-first default needs a companion rule saying when to switch, and the corpus has none. The result is that the unguent jar and the four-headed ram arrive in the same voice, at the same confidence, in the same week.
The honest position is that the image rule is a generator of hypotheses, not a settler of them. It's very good at producing candidates worth checking. It has no procedure for rejecting one.
A reading like this could be checked, and it's worth spelling out how, because the corpus never does.
You'd want a case where the iconography specifies something the tradition had no other way of knowing. The four parathyroids come closest: four is a specific number, the glands are small and easy to miss, and pre-modern anatomy did not reliably distinguish them. If the four-headed form predates any plausible knowledge of them, that's a genuinely hard fact to explain away.
But you'd need the dating, and you'd need to rule out the obvious alternative — that four heads meant four bau, the spirits of Ra, Shu, Geb and Asar, which is what the same teacher says elsewhere in the same session about the same image.11 He offers both readings minutes apart and never notices they compete.
That competition is the most useful thing on the page, because it's the one place the corpus supplies its own alternative hypothesis. Everywhere else it supplies only the reading it likes.
You're looking at a photograph of a relief and you're doing the thing where you read the caption first. "Tananet, consort of Montu, wearing the emblem of a cow's uterus."
Try it the other way. Cover the caption. Look at the object. Ask: if this were a picture of where something lives, where would it be pointing?
The answer comes fast and it's uncomfortable, because it's specific, and specific things can be wrong in a way that "emblem of fertility" never can.
You're in a museum and there's a case of small bronze figures. One holds a jar. One stands on a crocodile. One has a bird on its head.
You catch yourself sorting them: which of these is plausibly a report, and which is plausibly an emblem? And you notice you're deciding on vibes. You have no rule. Neither does the teacher. The difference is that you know it.
Somebody shows you a reading you find far-fetched. The old response was to dismiss it and move on. The new one is more work: what would have to be true for this to hold, and is any of that checkable?
Sometimes the answer is nothing, and you can say so cleanly. Sometimes the answer is a number, or a date, or a structure you could go and look at. Those are the ones worth keeping.
Take the composite form, because it's the one where the image rule and the anatomy claim have to carry each other.
Kainsu appears in a form with two hawk heads and two pairs of wings, standing on the heads of two crocodiles, holding the crook and flail.12 It's a strange object and the standard descriptions do not try very hard with it.
The teacher's reading: he is the thymus.13 The crook and flail are sovereignty. The crocodiles are entities moving through the body's fluid — things in the lymph that can hurt you — and standing on their heads is subduing them without destroying the medium they swim in.14
Now: is that iconography read as anatomy, or is it anatomy imposed on iconography? Honestly, the second. Nothing in the picture says thymus. The thymus arrives from elsewhere — from the observation that a second falcon-headed force sits beside the heart, and that the thymus is a lymphoid organ in exactly that position.15 The crocodiles are then interpreted in light of that siting.
That's a different epistemic operation from the four-headed ram, where the image supplied the number and the anatomy matched it. Here the anatomy supplies the frame and the image gets fitted to it.
Both are presented identically. Distinguishing them is the single most useful thing a reader can do with this corpus, and it is the thing the corpus is least equipped to do for itself.
The reading isn't therefore worthless. "Subdue what moves through the fluid without poisoning the fluid" is a decent description of adaptive immunity, and it is genuinely more interesting than "moon god, healer." But it should be labelled as an interpretation extended to an image, not as a reading recovered from one.
Strongest support. The four-headed ram against the four parathyroids, and the double-duty of Anuket's glyph as both embrace and sieve where the organ genuinely does both. In both, the image contributes information the anatomy didn't already contain.
The central tension. Literal-first is a defensible default and the corpus supplies no exit from it. Crowns, sceptres and the false beard are plainly emblematic; the method has nothing to say about them and does not notice the gap.
The self-supplied contradiction. Four heads = four parathyroid glands, and four heads = four bau of Ra, Shu, Geb and Asar. Same session, same image, minutes apart, both asserted, neither withdrawn.16 Preserved here rather than resolved — this is a real fork in the corpus's own reasoning and flattening it would destroy the evidence.
Open. Is there any Kemetic image the method would decline to read anatomically? Until one exists, the rule is a lens rather than a finding.
This page and The Name Is an Address are two channels of one method, and separating them exposes something the corpus obscures by running them together.
When both channels converge — Hatmehit's fish-shaped headdress and the word mehit meaning fish and a phonation claim — the case is genuinely stronger, because three independent lines are unlikely to coincide by accident. When only one channel carries the weight, it's much weaker. The corpus never grades its readings this way, so a reader has to do the grading, and doing it changes which claims survive.
Against Late-Period European Corruption — Critical Reading Method, there's a real inconsistency worth keeping visible. That method treats late-period material as suspect, conflated by foreign influence and needing correction. But the image rule treats depiction as reliable report. Those two commitments pull opposite ways on any late image, and the corpus applies whichever one supports the reading it wants. It never states a rule for choosing.
Map vs Territory — Go to the Conversations
The map-territory distinction is usually deployed as a warning: don't mistake the representation for the thing. Here it gets used in reverse, and the reversal is what makes the pairing worth having.
The image rule's bet is that these pictures were maps drawn from the territory — made by people who had the thing in front of them and were recording where it was — and that centuries of reading them as poetry has been the error. Not map mistaken for territory. Map mistaken for decoration.
That's a live possibility and the history of reading other cultures' images supports taking it seriously more often than it has been taken. Botanical accuracy in medieval manuscripts, navigational precision in Pacific stick charts, anatomical content in objects catalogued as ritual — the pattern of underestimating the informational load of non-Western depiction is well documented.
But the handshake earns its place by supplying the counter, which the corpus lacks. The map-territory frame insists you check the map against the territory, and that is exactly the step never taken here. Nobody dissects. Nobody dates. The readings are validated against other readings, and a map checked only against other maps is a map that can drift indefinitely while staying perfectly self-consistent.
That's the honest verdict on the whole method: not wrong, unchecked. And the difference matters more than either camp's confidence.
Sharpest implication. If ancient depiction is report rather than emblem, then the burden shifts and the question stops being "what did this symbolise" and becomes "what were they looking at." That's a harder question and a better one, and it can be wrong, which is its main advantage over the version it replaces.
Generative questions.