Chapters 38 to 42 are Brahma's own account of his beginning, told in the first person to his son Narada.1
He gives three of them.
Chapter 39: he came out of the lotus growing from the navel of the sleeping Vishnu.
Chapter 39, three sentences earlier: Shiva created him from the right side of his own body.
Chapter 40: he was born when a golden egg floating on the waters was struck open, one half becoming the sky and the other the earth.
Three different bodies, three different mechanisms, one speaker, four chapters.
And no version is presented as a correction of another. There is no "some say," no "in another kalpa," no reconciliation of any kind. They are simply told, one after another, to the same listener, in the same sitting.
The lotus origin is the famous one and it is also the politically loaded one.
It appears first not in Brahma's mouth but in Vishnu's, and it appears as a weapon.
In the Vidyeshvara telling of the quarrel, Brahma arrives, insults the sleeping Vishnu, and claims to be the grandfather of the universe. Vishnu's reply ends the argument:
Since the entire universe is located inside me, your thoughts are like those of a thief. You originated from a lotus in my navel. You are my son and are speaking in vain.2
In the Rudra telling, twenty-five chapters later, he says it again:
There is no doubt that you have originated from the lotus in my navel.3
This is the trump card of the whole episode, in both versions. Paternity is the currency of divine rank in this literature, and the navel claim settles the question of who outranks whom before any fighting starts.4
It is also, plainly, a Vaishnava claim. It makes the creator god a product of Vishnu's body.
And here is the sentence that should not be able to exist, from Brahma's own narration in chapter 39:
Shambhu created me from the right side of his body. O sage! Mahesha swiftly covered me in his maya. He created us in his pastimes, and I originated from the navel.5
Two origins. Three sentences apart. One speaker. One breath.
Not "some say I came from the right side, others from the navel." Both stated flatly, as fact, by the person they happened to, with a sentence about maya in between.
And the right-side version is not an aside. Shiva repeats it formally in chapter 41, in the treaty speech:
This Brahma, the grandfather of the worlds, was born from my right. I am the paramatman and Vishnu originated from my left... Hari was born from my left limb, and Vidhatri was born from my right limb.6
That is the Śaiva claim, and it does exactly the work a Śaiva claim needs to do. It makes Brahma and Vishnu siblings rather than father and son, and it makes both of them products of a third party. The navel argument stops working the moment you accept it.
The third is the oldest image of the three and the only one that belongs to neither sect.
That divine egg floated on the water for many years. At the end of one thousand years, it divided into two and Aja originated... From that egg was born the four-faced one who is known as Kah. He is the creator of all the worlds.7
Debroy's notes confirm both names are Brahma's.8
The golden womb — hiranyagarbha — is pre-sectarian material. In its received form the egg is the first thing, self-existent, floating in the waters, with nobody above it and nobody's body involved. What comes out of it is nobody's son.
Chapter 40 keeps the image and takes away its independence: the egg is made of the seed A, which came from the lingam, flung into the womb U, floating on water that came out of Vishnu two chapters earlier, and opened by a blow from Ishvara.9
So the third origin is the first origin, subordinated.
The chapter 39 passage deserves a section of its own, because a contradiction across chapters is ordinary in this literature and a contradiction across three sentences is not.
Read it in full, with the surrounding lines:
O sage! Parameshvara Shankara carefully exerted himself, as he had done earlier. Shambhu created me from the right side of his body. O sage! Mahesha swiftly covered me in his maya. He created us in his pastimes, and I originated from the navel. O son! Since I was born in this way, I came to be known as the one who was born from a golden womb.5
Three things are happening in one paragraph.
The Śaiva origin is asserted — right side of Shiva's body, with Shiva's exertion described.
The Vaishnava origin is asserted — from the navel.
And an epithet is derived from the second: born from a golden womb, which Debroy glosses as Hiranyagarbha.10
That last move is the tell. The epithet is attached to the navel origin — which means somebody has already done the work of explaining hiranyagarbha without an egg. And then one chapter later there is an actual egg, and Brahma is born out of it, and the epithet is available for the taking a second time.
The same title is derived twice from two different births, one chapter apart, by the same speaker.
That is not a text with three traditions lying inertly side by side. It is a text where a joining seam runs straight through the middle of a paragraph, and where somebody has been actively fitting explanations onto material that arrived with its own.
You are trying to establish when a company actually started, for a piece you are writing.
The website says 2011.
The founder, in an interview, says it really began in a flat in 2009, before there was a name for it.
And the incorporation documents say 2013.
None of these is a lie. The flat is where the idea was; 2011 is when the first product shipped; 2013 is when it became a legal entity. Everyone involved would defend their own date honestly.
But notice who tells which one, and when.
The 2009 story is told when the founder wants to claim priority over a rival who started in 2010. The 2011 date is on the site because it makes the company look established without looking stale. The 2013 filing is what appears in a due-diligence pack.
Nobody is coordinating this. Each date is deployed where it does the most good, and the reason nobody notices the discrepancy is that the three audiences never meet.
That is chapters 38 to 42 exactly, with one difference that makes it worse rather than better: here the three audiences do meet. It is one speech, to one listener, in one sitting.
The obvious answer is that the volume is a compilation, and the obvious answer is probably right.11
But there is a more specific reason available here, and it is the strongest thing on this page.
The Śaiva text cannot delete the Vaishnava origin, because its own founding myth requires it.
The column of fire is where lingam worship begins. Everything the volume builds — the installation rules, the arithmetic, the materials, the whole apparatus — rests on that episode.12 And the episode is a quarrel about rank.
A quarrel needs a claim. Vishnu's claim is the navel. Take away the navel and Vishnu has nothing to say when Brahma calls him a child, the argument does not escalate, the weapons are not released, and the column has no reason to appear.
So the Śaiva compiler is in an awkward position. He needs Vishnu to make a paternity claim that the text elsewhere flatly denies — because the claim is what generates the founding event.
His solution is not to remove it. It is to let it stand and surround it. Vishnu says you came from my navel, and Brahma says it too, and Shiva says actually you came from my right side, and none of them argues with the others.
A sectarian text turns out to be load-bearing on its rival's claim. That is a much sharper finding than "the Puranas are inconsistent," and it is checkable: wherever a tradition preserves a rival's doctrine unusually carefully, look for a story it cannot tell without it.
Laid out plainly, because the three are not interchangeable and each has a job.
| Origin | Makes Brahma | Serves | Where it is used |
|---|---|---|---|
| Navel of Vishnu | Vishnu's son | Vaishnava | the quarrel, in both tellings — the claim that ends the argument |
| Right side of Shiva | Shiva's product, Vishnu's sibling | Śaiva | the treaty speech, where the pantheon is arranged |
| Golden egg | nobody's son, then everybody's | neither, then Śaiva | the cosmogony proper, where the world is made |
Each does something the others cannot.
The navel makes a fight possible. The right side makes a hierarchy possible. The egg makes a world possible — because neither of the body-origins explains where sky and earth come from, and the egg does.
And that is worth stating as the general point, because it applies well beyond this volume: these three are not three answers to one question. They are answers to three different questions that happen to share a subject. Nobody reconciled them because, from inside, there was nothing to reconcile — each was correct for the purpose it was doing.
The contradiction is only visible to a reader who asks the one question none of the passages was written to answer: where did Brahma actually come from?
Strongest support is the chapter 39 paragraph, where the right-side origin and the navel origin appear three sentences apart in one first-person speech, with an epithet derived from the second.
Tension one. Two origins in one paragraph, a third one chapter later, no acknowledgement anywhere in the volume.
Tension two. Hiranyagarbha is derived twice — once from the manner of the lotus birth, once from an actual golden egg — by the same speaker, one chapter apart.
Tension three, and it is the structural one. The navel claim is denied by the volume's own theology and required by the volume's own founding narrative. It cannot be removed without destroying the column-of-fire episode.
Tension four. In the Vidyeshvara telling, Vishnu's navel claim wins the exchange and Brahma is punished for what follows. In the Rudra telling, the same claim is made and Shiva contradicts it two chapters later without comment.13
Open question, and it is decidable. In Volumes 2 and 3, is Brahma given a fourth origin — and if so, does it belong to a fourth constituency? A tradition that adds origins as it absorbs communities would show that. A tradition that has settled on one would show the opposite, and would date the settlement.
Three Incompatible Origin Accounts holds the same failure mode applied to the text rather than to the god: Brahma composed the Puranas and taught Sanatkumara / Vedavyasa composed them / Shiva composed this one and Vyasa merely abridged it. Three chains, never reconciled.
This page is that page's second instance, and the pair is stronger than either alone. The volume does the same thing to its own provenance and to the provenance of its creator god — which makes multiplicity-without-reconciliation a property of the text's construction rather than a local accident in one passage.
The Quarrel Over Who Is Whose Father documented the paternity claim as the currency of divine rank. It could not see that the currency is counterfeit by the volume's own standard — Shiva says in chapter 41 that Brahma came from his right side, which means Vishnu's winning argument was false when he made it, in a passage the same volume presents as decisive.
The Second Column of Fire is where this bites hardest. That page found the lie missing from the second telling of the founding myth. This one finds that the claim which starts the whole quarrel is contradicted by the frame it is told inside. Both tellings contain a falsehood; only one of them notices, and it notices the wrong one.
Apocryphal Birth-Prophecy Architecture (history). The Ashokavadana, composed in the second century CE — four hundred years after the Buddha and two after Ashoka — opens with a boy dropping dirt into the Buddha's begging bowl and the Buddha turning to Ananda to prophesy that this child will be reborn as a king named Ashoka who will distribute the relics and build the eighty-four thousand stupas. The conversion narrative that follows has been pre-authorised by the founder himself, several centuries after both of them died.
That is the mechanism running here, three times over.
Each of Brahma's origins is a beginning written backwards from a later need. The navel exists so that a quarrel can happen. The right side exists so that a pantheon can be ranked. The egg exists so that a world can have a sky and a ground.
The insight neither gives alone: the tell for a retrofitted origin is not implausibility, it is fit. Nothing about the dirt-offering story is internally odd; what marks it is how precisely it authorises the specific thing the text goes on to want. And the multiplicity here is the same signal at higher resolution — three origins, each an exact fit for a different downstream requirement, which is far better evidence of manufacture than one origin could ever be. A single origin story can be either true or constructed. Three, each perfectly suited to a different argument, can only be constructed — and that is a general test available wherever a tradition has been generous enough to keep all its drafts.
Beginning and Ending Written Last (creative-practice). Karlsson's finding is that an opening written first almost always has to be thrown away, because it was built as a setup for a piece that did not exist yet — and that once the piece is written, the real opening is usually already sitting somewhere in the middle, obvious in retrospect.
Cosmogonies are openings.
And this one behaves exactly as the craft rule predicts: it was written last, after the tradition it introduces already existed, which is why it is so well fitted to what follows and why there are three of them. Each was drafted to set up a different later passage, and nobody threw the earlier drafts away.
What the pair produces: the reason origin stories are the most contested part of any tradition is not that beginnings are metaphysically important. It is that beginnings are the last thing written and therefore the most current — they encode the settlement in force at the moment of composition rather than anything older. That inverts the usual assumption. A tradition's cosmogony is not its most ancient layer; it is its most recent, and often its most political, because it is the part rewritten every time the argument it introduces changes. Which suggests a practical instruction for reading any corpus: date the origin story last, not first, and read it as evidence about the period that produced it rather than the period it describes.
Sharpest implication. Brahma narrates his own beginning to his son and gives three of them: from the lotus in Vishnu's navel, from the right side of Shiva's body, and from a golden egg struck open on the waters. Two of those are three sentences apart in one paragraph, with an epithet — born from a golden womb — derived from the second, one chapter before an actual golden egg turns up and supplies the same title again. And each origin is doing a job the others cannot: the navel makes a fight possible, the right side makes a hierarchy possible, the egg makes a world possible. Which is why nothing was reconciled — from inside, there was nothing to reconcile. But the sharpest consequence is what the Śaiva compiler could not do: he could not delete the Vaishnava claim, because his own founding myth needs it. The column of fire is a quarrel about rank, a quarrel needs a claim, and Vishnu's only claim is the navel. Remove it and lingam worship has no origin story. The sectarian text is load-bearing on its rival's doctrine, and so it lets the claim stand, surrounds it, and contradicts it two chapters later without ever saying so.
Generative questions.