Chapter 46 is a catalogue of flowers. Which one to offer for which outcome, how many, and what each weighs.
It is generous about it. Lotuses, bilva leaves, dhattura, tulasi, mallika, karavira, shami, agastya — dozens of species, each with its result. And then a summary line that could not be more open:
There is no flower that is not loved by Shiva. With the exception of champaka and ketaka, everything can be offered.1
Every flower in the world is acceptable. Two are not.
One of those two is the ketaki.
The ketaki is not a random exclusion. The volume told us exactly how it got banned, thirty-one chapters earlier.
In the Vidyeshvara Samhita, Brahma is flying up the column of fire as a swan, failing to find the top, when he meets a ketaki flower falling from above. It has been falling for years. It was dislodged from Shiva's own head when he laughed and shook it, watching the two of them struggle.2
Brahma suborns it. He needs a witness to a lie he is about to tell — that he reached the summit — and the flower agrees. He prostrates himself before it, quoting scripture as he goes: "the instruction of the sacred texts is that in a time of calamity, the utterance of a falsehood is praised."
The lie works. Vishnu believes it and worships him with sixteen upacharas.
Then Shiva appears, and the sentence is passed:
O ketaki! You are deceitful. Go far away. From now on, I will no longer regard you as a beloved flower with which I can be worshipped.3
The flower appeals. Shiva partly relents — his followers may wear it, and it may hang in the canopies above him — but the core of it stands: "since I speak the truth, it is not appropriate that I should wear you."
That is the origin of chapter 46's exclusion, and chapter 46 does not mention it.
Now the problem.
Chapters 39 and 40 retell the entire column-of-fire episode. Same sleeping Vishnu, same child insult, same navel-paternity claim, same quarrel, same pillar, same boar downward and swan upward, same double failure.
And there is no ketaki. There is no lie. There is no false witness, no Bhairava, no sentence, no stripping of worship. Both gods simply fail, come back, and bow for a hundred autumns.4
That retelling is narrated by Brahma himself, in the first person, to his son.
So within the Rudra Samhita we have a founding myth with the crime removed — and, six chapters later, the punishment that the crime produced, still in force.
It is worth being exact about how close these two passages are, because the closeness is the argument.
Chapters 39 and 40 are the cosmogony. Chapter 46 is the flower catalogue. Both are in the Rudra Samhita. Both are ultimately Brahma speaking to Narada, reported by Suta.5 Six chapters and perhaps forty minutes of recitation separate them.
One says the episode had no crime in it.
The other enforces the penalty for that crime, by name, as a live rule about what a householder may put on a lingam this morning.
The Rudra Samhita does not know about the ketaki lie in chapter 39 and does know about it in chapter 46.
This settles something the vault could not settle before, and it is worth walking through carefully because the conclusion is narrow but firm.
When the second telling was built, three explanations were on the table for the missing lie.4
One: a different recension. The Rudra material came from a source that simply did not have the ketaki episode, and the two were bound together without anyone reading across.
Two: a different purpose. The Vidyeshvara version is arguing about worship — why Brahma has no temples — so it needs a crime. The Rudra version is arguing about cosmogony, and a cosmogony has no use for a courtroom.
Three: deliberate omission by the narrator, who is the accused.
Chapter 46 does not eliminate all three, but it eliminates the strongest form of the first.
The Rudra Samhita's own material contains the ketaki ban. So whoever assembled this Samhita had the episode, or at least its outcome, available. The tradition it drew on knew that the ketaki was excluded and knew it firmly enough to encode it in a worship manual as a standing rule.
Which means the absence in chapter 39 is not ignorance. It is either a different source within a compilation that also held the ban, or a choice.
And there is a further squeeze. Explanation two — that a cosmogony has no use for a courtroom — is a genuine and good argument, and it survives. But it now has to explain why the same Samhita, forty minutes later, is happy to enforce the courtroom's verdict without ever mentioning the trial.
A tradition that keeps a sentence and drops the case is doing something more specific than changing register.
You join a company and on the second day someone tells you never to book the small meeting room on the third floor.
You ask why. Nobody knows. It is not in any policy. The room is fine — it has a table, a screen, a window. And nobody uses it.
Six months later you find out. There was an incident, four years ago, involving two people who no longer work there, and one of them was badly treated, and the room is where it happened. The story was never written down anywhere and the people who witnessed it have mostly left.
But the rule survived them all.
That is the shape here, and the direction of inference is what matters. You did not learn about the incident and then observe the rule. You observed the rule first — an arbitrary, unexplained, faithfully-kept prohibition — and it told you that something had happened before you arrived.
Rules like that are the most reliable historical evidence a live institution produces, and for a specific reason: nobody maintains an arbitrary prohibition on purpose. Stories get retold and shaped and softened, and the shaping is the point of retelling them. A rule just gets obeyed. There is no incentive to improve it, so there is no mechanism by which it drifts.
Chapter 46's ketaki ban is that rule. And its presence tells you the episode was there, whatever chapter 39 says.
Three consequences for the build, stated precisely, with the limits marked.
One. The omission in chapters 39–40 is not the Rudra Samhita being unaware of the ketaki episode. Its own worship material enforces the outcome. Whatever produced the gap, it was not that the tradition had never heard of it.
Two. The strongest version of the "different recension" reading is weakened, not killed. A compilation can absolutely contain one block that lacks the episode and another that assumes it — that is what compilations do. But the two blocks are six chapters apart in one Samhita, and the burden has shifted: the compiler had both and joined them.
Three, and this is the one that matters most. The Second Column of Fire catalogued what the retelling drops — ketaki, lie, witness, Bhairava, sentence, staff, guru appointment. It could not say whether any of that was known to the Samhita doing the dropping. Now one item on that list is demonstrably known to it.
That does not prove the others were. Bhairava's creation, the negotiated sentence, the guru appointment — none of those leaves a trace in chapter 46 the way the flower does.
But the ketaki was the load-bearing one, because the ketaki is the witness, and without a witness the lie does not work. If the flower's exclusion was known, the reason for the flower's exclusion was available.
The general instrument, because it is transferable and it is the reason this page is worth its length.
Devotional literature is full of arbitrary prohibitions — this flower, not that one; this day, not that one; this material, this direction, this hand. They are usually treated as texture, the kind of thing you skim on the way to the doctrine.
They are the opposite of texture. A prohibition with no stated reason is a fossil, and it can be dated relative to the narratives around it.
The method is simple and it runs in one direction:
That works because rules and stories decay differently. A story is retold, and every retelling is an opportunity to change it — which is exactly what chapters 39 and 40 did. A rule is applied, and application is not an opportunity to change anything.
So in a corpus with multiple recensions, the ritual layer is more conservative than the narrative layer, and where the two disagree, the ritual layer is the better witness to what the tradition once held.
This build has been using narrative doublings and frame violations to detect composition. Chapter 46 adds a third instrument, and it is the only one that can recover something a text has removed rather than merely duplicated.
Strongest support is the exclusion itself: "there is no flower that is not loved by Shiva. With the exception of champaka and ketaka, everything can be offered" — a maximally generous rule with exactly one relevant exception, in a Samhita whose own cosmogony contains no reason for it.
Tension one. Chapter 46 enforces the sentence and never states the offence. The volume elsewhere explains its prohibitions freely, at length, and often more than once.
Tension two. Chapter 15's sentence was partial — Shiva will not wear the ketaki, but his followers may, and it may hang in canopies above him.3 Chapter 46's exclusion is flat, with no such relaxation. The penalty has hardened in transmission, which is the same directional drift the numbers show.6
Tension three, and it is unresolved. Champaka is banned alongside ketaka, and Volume 1 gives no origin for that at all. Either there is a second lost episode, or the pairing is conventional and the ketaki has been attached to an existing exclusion. Both are live and the text will not decide.
Tension four. This weakens the recension explanation without eliminating it. A compilation may hold a block that lacks an episode and another that assumes it — but the two are six chapters apart in one Samhita, and the compiler had both.
Open question, and it is the highest-value item outstanding. Do Volumes 2 or 3 narrate the ketaki episode, and if so, with or without the lie? A third telling would settle whether the Rudra omission is a recension or a choice — and this page makes that check specific rather than general.
The Second Column of Fire is the page this one exists to correct, and the correction is a strengthening rather than a retraction. That page found the crime missing from the second telling and honestly listed three possible explanations, noting that an argument from silence needs to establish what a document was for before weighing what it omits.
This page supplies the missing piece: the Samhita that omits the crime enforces its penalty. The omission is not ignorance.
Brahma's Lie and the Ketaki Flower holds the original episode, and its status changes here too. That page carries the boundary that its material exists in one Samhita only — which remains true of the narrative. But the flower's exclusion is in both, so the episode's effect is corpus-wide even where its telling is not.
The Sentence That Gets Negotiated Down documented the volume's habit of passing a harsh sentence and then softening it on appeal — Brahma stripped of worship and handed the staff of chastisement and mandatory guru status; the ketaki banished and partly reinstated.
Chapter 46 shows what survives that process at distance. The negotiation is gone. The flower's partial reinstatement — followers may wear it, canopies may hold it — has not travelled. Only the ban did.
Which is a finding in its own right about how this literature transmits: the concession is the first thing lost.
Behavioral Residue (behavioral-mechanics). The mechanism there is the visible trace consumption leaves behind — the Sent from my iPhone, the recognisable bottle on the table, the voting sticker — signalling identity to others and, crucially, back to the person themselves.
The ketaki ban is residue of a different kind, and setting the two side by side sharpens both.
Behavioural residue is a trace left on purpose, or at least left by an act whose visibility was part of its value. This is a trace nobody intended to leave. It survives because it is arbitrary — the rule has no content anyone is invested in, so there is no reason to revise it, and it outlasts the story that produced it by centuries.
The insight neither gives alone: the traces that survive longest are the ones nobody is curating. Deliberate residue is maintained, which means it is also edited — a signal kept current with what its owner wants to signal now. An unexplained prohibition is maintained by nobody and edited by nobody, so it preserves an older state than anything the institution says about itself. Which suggests a general search instruction for any tradition, corporate or religious: the honest history is in the rules that embarrass no one because nobody remembers what they are for.
Ashoka's Deliberate Genealogical Silence (history). Across the entire Ashokan corpus — Girnar, Maski, Brahmagiri, every edict — the king's father is never named. A silence so complete across so many independent inscriptions that it cannot be accident.
Both cases are absences in a record, and the difference between them is exactly the methodological point.
Ashoka's silence is legible because it is total. Every inscription, every site, no exceptions — which is what lets it be read as deliberate rather than as a gap. If one edict had named Bindusara, the argument would collapse.
The ketaki works the opposite way. The silence in chapters 39–40 is legible precisely because it is not total — because six chapters later the same Samhita acts on what it did not say. A complete silence is evidence of intent; an incomplete one is evidence of what was suppressed.
What the pair produces: these are the two ways an absence can be made to speak, and they need opposite conditions. A gap with no leak tells you someone decided; a gap with a leak tells you what was in it. Which gives a practical instruction for reading any corpus with omissions: do not look for more silence — look for where the omitted thing left a footprint in a layer nobody thought to edit. For Ashoka that would be an administrative or ritual document rather than a proclamation. Here it was a list of flowers.
Sharpest implication. Chapter 46 says "there is no flower that is not loved by Śiva" — and then excludes exactly two, one of which is the ketaka. That exclusion exists because of the ketakī flower's false witness to Brahmā's lie at the column of fire, an episode the volume narrates in chapter 15. But the Rudra Saṃhitā's own retelling of that episode, in chapters 39–40, contains no lie, no flower and no sentence at all. So one Saṃhitā, six chapters apart, drops the crime and keeps the penalty — which means the omission is not ignorance: the tradition that assembled this material had the episode, or at least its verdict, firmly enough to encode it as a standing rule about what a householder may put on a lingam. And what survived transmission is instructive: chapter 15's sentence was partial, with the flower allowed to Śiva's followers and in canopies above him, and none of that concession travelled. Only the ban did. This gives the build a third instrument alongside narrative doubling and frame violation, and the only one that can recover something a text removed — because a story is changed every time it is retold and a rule is only ever obeyed.
Generative questions.