Imagine a company with a list of problems and a chart showing who can handle each one. Most are covered. A few have a blank next to them, because the only person who could deal with those doesn't exist yet.
That is the argument Brahmā uses to get Śiva married, and it is startlingly literal.
"O Maheśvara! Some asuras must be killed by me, some must be killed by Hari and there are others who have to be killed by you. O great lord! Some will be killed by the son who results from your seed. There will be asuras who can only be killed through māyā."1
Read that as a table and it is a capability matrix. Every category of enemy has an assigned killer. Four of the five slots are staffed. The fifth is staffed by a person who has not been conceived, whose mother has not been chosen, and whose father is not interested.
The marriage is the procurement action for that slot.
And Brahmā names a reason Śiva cannot cover the gap himself, which is the part worth slowing down for:
"You are always engaged in yoga, devoid of attachment and hate. Thus, since they are the subjects of your compassion, some can never be killed by you."1
Not will not. Cannot. The disqualification is his virtue. He is too unattached to have enemies, which means the beings that need killing are, from where he stands, simply beings.
By chapter 68 this marriage has been argued for three separate times, to three different audiences, on three unrelated grounds. Nobody puts the three side by side.
To Nārada, in private, decades later: Brahmā was caught looking at his own daughter with desire, Śiva laughed at him in front of his sons, and he wanted it to stop — "when he accepts a wife, I will be happy and will no longer be miserable."2
To Devī, in chapter 63: "If Hara does not accept a wife, how will this auspicious creation continue?" — a cosmological necessity, framed entirely in the vocabulary Viṣṇu had used one chapter earlier to rebuke him.3
To Śiva, here: a specific military gap in a specific list of enemies.
Three arguments, and only the first one is true in the sense of describing why anyone is actually doing this. The other two are real problems — the text does not invent the asuras — but they are reasons found for a decision already taken, and the sequence proves it, because the grievance predates them by seven chapters.
Notice also that each version is pitched to what the listener cares about. Devī gets creation. Śiva gets a threat assessment. Nobody gets the grievance except Nārada, forty chapters after it stopped mattering.
Brahmā's speech is worth reading as a piece of persuasion, because it is unusually well built.
He opens by conceding Śiva's entire position: "You are always engaged in yoga, devoid of attachment and hate." The compliment is the argument — your detachment is exactly why we have a problem.
He then makes the problem structural rather than urgent: *"Until they are restrained, how can the equilibrium of creation, preservation and destruction be constantly maintained? It will always be unstable." Not a crisis, a permanent instability, which is harder to defer.
Then he gives the theological frame that makes the request modest:
"We are one in form. The differences only because there are differences in the tasks. Had a difference in tasks not existed, there would have been no need for differences in forms."1
That is a purely functionalist account of the three gods: form follows job. Which quietly reframes marriage as a job requirement rather than a change of state. If you are the destroyer by function, and the function now includes a fifth category of enemy, then acquiring a wife is no more personal than acquiring a weapon.
And he closes on symmetry: "The one who resides on the lotus is Viṣṇu's companion and Sāvitrī is mine. O Śambhu! In that way, accept a beloved now." Everyone else on the management team is married.
Then Brahmā plays the strongest card, and it is not an argument at all. It is a quotation.
*"In ancient times, in your form as Śiva, you yourself told us this. 'In the form of Rudra, I will assume guṇas and will carry out the task of dissolution. I will carry out the excellent task carried out in the worlds of taking a woman as a wife.' I have thus remembered what you yourself said. Now please fulfil your pledge."*1
Śiva already promised. In an earlier form, in an earlier age, to this same audience.
So the request is not please do this new thing. It is you are in breach. And Śiva accepts the framing immediately — his consent, when it comes, cites it: *"In addition, I must fulfil my pledge."1
This is the second pledge in the same chapter, and the pair is the chapter's real architecture. Śiva is held to an old promise, and the price of honouring it is a new promise — the abandonment clause — which he will honour just as literally, at Satī's expense.4 The volume's supreme god is characterised, in the space of one conversation, as a being who keeps his word exactly and never revisits it.
Put the three arguments in order of appearance and something specific shows up.
Chapter 53–60: the grievance. Concrete, personal, sufficient to explain everything, and told to nobody who could act on it.
Chapter 63: cosmic continuity. Told to Devī, whose consent is required, and pitched at the level of the universe.
Chapter 68: the asura gap. Told to Śiva, whose consent is required, and pitched at the level of operations.
The grievance has one property the other two lack: it is the only one that explains the timing. Creation had been continuing fine. The asuras were not new. Nothing in chapters 62–68 has changed about the universe. What changed is that a god was laughed at.
And here is the thing that makes it worth a page rather than a note. Both later arguments are probably true. The asuras really do need killing; Kārttikeya really will kill Tāraka. The volume is not showing us a fabrication — it is showing us a real need that gets discovered because someone was looking for a reason.
That is a much more common shape than lying, and much harder to see: the justification is sound, the decision preceded it, and no dishonesty was required at any point.
You will recognise this from any organisation that has ever restructured.
Somebody wants something for a reason they will not say. They do not fabricate a case. They go and find one, and because the world is complicated, they find a good one. The case is checkable, the numbers hold, and everyone who examines it agrees it is sound — because it is.
You cannot beat that on the merits. Auditing the argument returns the argument is fine, which is true and useless.
The tell is timing, not content. Ask when the problem was discovered relative to when the preference appeared. A genuine problem gets noticed roughly when it becomes a problem. A recruited problem gets noticed the week after someone needed one — and it will usually be a problem that has been quietly sitting there for years.
Second tell: count the audiences. A real case gets told the same way to everyone, because it is the reason. A recruited case gets tailored — creation for the goddess, threat assessment for the destroyer — and the different versions never meet, because the person telling them is the only one in every room.5
The practical move is not confrontation. It is a single flat question: "what changed?" Not is this a problem — it is — but what happened recently that made it one. Brahmā has no answer to that question, and the volume knows it, which is why he never has to face it.
Four beings in chapter 68 know the real reason and none of them says it.
Brahmā obviously will not. Viṣṇu told him to his face six chapters ago that the whole thing was hatred — "taking Śaṅkara to be an ordinary god, you hate him" — and here he speaks second, endorsing the procurement case in full: "what Brahmā has said represents everything that I have to say."6
That is the same reversal vishnu-tells-brahma-he-hates-shiva documents at the level of one speech, now visible at the level of the plot. Viṣṇu did not change his mind. He simply arrived at a meeting where the item was on the agenda and the correction was not.
Śiva does not ask. He raises no question about why this has come up now, who wants it, or what happened. He negotiates terms.
And the fourth is Devī, who has already agreed to be born for exactly this purpose, having named it a delusion while agreeing.7 She is not present, and her prior objection is not represented by anyone in the room.
Four accurate readings of the situation, none of them in the conversation where the decision is made. That is the Khaṇḍa's standing arrangement, and this chapter is its cleanest instance.
The last thing, and it is easy to read past.
The son is specified before the wife is. Brahmā's case names a capability — "the son who results from your seed" — as an existing item on a list of enemies and their killers. Only after Śiva agrees does the conversation turn to who the mother will be, and Brahmā produces Satī the way one produces a candidate: she is available, she is suitable, she is already performing austerities, and she wants the job.8
Which orders the whole sequence strangely. The requirement is the child. The wife is the means. And Satī — who has spent twelve months of tithi-by-tithi vows to be exactly here9 — is introduced into the negotiation as the solution to a staffing problem she has never been told about.
She does not get the son either. Satī dies childless at her father's sacrifice; the asura-killer arrives from the second marriage, generations of narrative later.10 The reason the marriage was authorised is not satisfied by the marriage it authorised.
Strongest support is the capability list itself — five categories of asura, five assigned killers, one of whom does not exist — together with the disqualification clause ("since they are the subjects of your compassion, some can never be killed by you") and Brahmā's citation of Śiva's own prior pledge.
Tension one, and it is the page. This is the third distinct justification given for one marriage, to a third audience, and only the first — Brahmā's humiliation — explains the timing. The other two are sound, which is what makes them hard to see.
Tension two. Śiva's disqualification is his virtue. Detachment is the property the whole volume is built to praise, and here it is the reason he cannot do the job, so a person has to be manufactured who lacks it.
Tension three. Viṣṇu endorses the plan in full six chapters after naming it as hatred, and nobody notes the reversal.6
Tension four. The son the marriage was authorised to produce is not produced by it. Satī dies childless.10
Open question. Does the Pārvatī Khaṇḍa ever cite chapter 68? If the second marriage is argued on the same asura-list — and Tāraka is precisely the enemy in play there — then the volume ran the same procurement case twice and only the second one worked, which would make the Satī Khaṇḍa a failed acquisition from the standpoint of its own stated purpose. Chapters 96–150 are unread and this is checkable in one pass.
So As to Confound Hara holds the real reason, and this page holds the reason that was given. The pair is the volume's clearest demonstration that a decision and its case can be entirely separate objects, with no lying anywhere between them.
If She Does Not Trust Me, I Will Abandon Her is the same conversation, four paragraphs later. Śiva is held to an old pledge and pays with a new one — and the new one is collected from Satī rather than from him.
Śiva Refuses to Create has him declining to make beings who will suffer, and handing the task on with an indemnity. Here he declines the marriage and is talked round. The refusal is his standard first move, and it has never once held — which suggests the refusals are doing something other than refusing.
Viṣṇu Created to Bear the Burden is the volume's other case of manufacturing a being to fill a functional gap, and it is the closer parallel than any of the marriage material: something is needed, so a person is specified, named and tasked. The difference is that Viṣṇu was made from Śiva directly and this one requires a wife, a father-in-law, and forty chapters of politics.
The Founder Dilemma and Succession Paranoia (history). The classical problem: transfer power to a successor without creating a rival who challenges you, or a successor too weak to hold what you built. Every solution trades one risk against the other, and founders reliably choose badly because the fear is about themselves.
Chapter 68 is the succession problem with the paranoia surgically removed, and the removal is what makes it legible.
Śiva has no stake. He is not dying, not being replaced, and gains nothing from the son. The heir is not commissioned to succeed him but to do a job he is disqualified from by character. So none of the founder's usual anxieties apply — no rival risk, no legacy risk, no timing risk.
What the pair produces: the succession problem is not really about succession, it is about the principal's stake, and when the stake is removed the whole thing collapses into ordinary procurement — a spec, a capability gap, a hire. Which cuts both ways, and the second direction is the useful one: a founder who cannot discuss their successor in procurement language is telling you the role is about them. The test is cheap. Ask what the successor is for. If the answer is a task, you are looking at a hire. If the answer is a position, you are looking at the dilemma, and the paranoia is already installed.
Indirect Agency and Deniable Power (behavioral-mechanics). Power exercised openly invites opposition, because the agent is identifiable. Power exercised through intermediaries accomplishes the objective without triggering the defensive reflex, because there is nobody visible to oppose.
Śiva's disqualification runs this in reverse, and honestly.
He cannot kill certain asuras because he has no hostility toward them — they are "the subjects of your compassion." The remedy is to produce an agent who has not inherited the relationship. The son will kill what the father cannot, not because the father wants deniability but because the father genuinely lacks the enmity required.
The insight neither gives alone: indirect agency and conflict-of-interest recusal are the same manoeuvre viewed from opposite sides of the intent line, and from outside they are indistinguishable. Both produce a principal with clean hands and an intermediary who does the killing. Which means the hidden hand is not a reliable diagnosis of scheming — sometimes it is the honest arrangement, and the only thing that separates them is a fact about the principal's interior that no observer can check. Practical consequence: stop trying to read intent off the structure. Ask instead whether the principal disclosed the disqualification. Śiva's is stated out loud, by someone else, in front of everyone. That disclosure is the whole difference, and it is the part a scheme cannot easily fake.
Sharpest implication. The marriage is sold to Śiva as procurement. Brahmā lays out a capability matrix — some asuras killed by me, some by Hari, some by you, some by "the son who results from your seed", some only by māyā — and names Śiva's disqualification as his virtue: "since they are the subjects of your compassion, some can never be killed by you." Then he closes with a quotation, holding Śiva to a pledge made in an earlier age to take a wife: "Now please fulfil your pledge." Śiva accepts on exactly that ground, and in the same breath issues the pledge that will destroy Satī. This is the third justification given for one wedding — grievance to Nārada, cosmic continuity to Devī, threat assessment to Śiva — and only the first explains the timing, because nothing about the asuras changed; a god got laughed at. And the later two are sound, which is the difficult part: a real need discovered because someone was hunting for a reason. Four beings in the room know the real story and not one says it, including Viṣṇu, who named it as hatred six chapters earlier and here endorses the case in full. And the son is specified before the wife, then never arrives — Satī dies childless, and the asura-killer comes from the second marriage.
Generative questions.