Ten thousand young men are at a lake called Nārāyaṇa-saras, performing austerities before taking up the work their father assigned them: populate the world. They are not in doubt. They are not seeking instruction. They are doing preparatory practice for a task they've accepted.
Nārada arrives uninvited and opens with an insult:
"O Haryashvas! How can you possibly create subjects? Although you are protectors, you have not seen the ends of the earth and are foolish."1
Then the riddles, delivered in a rush, with no explanation of what they mean or why he's asking:
"There is a kingdom with a single man. There is a hole whose exit cannot be seen. There is a woman who assumes many kinds of forms. There is a man who is the husband of a harlot. There is a river that flows in both directions. There is an extraordinary house that is made out of twenty-five materials. There is a swan with a wonderful tale. There is a disc made out of razors and it revolves on its own."2
And a closing challenge: "Your learned father has commanded you, but you are ignorant. Without understanding what is appropriate for you, how can you possibly create?"3
The Haryaśvas work them out, and the text gives their reasoning, which is how we know what the riddles meant. Each is a compressed argument, and the solutions are technical:
The kingdom with a single man — the field of embodied existence, occupied by the one soul, which binds itself without a second party doing the binding. The hole with no visible exit — heaven, from which one returns to rebirth, so the apparent escape is a cul-de-sac. The woman of many forms — the intelligence, described as having "the gunas of a wanton woman," shifting shape and never settling. The husband of a harlot — the person attached to that intelligence, who "loses all one's powers and roams around in saṃsāra."
The river flowing both directions — māyā, running toward both creation and destruction at once. The house of twenty-five materials — the body, built from the twenty-five tattvas of Sāṃkhya. The swan with a wonderful tale — the account of the supreme being, the fourth beyond the three states. The self-revolving razor disc — time, turning independently, cutting whatever it reaches.
Eight images, each carrying a full doctrinal position, none explained. The riddles function as a compression format: an entire metaphysics delivered in a paragraph, unpacked by the listener rather than transmitted directly.
The doctrinal content isn't what moves the Haryaśvas. Look at how their reasoning is actually structured — every single solution terminates in the same question:
"Without having witnessed its emancipation, why should one indulge in acts that are transient?... If that being without an origin has not been seen, why should one indulge in acts that are transient?... Without realizing the transcendental abode, why should one indulge in acts that are transient?"4
The phrase repeats after every riddle. Eight solutions, eight identical closings.
That repetition is the mechanism, not the metaphysics. Each riddle establishes a different fact about existence, and each fact gets converted into the same conclusion by the same rhetorical move — given this, why do the transient thing? The Haryaśvas aren't reasoning their way to eight independent conclusions that happen to converge. They're applying one template eight times, and the template has the answer built into it.
Which makes the riddles less a teaching method than a drilling method. The doctrinal variety is genuine; the conclusion was fixed before the first riddle was posed.
Dakṣa fathers a thousand more sons and sends them to the same lake. Nārada arrives again, and this time he doesn't even bother with the riddles:
"O sons of Daksha! Listen to my words of instruction. You are devoted to your brothers. Therefore, you must search out the footprints of your brothers. A brother who knows about dharma must generally follow his brother."5
That's the entire second operation. Not an argument about the nature of existence — an appeal to family loyalty, pointing at a precedent. The text notes flatly: "The sighting of Narada does not fail."
It's worth registering what that substitution reveals. If the riddles were doing the real work the first time, the second batch should have needed them too. They didn't. What was needed was a reason to go, and the second group already had one available in the form of their brothers' example. The elaborate metaphysics turns out to have been optional equipment.
There's a strong tradition, in this text's own civilization, of knowledge transmitted through question and answer — a student asks, a teacher responds, and the asking is what makes the answer usable. This entire Purāṇa is structured that way: Parīkṣit asks, Śuka answers, and the frame repeats at every level of nesting.
Nārada's riddles invert it. Nobody asked. He arrives at a lake where young men are doing exactly what they're supposed to be doing, opens by calling them foolish, poses questions they didn't request, and leaves once they've drawn the conclusion he came to produce.
That inversion is worth naming precisely, because the text elsewhere treats the question-first format as normative and here treats its violation as successful. Both can't be the tradition's actual position on how transmission works.
The detail that most complicates a charitable reading of the riddles is what the text shows Nārada doing the moment the Haryaśvas leave:
"The sage continued to roam around the worlds, fixing his undivided mind on Hrishikesha's lotus feet and chanting the tune of the brahman."6
He doesn't follow them. He doesn't check whether the detachment held, doesn't accompany them into the practice he's just redirected their entire existence toward, doesn't establish any ongoing relationship with eleven thousand men whose lives he has permanently altered. He moves on, singing.
Read one way, that's the appropriate detachment of a teacher who has done what was needed and claims no ownership of the outcome. Read another, it's the signature of an intervention whose success was measured entirely at the moment of departure — did they leave, yes — with no interest in anything downstream.
The text gives no basis for choosing between those readings, and both are consistent with everything shown. What can be said is narrower and worth stating plainly: no follow-up occurs, and the text does not appear to consider follow-up relevant to whether the operation succeeded. Success is defined as the sons leaving. Whether the detachment they carried away was the durable kind is a question the narrative never returns to, from either direction.
You have something you're certain about, and someone in front of you who hasn't asked and isn't in doubt.
The riddle format is genuinely tempting here, because it feels less coercive than simply telling them. You're not lecturing — you're asking questions, letting them reach it themselves, respecting their capacity to work it out. That framing is why the technique is durable and why it survives in every coaching methodology and every Socratic classroom.
Nārada's operation is worth holding as the boundary case. His questions are real questions with real answers, and the Haryaśvas genuinely solve them. And the conclusion was still fixed in advance, embedded in the repeated closing move rather than in any of the eight solutions, which is why a completely different approach — pure family-loyalty appeal — produced the identical outcome on the second group.
The diagnostic that separates genuine inquiry from drilling: does the format admit an answer you didn't want? If every path through the questions arrives at the same place, the questions are delivery, not inquiry. That's not automatically illegitimate — sometimes the conclusion really is correct and the format is just gentler than assertion. But it should be named accurately rather than mistaken for the student's own discovery.
Every branch terminates identically. Watch for the repeated closing move. Eight different premises reaching one conclusion by one identical step means the step, not the premises, is carrying the argument.
The elaborate version turns out to be optional. When a simplified approach on a second attempt produces the same result, the sophistication of the first attempt wasn't load-bearing.
Nobody asked. The strongest signal, and the easiest to overlook, because unrequested teaching frequently arrives framed as generosity.
Strongest support: the text supplies the Haryaśvas' full reasoning rather than merely reporting that they were convinced, which is what makes the repeated closing formula visible. Without their worked solutions, the uniformity of the conclusion would be invisible.
Tension, unresolved. The riddles are presented approvingly, and Nārada is an authoritative figure throughout this corpus. Yet the immediately following chapter has Dakṣa charge that this exact method produces false detachment, and the text never answers him. Whether the riddles are a legitimate teaching method or a sophisticated form of what Dakṣa calls installing conclusions is left standing.
Open question. The eight riddles are solved in the text, but Nārada's own framing mentions a ninth element the solutions don't clearly address — the ordering and count vary depending on how the images are parsed. Whether the traditional count is eight or nine, and whether any image goes unsolved, would need checking against a commentarial edition this translation doesn't reproduce.
Compare how this text handles its own primary frame. Parīkṣit asks and Śuka answers, repeatedly, at every level — and when Parīkṣit raises an objection, as he does about the futility of atonement, the objection genuinely shapes what comes next. The asking is load-bearing there. Nārada's riddles operate on the opposite principle within the same volume, arriving unrequested at people in no doubt, and the text presents both as effective transmission without ever noting they're incompatible accounts of how understanding gets produced.
With Question and Answer as the Native Transmission Method (creative-practice). That page documents a practitioner who has kept a five-year rule of not answering until someone asks, and names the payoff as peace rather than better pedagogy. It also states the tradition-level claim directly: question and answer is "one of the most fundamental ways of transferring knowledge in our civilization... It is never lectures. We have never done lectures."
Nārada violates that rule completely and succeeds, twice, in this text's own telling. What neither page states alone: the "never lectures" claim and the Nārada episode both come from inside the same tradition, and they can't both describe its actual practice. The resolution isn't that one is wrong — it's that the tradition appears to hold two distinct transmission modes and to license the unrequested one specifically when the recipient is understood to be heading somewhere harmful. Nārada doesn't wait to be asked because, on the text's own framing, the Haryaśvas' assigned task is itself the problem.
Which sharpens the other page's rule into something more precise than "wait to be asked." The waiting discipline appears designed for cases where the questioner's own trajectory is fine and premature answering would merely be unhelpful. It says nothing about the case where you believe the person's entire direction is wrong — and that's exactly the case where the temptation to speak unasked is strongest, and where this text supplies a precedent for doing so that its own next chapter then attacks.
Sharpest implication. A teaching format that lets someone reach a conclusion themselves can still have the conclusion fixed in advance — and the test isn't whether they did real reasoning, but whether the format could have produced a different answer. Eight genuine solutions reaching one predetermined endpoint is drilling wearing inquiry's clothes.
Generative questions.