Most moral instruction tells you what to do. This passage does something rarer: it enumerates the specific ways an action can pass as dharma while not being it, and gives each failure mode a name.
The five are not sins. Nobody in this taxonomy is doing anything obviously wicked — no theft, no violence, no lie. They are ways of being correct in form and wrong in substance, which is why they need naming at all. An obvious wrong does not require a taxonomy.
The enumeration:
"A person who knows about dharma must avoid five kinds of adharma — vidharma, paradharma, abhasa, upama and chhala. These are five branches of adharma."1
The definitions, which follow immediately:
"Obstructions to dharma are known as vidharma. Dharma urged by someone else is known as paradharma. Upadharma is that followed by insolent heretics. Debating with words is chhala. If a man deviates from his own dharma and wilfully does whatever he wants, that is known as abhasa."2
The fourth term does not match itself. The list names upama; the definitions name upadharma. One of the two is what the passage means and the other is not, and Debroy's English does not reconcile them. This build's own PRD silently harmonised the pair to upadharma — a small correction, recorded here because the vault's practice is to report the text rather than the tidied version.
The definitions also arrive out of the listing order — vidharma, paradharma, upadharma, chala, ābhāsa against vidharma, paradharma, ābhāsa, upama, chala — which is a further sign that the passage is transmitting a fixed list rather than composing one.
Vidharma — obstruction. Not doing the wrong thing but doing something that prevents dharma. The category covers acts that are unobjectionable in themselves and whose effect is to stop something necessary from happening.
Paradharma — someone else's dharma. Dharma urged by someone else. The failure is not that the conduct is bad; it is that it belongs to a different person's role and has been adopted. This is the Gītā's position — better one's own dharma imperfectly than another's well — restated as a species of counterfeit rather than as a preference.
Upadharma — the heretics' version. That followed by insolent heretics. Structurally the most fragile item, because it is defined by who practises it rather than by what it does, and therefore requires an authority to adjudicate membership.
Chala — debating with words. Verbal manoeuvre. Not lying, and not necessarily arguing badly — sophistical technique, the win obtained on the word rather than on the matter.
Ābhāsa — semblance. Deviating from one's own dharma and wilfully doing whatever one wants. The most interesting of the five, because it names the case where a person's conduct is unconstrained and still looks like a practice — self-will operating under a form.
Four of the five require somebody to make a determination. Obstruction requires knowing what was obstructed. Another's dharma requires knowing whose is whose. The heretics' version requires an authority to identify heretics. Verbal manoeuvre requires distinguishing a good argument from a clever one.
Ābhāsa alone is checkable by the person doing it, and the chapter supplies the test in the sentence immediately after the list:
"If one follows one's natural and ordained dharma, why is that not capable of bringing about serenity?"3
Serenity is the diagnostic. Not correctness of doctrine, not approval, not results — the presence or absence of settlement. If the practice is not producing it, the practice is a semblance of one.
That is a genuinely usable test and it is also, read carefully, a rather large concession. It makes the outcome in the practitioner the arbiter of whether the practice was real, which sits oddly beside a taxonomy whose other four items are adjudicated from outside.
The five are not left abstract. The surrounding chapter runs a worked example immediately before them and a set of applications immediately after.
Before: an extended argument that meat must not be served or eaten at śrāddha ceremonies — a rite that is unimpeachably correct in form and, the passage argues, disqualified by what it costs. "On seeing those who sacrifice with sacrificial objects, creatures become terrified and say, 'This one is without compassion and ignorant. He will certainly kill me, desiring to take life.'"4 A perfectly performed rite, named as a counterfeit by its content.
After: a list of āśrama-specific frauds, which is the taxonomy applied to stations rather than acts. "Avoidance of rites by a grihastha, the giving up of vows by a brahmachari, residence in a village by an ascetic and the agitation of the senses by a mendicant — these are indeed the vilest in the respective ashramas and amount to fraud in the ashramas."5
Four stations, four characteristic frauds, each consisting of retaining the title while dropping the constitutive practice. And then, unusually, an instruction about how to treat such people: "They are confounded by the divinity's maya. Out of compassion, one should disregard them."6 Not exposure and not correction — disregard, with the reason given as their being deceived rather than dishonest.
It is worth asking what work a list of counterfeits does that a list of virtues cannot, because the chapter supplies both and they are not redundant.
The thirty signs at 7(11) tell you what dharma consists of. They are useless against a counterfeit, because every counterfeit satisfies them at the level of description. A man performing another's dharma is being truthful, austere, pure and disciplined; he is doing all of it. A man debating with words is studying, discriminating and inquiring. The positive list cannot catch them because the positive list is about content and the failure is about fit.
That is the taxonomy's actual claim, and it is a strong one: dharma is not a property of an action but of an action-in-a-position. The same act is dharma for one person and paradharma for another; the same argument is inquiry in one mouth and chala in another; the same rite is worship performed by a householder and upadharma performed by someone the tradition does not recognise.
Which explains why three of the five need an adjudicator. If correctness depends on fit, then somebody has to say what the position is — and the only counterfeit that escapes this is ābhāsa, because a person deviating from their own dharma is the one case where the position is not in dispute. They know what theirs was. They left it.
The five-fold taxonomy is not the Bhāgavata's invention; the terms belong to a wider dharmaśāstra vocabulary that the vault does not currently carry, and the compression shows in exactly the place a compression would show — the fourth term is unstable between the enumeration and the definition.
That instability is the most useful thing here for the vault's own practice. A build that harmonises upama to upadharma, as this build's PRD initially did, produces a cleaner page and loses the one piece of evidence that the passage is transmitting a received list rather than composing one. The out-of-order definitions point the same way. Both are small, both were nearly smoothed, and together they are the difference between a page that reports a taxonomy and a page that reports a text carrying a taxonomy.
The wider point generalises to the corpus: where a Purāṇa states a technical list, the discrepancies are the evidence of transmission, and they are the first thing an editorial hand removes.
Run the paradharma check on your obligations, not your ambitions. Dharma urged by someone else is the item most likely to be live in an ordinary life, and it does not feel like a moral failure at all — it feels like being conscientious about something you were handed. The question is not is this good but is this mine, and the two come apart most often in work and in family.
Use serenity as the ābhāsa test, with the caveat attached. The chapter's diagnostic is that one's own proper practice settles a person. That is checkable and it is not infallible: a practice can be genuinely yours and still be difficult for a long stretch. Use it over months rather than days, and use its absence as a prompt to examine rather than as a verdict.
Notice which counterfeits require an authority. Three of the five do, and one of those — upadharma — is defined entirely by who is practising. Any taxonomy of false practice containing a category whose membership is decided by the party issuing the taxonomy is doing two jobs, and it is worth knowing which one is operating on you.
And take the disposal instruction seriously, because it is counter-intuitive. The text's response to āśrama fraud is neither exposure nor reform. It is to disregard, on the grounds that the person is deceived rather than deceiving. That is a genuine and unusual position on how to handle people whose practice has become nominal, and it costs nothing to try.
The form is impeccable and something is missing. All five counterfeits pass inspection at the level of form. Where the form is what you have been checking, you have not been checking.
A category is defined by its practitioners. That followed by insolent heretics. Definitions of this shape identify a group rather than a behaviour, and they can be applied to anyone the definer wishes.
The argument is being won on the word. Chala. Watch for the point at which a disagreement resolves into a dispute about a term, and notice who benefits from the resolution.
The title is intact and the practice has lapsed. The four āśrama frauds are all this shape: householder without rites, student without vows, forest-dweller in a village, mendicant with agitated senses. The tell is always the retained name.
Strongest support. The list and its definitions are consecutive; the śrāddha-meat argument immediately precedes them as a worked case; and the āśrama-fraud list immediately follows as an application, with a stated disposal instruction. The taxonomy is used, not merely stated.
Tension the page preserves. The enumeration says upama and the definitions say upadharma. One of them is wrong and the passage does not say which. The vault's own PRD harmonised them; this page reports the discrepancy instead.
A second tension. Four of the five counterfeits require external adjudication and the fifth is validated by the practitioner's own serenity. A taxonomy mixing those two authorities has not decided who is entitled to declare a practice false.
Absence-claim discipline. No claim is made about whether the Bhāgavata Purāṇa elsewhere resolves the upama/upadharma discrepancy. Within the range read for this build — Skandhas 5–9 complete and 10(1–11) — the pair appears once.
Open question. Ābhāsa is glossed as wilful deviation from one's own dharma, which makes it nearly identical to paradharma under a different description. Are these two failure modes or one seen from two sides, and does the Sanskrit distinguish them more sharply than this rendering does?
With The Respectability Shield (behavioral-mechanics). The mechanism there is that correct form functions as armour: a party whose conduct is procedurally impeccable becomes very difficult to challenge on substance, because the challenge has to begin by disputing something everyone can see is in order.
This taxonomy is the same observation from the defender's side, and it is unusually well built. Each of the five names a way the shield can be intact while the thing behind it is absent, and the āśrama-fraud list makes the mechanism explicit: the title retained, the constitutive practice dropped. The householder who has stopped the rites is not doing anything anyone can point at. He is a householder.
What neither page states alone: a system that hands out durable titles necessarily generates a counterfeit for each one, because a title is precisely a claim that survives the lapse of what earned it — so the taxonomy of frauds can be derived from the taxonomy of stations without observing anybody. And the text's disposal instruction, out of compassion, one should disregard them, is a striking practical conclusion: it concedes that the shield cannot be pierced from outside and declines to try. [n=1 OBSERVED]
Sharpest implication. Five named ways to be correct in form and wrong in substance, three of them requiring an authority to adjudicate, one validated only by whether the practitioner is at peace — and the text's recommended response to the fraud it describes most precisely is to leave the person alone.
Generative questions.