Read a contract and your eye skips over etc.
It shouldn't. In a document where every clause is fought over, etc. is the only clause that can grow. Everything else has fixed edges. That one doesn't.
The definition of pure devotion has one. The verse forbids jñāna-karmādi — knowledge, action, and so on. Two named exclusions and then a door left open.
And when the commentators walk through that door, what's behind it is not a handful of odds and ends. It's three entire disciplines, each with its own texts, its own teachers, and its own claim to be the highest thing a person can do.
Viśvanātha spells the ādi out. Three items, plus a gesture at more.1
Phalgu-vairāgya — renunciation that doesn't help devotion. Phalgu means barren, worthless, a river that's dry in the sandy bed. Renunciation that produces nothing.
Aṣṭāṅga-yoga — the eightfold system. Posture, breath, withdrawal, concentration, absorption. Patañjali's whole architecture, named and set outside.
Abhyāsa-yoga — the repeated meditation on abstract spirit prescribed in the sāṅkhya texts.
Then: and other practices. All these are also prohibited.
Sit with the size of that.
Two words in a verse have just excluded classical yoga and classical sāṅkhya — two of the six orthodox systems of Indian philosophy — from counting as devotion.
Not condemned.
Not called false.
Excluded from the category.
That distinction matters more than it looks, and the tradition is careful about it in a way that gets flattened when it's summarised.
Nowhere here does anyone say the eightfold yoga doesn't work. Nowhere does anyone say sāṅkhya's meditation fails to deliver what it promises. The claim is narrower and stranger: these things deliver, and what they deliver is not this.
It's the difference between saying a road is impassable and saying it goes somewhere else.
Which means the exclusion list isn't a polemic. It's a set of directions. Each named item is a discipline that a serious person might reasonably confuse with this one, because both involve renunciation, discipline, inwardness, long practice, and a stated goal beyond ordinary life. The surface similarity is high. That's exactly why the naming is needed.
Of the three, phalgu-vairāgya is the one I find hardest to dismiss, because it doesn't name a rival school. It names a way of being that anyone can fall into.
Renunciation that produces nothing. You give things up. The giving-up becomes the thing. And nothing grows in the bed of it.
The tradition has a paired term it uses elsewhere — yukta-vairāgya, renunciation that's connected, where what's given up is given up for use. The tāpasa devotees in Text 18 are described as practising exactly this.2 So the contrast is available in the same book: two renunciations, identical from outside, distinguished only by whether anything is being served.
And notice the failure mode this names. Not indulgence. Austerity performed as an end. The person failing here is doing more than most, harder than most, and the barrenness is precisely what their effort produced.
Here's where the clause gets interesting rather than just severe.
Having swept karma into the exclusion list, Viśvanātha immediately protects an enormous amount of it. All the daily and occasional duties of the smṛti texts are forbidden. All devotion mixed with reward-seeking action or with speculative knowledge is forbidden. But the activity of sevā-paricaryā — service and attendance on the Lord — is not karma in the forbidden sense at all, and can never be prohibited.3
The reasoning is clean. Those activities relate to devotion. They fall inside kṛṣṇānuśīlana. So they were never in the excluded category to begin with — not exempted from it, never in it.
Which reframes the whole clause. The verse isn't banning action. It's banning action performed under a different theory of what action is for. Sweep the temple floor believing it's service, and it's a limb of devotion. Perform the identical sweep believing it discharges an obligation you'd incur sin by skipping, and you've just done the forbidden thing.
Same broom.
Monday, 5:40 a.m. You're sitting for twenty minutes of breath-work before the chanting because a physiotherapist said it would help your back and it does.
Nothing about this is forbidden. It isn't aṣṭāṅga-yoga in the excluded sense, because you're not doing it as a path anywhere. It's maintenance on the instrument.
Now shift one thing. You start noticing the breath-work is better — cleaner, more absorbing, more reliably good than the chanting, which has been dry for weeks. You extend it to forty minutes. The chanting slips to the end and then to whenever.
Nothing external has changed. The mat is the same, the room is the same, the back is the same. But the twenty minutes has become a path, and the clause has closed on it. Not because breath-work is bad. Because it's now the thing you're going somewhere on.
Third scene, subtler. You've given up alcohol, then television, then most of your friendships, and you notice a small hard pleasure in each subtraction. Someone asks what the space is for and you don't have an answer that isn't the subtraction itself.
That's the dry riverbed, and it took eighteen months of genuine discipline to build.
There's an obvious question the clause raises: why bother naming the eightfold yoga? Nobody drifts accidentally into Patañjali's system. It takes instruction, effort, a teacher.
I think the answer is in who this book was written for. This is a manual for people already inside a renunciant culture — men who have left households, who live near temples, who spend their days on disciplines, and for whom the available prestigious options are exactly aṣṭāṅga-yoga, sāṅkhya meditation, and hard renunciation. These are not exotic temptations. They are the neighbouring professions.
And they carry something devotion at this stage often doesn't: measurable progress. Postures you can hold longer. Breath you can extend. Concentration that visibly stabilises. Whereas the early stages of devotion, on this book's own account, feel like effort in the dark with anartha still thick in the heart.
So the clause is guarding against the most likely defection, not the most exotic one. The practitioner who leaves does not leave for pleasure. He leaves for a discipline with a progress bar.
One last structural point, and it's the reason this word deserves a page.
Rūpa could have written a closed list. He wrote ādi. That means the exclusion is defined by principle rather than by enumeration — anything that covers devotion is out, whether or not anyone thought of it in 1541.
That's a considerable amount of trust to place in later readers, and it cuts both ways. It makes the definition durable: therapy, self-optimisation, whatever arrives next, all get tested by the same criterion rather than escaping on a technicality. It also makes it expandable by whoever holds the pen. An open exclusion clause in a doctrinal text is a standing licence to exclude, and the licence sits with the commentator.
The book uses it modestly here — three items, all of them genuinely near-neighbours. But the mechanism has no brake in it.
Strongest support. The clause is unpacked with named disciplines rather than left gestural, and the immediate carve-out for service-related action shows the criterion is doing conceptual work rather than blanket prohibition. Phalgu-vairāgya has a stated positive counterpart (yukta-vairāgya) in the same book, so the contrast isn't invented here.
The tension the book does not close. The excluded practices are excluded as paths, not as activities — and the book never gives a test for when an activity has become a path. The line between breath-work for your back and breath-work as a route is entirely internal, and by the time you'd want the test you're the least reliable person to apply it.
A second tension, sharper. Aṣṭāṅga-yoga is excluded here in Text 1. But Text 4 will prescribe sixty-four limbs including posture-adjacent practice, controlled breathing in japa, and sustained one-pointed remembrance that the book itself calls dhyāna. The techniques recur under devotional description. What's excluded is the system and its goal, not the operations — and the book doesn't mark that, which leaves a reader with the impression of a prohibition much broader than the one actually in force.
Open question. Is ādi's content Rūpa's or Viśvanātha's? The verse says only and so on. Every item on the list — the barren renunciation, the eightfold yoga, the sāṅkhya meditation — is supplied a century later by a commentator with his own opponents. The word is Rūpa's; the list may be entirely Viśvanātha's map of his own moment.
Come to this from the liberation-models material and the experience is disorienting. There, the eightfold yoga and sāṅkhya's discriminative meditation are models of liberation — serious, ranked, argued between, each with a defensible account of what freedom is and how you get there. You spend that reading weighing them.
Arrive here and nobody weighs anything. They're on a list of things that must not be lying on top of something else. The evaluative frame has vanished and been replaced by a spatial one — covering, uncovered, on top of, underneath.
That shift is the actual content of the disagreement, and it's easy to miss because no argument is made. The rival systems aren't refuted. They're relocated — from competing answers to the same question, into contaminants of a different one. Whether that's a stronger move or a way of declining the contest is exactly what the surrounding literature has spent centuries on.
An open exclusion clause outlives its author and transfers the power to exclude. Hold this beside Narrative Premise as Meta-Filter. A premise governs what may enter a story, and a good one is stated as a principle rather than a list of banned scenes — which is what lets it keep working on material the author never imagined. Ādi is that same construction inside a definition: not these three practices are excluded but whatever covers this is excluded, applicable forever.
What the pairing surfaces that neither gives alone: a principle-stated filter is durable and unbounded in the same stroke, and the unboundedness accrues to whoever inherits it. A story's premise-filter is applied by the author, who bears the cost of applying it badly — cut too much and the story dies in your hands. A doctrine's exclusion clause is applied by commentators who bear no comparable cost; excluding too much makes the tradition narrower, and narrower is rarely experienced from inside as a failure. So the same design that makes a filter future-proof makes it a standing instrument for whoever holds the pen next. Viśvanātha uses it on three genuine neighbours. Nothing in the mechanism required him to stop there, and nothing in it would have told him if he hadn't.
Sharpest implication. The interesting exclusions in any tradition are the ones aimed at its neighbouring professions, not at its opposites. Nobody guards against pleasure — pleasure is obviously not the thing. What gets named is the adjacent discipline that shares the surface and offers measurable progress, because that's where the serious people actually go.
Generative questions.
Phalgu-vairāgya names austerity that produced nothing, built by genuine effort. What's the equivalent in a domain you actually work in — the discipline that shows every sign of rigour and irrigates nothing?ādi's list is Viśvanātha's rather than Rūpa's, it's a map of one century's rivals presented as timeless. What would an honest 2026 list contain?anāvṛta covering test that governs how it appliesphalgu-vairāgya, arrived at from a different direction