Think of somebody you know who is genuinely kind. Not performing it — actually kind, in the small unwatched moments where there is no credit available.
Now try to imagine them working on it. Setting a reminder. Reviewing their kindness at the end of the week.
It does not fit, and the reason is not that they never tried. The ones who got there mostly got there sideways. They were doing something else — raising somebody, nursing somebody, staying somewhere hard for a long time — and the quality arrived as a consequence of the thing they were actually doing.
Whereas the people you have watched work directly on a virtue usually end up with a performance of it, and often with something worse than they started with.
Madhya 22 makes that an explicit doctrine, and then does something with it that undercuts it.
The chapter lists the qualities of a devotee at length — truthfulness, cleanliness, equanimity, gravity, compassion, the whole inventory.1
And the claim about them is that they are not cultivated. They arrive. A person who develops devotion is found to have these qualities; the qualities are not a syllabus and were never worked on directly.
Two images carry it.2
Water the root, and the trunk, branches and twigs are fed. You do not water branches. There is one point of application and everything downstream is supplied from it.
Feed the stomach, and every sense is satisfied. You do not feed the eyes.
Both say the same thing and the second is sharper, because a stomach is not merely upstream of the senses — it is the only route to them. You cannot feed an eye directly even if you want to.
This is a claim about causal architecture rather than about virtue, and it is worth pulling apart before deciding whether it is true.
It removes ethics as an independent activity. If the qualities are downstream of one practice, then working on them directly is not merely inefficient — it is watering branches. The corpus does not say ethics is unimportant. It says ethics is not a place you can act.
It makes one point of intervention carry everything. Efficient when it works, and brittle in a specific way: there is no second route. A person whose devotion is not progressing has, on this account, no available path to becoming less cruel, because the only tap is the one that is not producing.
And it converts character into evidence. If the qualities follow devotion and nothing else, their presence is a reading on devotion and their absence is one too. That diagnostic is deployed nowhere in the corpus's disciplinary chapters, where it would have decided several cases differently.
Here is the problem, and it is internal.
The same chapter that says the qualities arrive as a consequence also grades practitioners by them. Three tiers of devotee, sorted by scriptural competence and faith, with the behavioural symptoms attached as the sorting criteria.3 Elsewhere the corpus supplies its one genuinely observable test of a first-class Vaiṣṇava: the man whose presence makes others chant.4
Those two moves are not compatible in the way the chapter needs them to be.
If good qualities are strictly downstream of devotion, they are a perfect proxy and selecting on them is fine. But then the corpus's insistence that the real criterion is an invisible interior state is redundant — you could simply look at the person.
And if the interior state is the criterion and the qualities can appear without it — which the corpus asserts directly elsewhere, in its doctrine that convincing symptoms occur in people who lack the thing entirely5 — then the qualities are not a reliable proxy and grading by them is unsound.
The corpus holds all three positions: the qualities follow devotion, the qualities can be counterfeited, and practitioners may be graded by the qualities. Any two of them work together. The three do not.
It is worth saying what the doctrine gets right, because the observation behind it is sound and unusually well chosen.
Directly pursued virtues have a characteristic failure mode, and everybody has met it. Humility worked on becomes a performance of humility. Generosity as a project becomes a ledger. The pursuit installs a monitor, and the monitor is the thing that was supposed to be absent.
The root-and-branch image handles this correctly, and it is the same structure as this corpus's other good practical instruction. Its remedy for offence is a fence rather than resolve; its account of practice makes the practitioner responsible for supply and not for outcome; and here, the qualities are what happens rather than what you attempt.6
This tradition consistently declines to make the target the object of effort, and puts effort somewhere upstream instead. That is a real and coherent theory of how change happens, and it is more sophisticated than most of what surrounds it.
One thing about the list is easy to miss because it is so familiar.
Nothing on it belongs to this tradition. Truthfulness, cleanliness, compassion, equanimity, gravity — this is the common virtue inventory of the whole culture the corpus is arguing inside. Its rivals name the same qualities. So do its opponents.
Which means the doctrine is not claiming a distinctive product. It is claiming a distinctive route — everybody agrees on what a good person looks like, and the corpus is asserting that only one road gets there and the others are watering branches.
That is a much harder claim than it first appears, and it fails in a specific way. A route claim cannot be checked from the destination. If two people are equally truthful and one arrived by devotion and one by a rival discipline, nothing in the outcome distinguishes them, and the corpus's own counterfeit doctrine says so explicitly.
So the tradition has staked its argument on the one part of the process that leaves no trace in the result — which is convenient, and is also, given everything else here, entirely characteristic.
Layer A gives the quality list and the two images compactly, inside the process instruction, without argument.7
Layer B expands the list into a portrait and presses the inference in the other direction — from an absence of qualities back to a diagnosis of the practitioner.8
That reversal is not in the verse. The verse says the qualities follow devotion. Reading absence as proof of deficient devotion is a further step, and it is the step that makes the doctrine usable as an instrument on other people.
It is also the same shape as the corpus's other diagnostic reversals, where a method that produces no result establishes something about the practitioner rather than about the method. Layer A supplies the causal claim; the disciplinary use of it is largely Layer B's.
You have decided to become more patient and it is not going well. Check whether you are watering a branch. Some qualities are not reachable directly, and the pursuit installs exactly the self-monitoring that the quality requires the absence of.
You are looking for the root. The corpus's answer is one specific practice; the general form of the answer is that there is a smaller number of upstream things than you think. Finding yours is worth more than a list of virtues, and the test is whether working on it changes things you were not aiming at.
You are assessing somebody by their conduct. If your framework says conduct is a byproduct, it also says conduct can be produced other ways, and you have no way to tell a byproduct from a direct imitation by looking at it. That is the corpus's own position and it does not act on it.
Somebody's difficulty is being read as evidence about their inner state. Notice the direction of that inference. Qualities follow the practice does not license no qualities, therefore no practice — that is the contrapositive of a claim nobody made, and it is where a description of how change works turns into a way of judging people.
The evidence is a quality list plus two analogies, stated inside the process instruction and treated as settled.9 Neither image is defended; both are offered as obvious.
The tension the source doesn't close: the corpus holds that qualities follow devotion, that qualities can appear without devotion, and that practitioners may be graded by qualities. The three cannot all stand.
A second one. The doctrine supplies the most usable diagnostic in the corpus — look at the person's character — and the disciplinary chapters do not use it, deciding instead on interior states nobody can inspect.
Open: what a practitioner with no upstream progress should do about their conduct in the meantime. The doctrine forecloses direct work and nothing in the range read supplies an alternative.
This page and The Weeds Beside the Creeper are the two halves of one account of what practice produces. The same water grows the crop and the parasites, and the crop's good qualities arrive unbidden. Put together they describe a practice that is indiscriminate in what it feeds and precise in what it yields, which is an odd pair of claims about one input, and the corpus makes both about the same watering.
Against The Two Formidable Enemies it is the constructive counterpart. That rule protects the practice by removing people; this doctrine says the practice generates its own results without needing anything added. The corpus's positive mechanism is generous and its protective mechanism is severe, and they are three verses apart.
And it belongs beside Sādhana Awakens What Is Already There, which makes the same structural move about love itself: not produced, uncovered. The corpus prefers to describe every good thing as arriving rather than as being made — which is consistent, and worth noticing as a habit rather than reading as a finding each time.
Three Products of Evolution: Adaptations, Byproducts, Noise — that page's distinction is between traits selected for a function, traits that come along attached to something that was selected for, and traits that are neither. The classic example is bone whiteness: nobody was selected for white bones; calcium was selected for hardness and is white.
The corpus is making a byproduct claim in exactly that technical sense, and the comparison shows what the claim needs in order to be checkable.
In evolutionary work, the byproduct claim is testable — badly, expensively, contentiously, but testable. You look for the selection pressure. If you cannot find one for the trait itself, and you can find one for the thing it is attached to, the byproduct reading gains.
Here there is no equivalent move. A person is truthful and devout. The doctrine says the truthfulness came along with the devotion; the alternative is that it was there already, or was worked on directly, or is being performed. Nothing distinguishes these from outside, and — this is the part the corpus itself supplies — nothing distinguishes them from inside either, because the same corpus holds that convincing symptoms occur in people who have none of the thing.
Which gives the general finding. A byproduct claim is a claim about causal architecture, and its whole practical value is that it tells you where to intervene. That value survives being unverifiable. But the second use — reading the trait backwards as evidence of the cause — requires exactly the verification that is missing, and it is the use every tradition actually makes. The evolutionary literature keeps those two apart as a matter of method. Here they run together, and the doctrine that tells a practitioner where to put their effort is the same sentence that licenses a community to read somebody's character as a verdict on their interior life.
Sharpest implication. Good qualities follow the practice is excellent advice about where to spend effort and a terrible instrument for judging anybody, and the two uses are the same sentence read in opposite directions. Traditions almost always end up making the second use, because it is the one that scales to a community.
Generative questions.