Karma in this chapter is not a balance. It is a crop, and it has three stages, and what you can do about it depends entirely on which stage it has reached.
Both good deeds and wicked ones have three stages — the part that is the seed, the part that is the increase and the part where they are enjoyed.1
Seed: planted, nothing showing. Increase: growing, visible, still in the ground. Enjoyment: ripe, and arriving whether or not you want it.
The whole doctrine is what happens at each stage, and the striking part is the last one.
At the seed stage the remedy is knowledge, and the remedy is total.
As long as they are in the form of seed, they can be destroyed through jnana.2
Destroyed. Not reduced, not mitigated — gone, as though never planted.
This is the standard claim of the knowledge traditions, and here it is scoped precisely rather than asserted in general. Jnana does not destroy your karma. It destroys the part of it that has not started growing yet.
That scoping is more interesting than the unlimited version, because it tells you what jnana is being understood to do. It is not a solvent applied to the past. It is something that stops a thing from developing — which means it operates on potential rather than on history.
And it produces the practical asymmetry that runs through the rest of the passage: the earlier you get to a piece of karma, the cheaper it is, and the cost curve is not gentle.
At the increase stage, the remedy shifts from knowledge to labour.
In the stage of increase, they can be destroyed through the steps mentioned.3
The steps mentioned are the ones a few sentences earlier, sorted by the medium of the offence: dhyana for a sin committed in the mind, and japa, austerity and donated wealth for one committed in words.4
So the middle stage is workable but expensive. It costs recitation, physical discipline, and money.
And notice that the classification changes character between stages. Stage one is handled by a state — you either have jnana or you do not. Stage two is handled by acts, which can be counted, purchased and delegated in a way a state cannot.
Which means the tradition's whole ritual economy — the japa counts, the donations, the fasts — operates almost entirely at the middle stage. Everything in this volume that has a number attached to it is stage-two equipment.
Then the sentence the doctrine exists for, and it is the least comfortable thing in the chapter.
But in the stage of enjoyment, they have to be enjoyed. Even crores of good deeds will not suffice then.5
Have to be.
Crores of good deeds — the volume's own maximum unit, the number it uses for emancipation itself6 — will not suffice.
This is a hard limit, and the volume almost never states hard limits. Its characteristic move is the escape clause: build a rule, then void it.7 Ash purifies involuntarily. A thumb will do. There are no prohibitions. A plate of prasada clears brahmanicide.
Here there is no clause. Ripe karma arrives, and the arrival cannot be bought off at any price.
Three consequences follow, and the chapter draws none of them.
It puts a floor under the whole remedial apparatus. Every other page in this volume tells you a sin can be removed. This one says: only before it lands.
It explains suffering without blaming the sufferer twice. A person in the enjoyment stage is not failing at their practice. Their practice is working fine; the crop was planted years ago. That is a genuinely kind implication, and it is the same structure that makes the doctrine hard.
And it makes the timing of knowledge the whole game. Two people with identical karma, one who acquires jnana early and one who acquires it late, get completely different lives — not because one is better but because one got there while the seeds were still seeds.
You have been doing the practice for eleven years and this year your son died.
The first thing you do is audit yourself, because that is what the practice trains you to do. What did you fail to remove. Which count did you fall short on. Which morning did you skip.
And chapter 19 tells you: stop.
Not because you are blameless. Because whatever this is, it was planted long enough ago to have ripened, and the stage it has reached is not one your japa reaches. Crores of good deeds will not suffice. That is not a failure report. It is a statement about which tools work on which stage.
So you do not get relief and you do get something else, which is the end of a particular kind of self-interrogation.
What remains available is smaller and the chapter names it:
Worshipping devas, donating to brahmanas and performing austerities makes it easier for men to tolerate the period of enjoyment.8
Not removal. Tolerance.
You can make the year more bearable. You cannot make it not happen.
And on the other side of it — the years you are not yet in — the seeds are still seeds, and that is where the work goes now.
The phrasing deserves its own look, because of what it rules out.
The volume's standard remedial move is substitution at scale: enough japa, enough lingams, enough offerings, and the balance shifts. The arithmetic chapters run on exactly this — a crore of Parthiva lingams for emancipation, five crore japas to equal Sadashiva, thirty-six thousand offerings to end rebirth.9
Crores of good deeds will not suffice is a direct refusal of that method, in its own units.
Which makes this passage one of the very few places where the volume argues against its own dominant register. It is not saying the counts are wrong. It is saying they operate on a different stage of the process, and that a stage exists where they do nothing.
And it slots the arithmetic into place. All that counting is prophylactic. It works forward, on what has not ripened. It has never been a remedy for what is currently happening to you, and the chapters that sell it never say so.
Two clauses soften the wall without removing it, and they should be read precisely.
When seeds or sprouts are destroyed, only what remains needs to be enjoyed.10
That is a reduction of scope, not of the ripe portion. Clear the seeds and the sprouts and the harvest still comes — but nothing further is queued behind it. The practice shortens the tail rather than the present.
And:
Worshipping devas, donating to brahmanas and performing austerities makes it easier for men to tolerate the period of enjoyment.11
That is a change in the sufferer, not the suffering.
Put together, the honest offer of this doctrine is: you cannot alter what is arriving, you can stop more of it being on the way, and you can be a person better able to stand it. Which is a considerably more modest claim than the surrounding chapters make, and considerably more likely to survive contact with an actual life.
The chapter closes the passage with the plainest sentence in it — "therefore, a person who desires happiness must refrain from committing wicked deeds"12 — which, after all the machinery, is advice about seeds.
The structural reading, and it is what makes this page more than a technical note.
Read the three stages as three relationships to time.
Seed is the future — nothing has happened, everything is still open, and the tool that works there is knowledge, which is a way of not starting things.
Increase is the ongoing — a process underway that can still be interfered with, at a cost, by labour.
Enjoyment is the present arriving — closed, unnegotiable, already the case.
On that reading the doctrine is not really about morality at all. It is a claim that the past is closed, the future is open, and the difference between them is not moral but temporal — and that most religious effort is misfiled because it is aimed at the wrong tense.
That would explain the passage's tone, which is unusually sober for this volume: no threats, no hell, no supremacy claim, and a hard limit stated without any attempt to sell around it. It reads like a piece of inherited analysis that got absorbed rather than composed, and it sits in chapter 19 immediately after the warning that sin in a sacred place cannot be dislodged — another hard limit, in a book that mostly does not have them.13
Two absolute claims, adjacent, in the volume's most concessive text. Worth noting where the concessions stop.
Strongest support is the doctrine's internal consistency: three stages, three remedies of decreasing power, and an explicit statement that the tradition's own maximum unit fails at the third.
Tension one, and it is the big one. Crores of good deeds will not suffice is a direct contradiction of the counting soteriology that occupies most of this Samhita, stated in the same units.
Tension two. Jnana destroys seed karma outright, which makes knowledge categorically more powerful than any quantity of ritual — in a volume that ranks lingam worship above everything combined.14
Tension three. The passage's only concessions — shorter tail, better tolerance — are the two weakest offers in the book, made by a text that elsewhere promises liberation for overhearing a recitation.
Open question. Is the three-stage model this text's own, or standard sanchita / prarabdha / kriyamana karma theory in different vocabulary? The correspondence is close — Debroy glosses prarabdha elsewhere in the volume as "that bit of accumulated karma that has come to fruition in the present life," which is precisely the enjoyment stage.15 But the vocabulary here is agricultural rather than technical, and he does not connect them. If it is the standard model, the interesting question becomes why this chapter reaches for a crop metaphor instead of the received terms.
The vault holds this doctrine from the practitioner side, without the hard limit.
God Is Not a Fish Market argues that grace cannot be compelled by quantity. Chapter 19 supplies a mechanism for why — not because the god refuses, but because a portion of what is coming has already left the realm where quantity operates.
That is a stronger version of the same objection, and it comes from inside the tradition rather than from a modern teacher pushing back on it.
The Seed and Shoot Circularity of Grace uses the identical agricultural image for a different purpose — devotion producing favour producing devotion, "like a seed resulting from a shoot and a shoot resulting from a seed."16 Same metaphor, opposite structure: there it is a loop with no beginning; here it is a one-way sequence with a point of no return.
The vault should hold both, because the volume evidently found the seed image good for two incompatible jobs and used it for both within eight chapters.
Present Bias and Hyperbolic Discounting (psychology). The behavioral material shows that people systematically underweight distant consequences, and that the underweighting is steepest at the near end — which is why the cheap early intervention is the one nobody makes.
Chapter 19's cost curve is the exact inverse of that discount curve. Intervention is free at the seed stage, expensive at the increase stage, and impossible at enjoyment. So the cheapest action is the one furthest from any felt consequence, and the moment the consequence is felt, action has become useless.
What the pair produces: a doctrine of graded remediability is a direct countermeasure to present bias, and it is the only kind of moral teaching that could be — because a system that lets you fix things once they hurt has removed the reason to act before they do. The psychology pages describe the bias as a failure of judgement. This shows a tradition building its ethics around the bias as a fixed feature, by making the window close. Which also explains why the doctrine has to state the wall so bluntly: a hard limit is the only thing that reprices a distant risk.
Renegotiation vs Reenactment (psychology). The therapeutic material distinguishes material that can still be reworked from material that is being lived through — and holds that the intervention appropriate to one is actively harmful applied to the other, because treating an unfolding experience as a fixable error compounds it with self-blame.
Chapter 19 draws the same line and puts it in a soteriology. In the enjoyment stage the correct response is explicitly tolerance, not correction — worship and austerity make the period bearable, and nothing makes it stop.
The insight neither gives alone: a moral system that cannot say "this one is not yours to fix" will convert every misfortune into evidence of failed practice. That is the failure mode of any purely remedial religion, and this chapter installs the brake — the sufferer in the enjoyment stage is not doing it wrong, and the text says so in its own strongest units. The therapeutic corpus arrives at this as clinical caution; the Purana arrives at it as cosmology; both are protecting the same person from the same second injury.
Sharpest implication. Karma here is a crop with three stages, and what can be done about it collapses across them: seed karma is destroyed outright by jnana, growing karma is workable by japa, austerity and donation, and ripe karma has to be endured — "even crores of good deeds will not suffice then." That last clause refuses the volume's own dominant method, in the volume's own maximum unit, which makes all the counting in the surrounding chapters prophylactic rather than remedial: it works on what has not arrived and has never touched what is currently happening to you. What remains on offer is a shorter tail and a better-prepared sufferer. It is the most modest claim in the book, and the only one likely to survive contact with a bad year.
Generative questions.