The Hanumanic Strategy corpus tells its reader two opposite things about anger, in two documents, and neither mentions the other.
Document one — keep it. The Shadow-Devotion Paradox holds that "authentic devotion requires integrating shadow energies rather than transcending them," that true bhakti includes controlled aggression in service of dharma, and it proposes a Sacred Anger Meditation for "channelling righteous anger into protective energy for dharmic purposes."1 Anger is fuel. It stays anger and gets aimed.
Document two — convert it. The implementation system's months 1–2 introduce "simple anger-to-fire rasa practice (Raudra transformation)", and months 3–4 specify the destination: "Anger management through Raudra-Shānta progression."2 Śānta is the peaceful rasa. Anger is raw material, and the practice's success condition is that it stops being anger.
These are not two emphases. One says the rage is the resource and must be preserved to be useful; the other says the rage is the input and success is measured by its disappearance. A practitioner cannot run both.
It would be easy to say both mean "handle anger well." They do not, and the divergence shows up the moment either is operationalised.
The integration account produces a practitioner who, presented with an injustice, becomes more capable of fierce response and better aimed. Its own recognition marker is the ability to say no with love and enforce boundaries with devotion.1
The conversion account produces a practitioner whose marker is "the ability to transform negative emotions within minutes rather than hours."2 Speed of dissipation. On that measure, a person who stays angry at something worth being angry at has failed the month.
They score the same behaviour opposite ways, which is the test for whether two framings genuinely conflict. And the conflict lands hardest on the one case both documents care about — a real injustice, witnessed, requiring response.
The corpus did not invent this tension. It reproduced one that runs through its primary source, and the source's version is sharper.
Ramlal's treatise gives seven registers of one devotion — memory, vow, joy, speed, roar, battle-cry and wrath — with wrath as the seventh, presented as a pitch of devotion rather than a failure of it.3 Nothing there suggests wrath is a stage to be transited.
The Śaṅkarācārya of Dvārakā, in the same volume, treats Hanuman's nature as fundamentally sāttvika and warns against directing his fierce worship toward māraṇa or mohana.4 And Sarvānand Pāṭhak licenses the six fierce rites only when redirected inward, against desire, anger, greed, delusion, arrogance and envy — which is the conversion account exactly: the anger is not aimed at an enemy, it is aimed at your own anger.4
Three settings of one dial, in one book, unreconciled. This build has already recorded that as a preserved contradiction of the Aṅka. What is new here is that the extension corpus inherited both ends of it without noticing, and now transmits the incoherence forward.
The conversion account rests on a move that needs checking before it goes any further, because the vault holds the relevant material.
Raudra→Śānta, Bhayānaka→Vīra, Śṛṅgāra→Bhakti are written as progressions — one rasa converting into another as the outcome of a practice.2
That is not what rasas are. In the Nāṭyaśāstra system the rasas are aesthetic registers, the flavours an experience can carry. They are a palette, not a ladder. Raudra does not mature into Śānta any more than red matures into blue; a work can be in the raudra register or the śānta register, and a person can move between them, but the movement is a change of key rather than an advance.
So the arrows are doing something the system does not license. What the corpus appears to want is sequencing — first this state, then that one — and it has borrowed a taxonomy of registers to express it, in the same move it made with the Gauḍīya bhakti terms.
The pattern is now visible twice: a real technical vocabulary is picked up and arrows are added between its terms. That is not misattribution, since no text is claimed to say it — but it is the second instance of the same habit, and a reader who knows the Nāṭyaśāstra will read the arrows as sourced.
Here is the detail that makes the whole thing stranger, and it is the strongest single observation on this page.
Months 10 to 12 of the same implementation document abandon sequence entirely. Month 10 is "sequential rasa orchestration", month 11 is "simultaneous rasa harmonisation — holding multiple emotional states like archetypal chords", and month 12 is "spontaneous constellation response".2
Archetypal chords. Multiple rasas held at once, as a chord rather than a sequence — which is exactly the structure the kastvam verse gives for the three devotional registers, and exactly the structure the corpus's bhakti scheme refuses by arranging its three stages as a ladder.
So the same author, in the same corpus, reaches simultaneity in the rasa domain and imposes sequence in the bhakti domain. And in the rasa document the simultaneity is placed at month eleven of thirty-six — an advanced attainment — while the sequence-imposing bhakti scheme treats its own third stage as the destination.
He had the right shape. It is in his own vocabulary, in his own file, applied to the neighbouring material.
The stake is not theoretical, because the corpus has a practice at the far end of this.
Its Level Three protocol, Embodying Divine Wrath for Dharmic Protection, directs the raised energy outward, toward "the situations or forces that require protective response", for thirty to forty minutes, using the twelve-syllable tantric mantra.5
That is unambiguously the integration account. The anger is preserved, amplified, aimed.
But the curriculum that gates access to it runs on the conversion account, and rewards minutes-to-dissipation as its month-three recognition sign.2
So a practitioner is trained for two months to make anger disappear quickly, assessed on that, and then advanced into a protocol whose entire method is to sustain and intensify it. The preparation trains the opposite capacity to the one the practice requires, and no document in the cluster notices, because no document holds both.
Worse: on the conversion account's own logic, the practitioner best qualified to reach Level Three — fastest at dissolving anger — is the one whose training has least prepared them to hold it. The gate selects against the skill.
Sorted by what each is for, they stop competing and become a choice the corpus could have made explicitly.
Conversion is a regulation technology. It suits the practitioner whose anger is frequent, disproportionate, and directed at things that do not warrant it — which describes most anger most of the time, and describes the corpus's stated reader, who lives with a mood disorder and a condition characterised by difficulty regulating anger.
Integration is a mobilisation technology. It suits the practitioner whose problem is the opposite: real injustice, and a devotional training that has left them without a permitted response to it.
Those are different patients. A corpus that offers both without diagnosing which one it is talking to has handed a regulation problem an amplifier, or a mobilisation problem a sedative, with equal probability.
Naming the two and stating who each is for would cost a paragraph. That is the whole repair.
⛑⛑ SOURCE-UPDATE, 2026-08-18 — A THIRD POSITION, AND IT IS THE ONLY ONE SPOKEN BY THE AGENT ABOUT HIS OWN CONDUCT. Amirapu Natarajan's 2008 text has Hanumān deliver the verdict himself, immediately after firing Laṅkā.6
⭐⭐ *"I have done a terrible mistake in burning down this city in a fit of anger. A man who is unable to conquer his anger will commit any sin. He may even kill his Guru or insult a man of virtue."*6
⚑ Neither of the two accounts above has this. ⭐ The regulation account treats anger as a state to be managed; the integration account treats it as an energy to be directed. ⚠ This one treats a specific act as a lapse, in the first person, with the general principle stated as a consequence of it.
And it goes further than regret: he contemplates self-immolation — "now, I too will jump into this fire and kill myself" — and reasons his way out on the ground that Agni did not burn his tail, so would not have burned Sītā.6
⭐⭐⭐ Then the definition, and it is the most portable thing in the passage: "just as a snake sheds its old skin, whoever casts off his anger is called a PURUṢA" — with the book's own note that the term applies to women equally, as with puruṣakāra.6
⚑ That converts anger from a state into a shed. ⭐ Not suppressed, not channelled, not integrated — discarded, and the discarding is what confers the status. ⚠ Which is a third mechanism, incompatible with both of the above, and it comes with Vivekananda's flat maximalism attached: "anger is a great sin, even if it is righteous" — an explicit closing of the righteous-anger exemption that the integration account depends on.6
🔴 Three positions now, in three sources, none citing the others. Preserved.
When one body of work gives two treatments of the same material, do not average them — find the case where they prescribe opposite actions and check which one the source's own downstream practice runs on. Averaging produces a mush that is neither, and it hides the choice. The divergence case here is a real injustice, witnessed, and the answer is that the flagship practice runs on integration while the curriculum grades on conversion.
Then ask the diagnostic question the corpus skipped: which failure is this reader's? Too much anger too easily, or too little where it is warranted. Regulation and mobilisation are opposite interventions and the same vocabulary — transforming anger, working with anger, sacred rage — covers both, which is exactly why the confusion survives.
Two personal-notes documents for the two accounts,12 one for the practice they gate,5 and two Aṅka essays for the source-side split.34
Preserved, not resolved. This page picks neither account. Both are in the corpus, both have a defensible use, and CLAUDE.md's constraint against silently resolving contradictions applies with force here because the resolution would be a clinical judgement about a specific reader that this vault is not positioned to make.
Bounded. The claim that no document in the cluster holds both accounts is a claim about the thirteen documents read, 13 of 106. A rasa-focused document elsewhere in the corpus — the Advanced Hanumanic Rasa Alchemy material is named in the folder listing and was not read — is where a reconciliation would most plausibly live.
Open question. Does the Raudra→Śānta pairing come from the Nāṭyaśāstra rasa system or is it the corpus's construction? The vault holds a Nāṭyaśāstra corpus, and rasa theory does not conventionally treat the rasas as a transformation sequence with one converting into another — they are aesthetic registers, not stages. If the pairing is invented, that is worth knowing before it reaches a book.
The two documents were written for different purposes and it shows in their machinery.
The Shadow-Devotion document is a topic generator — five deep-dive ideas for content, each with a core insight and a why-it's-high-value line. It is arguing for the interest of a position.
The implementation document is a schedule. Weeks, months, gates, recognition signs. It is telling somebody what to do on Tuesday.
A position that is interesting to write about and a practice that is safe to run are different products, and the corpus's failure is filing them in one folder without a boundary. The integration account makes better content — sacred rage, shadow alchemy, the resolution of a split in contemporary spirituality. The conversion account makes better practice for the reader described. The thing that got optimised was not the practice.
Seven Registers of One Devotion. Ramlal's taxonomy is the integration account in its strongest form — wrath as the seventh pitch of a single devotion, not a lapse and not a stage — and the volume prints the collateral cost beside it rather than resolving it away. What the pairing produces is the reason the corpus's two accounts cannot be averaged: the tradition's integration account comes with a casualty list, and its conversion account comes with a prohibition, and each account's safeguard is unusable by the other. Inward redirection is meaningless if the wrath is meant to stay aimed outward; a casualty list is meaningless if the wrath is meant to dissolve in minutes. A practitioner running a blend of the two has the safeguards of neither, which is the state the cluster currently leaves them in.
Discernment as the Real Skill. The volume's own answer to when is force the right instrument is not a doctrine about anger at all — it is the honest-obstacle criterion, where Surasā is converted and Kālanemi is killed and the variable is whether the obstacle's presented self is real. The pairing yields what both of the corpus's accounts are missing: the question is not what to do with your anger, it is what kind of thing you are facing. Both corpus documents locate the decision inside the practitioner's emotional processing. The tradition locates it in an assessment of the object — which is checkable, can come out wrong, and does not require the angry person to be a reliable witness to their own motives.
The genuinely hard version of this, which neither account faces.
Some things warrant sustained anger for years. A practice that measures success in minutes-to-dissipation will train that out, and will feel like progress while doing it. A practice that preserves and amplifies it will produce somebody carrying a live charge for years, which is its own damage and is not obviously better.
The tradition's figure had a target supplied by somebody else and an end date supplied by the war. Neither condition is available to a person with a grievance and a meditation cushion, and both of the corpus's accounts quietly assume one of them.